INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI films
the text directly from the original or copy submitted. Thus, some thesis and
dissertation copies are in typewriter face, while others may be from any type of
computer printer.

The quality of this reproduction is dependent upon the quality of the
copy sutunitted. Broken or indistinct print, colored or poor quality illustrations
and photographs, print bleedthrough, substandard margins, and improper
alignment can adversely affect reproduction.

In the unlikely event that the author did not send UMI a complete manuscript
and there are missing pages, these will be noted.

Also, if unauthorized

copyright material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, beginning at the upper left-hand comer and continuing
from left to right in equal sections with small overlaps.

Photographs included in the original manuscript have been reproduced
xerographically in this copy.

Higher quality 6" x 9” black and white

photographic prints are available for any photographs or illustrations appearing
in this copy for an additional charge. Contact UMI directly to order.

ProQuest Information and Learning
300 North Zeeb Road. Ann Arbor, Ml 48106-1346 USA
800-521-0600

UMI

THE PERCEPTION OF EMOTIONAL EXPRESSIONS IN INCARCERATED YOUTH
by
Evan Vike
B.A. (Honours), Trinity Western University, 1991

THESIS SUBMITTED IN PARTIAL FULFILMENT OF
THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE
in
PSYCHOLOGY

© Evan Vike, 2001
THE UNIVERSITY OF NORTHERN BRITISH COLUMBIA
March 2001
All rights reserved. This work may not be
reproduced in whole or in part, by photocopy or
other means, without the permission o f the
author.

1^1

National Library
of C anada

Bibliothèque nationale
du C anada

Acquisitions and
Bibliographic Services

Acquisitions et
services bibliographiques

395 WaWngton Street
Ottawa ON K1A0N4
Canada

395, rue Wellington
Ottawa ON K1A0N4
Canada
Y o u rm YotmrUé ftne»

Ouril» NotmréfétmKê

The author has granted a non­
exclusive licence allowing the
National Library of Canada to
reproduce, loan, distribute or sell
copies of this thesis in microform,
paper or electronic formats.

L’auteur a accordé une licence non
exclusive permettant à la
Bibhothèque nationale du Canada de
reproduire, prêter, distribuer ou
vendre des copies de cette thèse sous
la forme de microfiche/film, de
reproduction sur papier ou sur format
électronique.

The author retains ownership of the
copyright in this thesis. Neither the
thesis nor substantial extracts from it
may be printed or otherwise
reproduced without the author’s
permission.

L’auteur conserve la propriété du
droit d’auteur qui protège cette thèse.
Ni la thèse ni des extraits substantiels
de celle-ci ne doivent être imprimés
ou autrement reproduits sans son
autorisation.

0-612-62542-7

CanadS

APPROVAL

Name:

Evan Vike

Degree;

Master of Science

Thesis Title:

THE PERCEPTION OF EMOTIONAL EXPRESSIONS
IN INCARCERATED YOUTH

Examining Committee:

Chair: Dr. Don Munton
Professor and Chair, International Studies Program
UNBC

C. 0 xW rL

A Û Û -

Supervisor: Dr. Glenda Prkachin
Associate Professor, Psychology Program
UNBC

Conlimftee Member: Dr. Kenneth Prkachin
Professor and Chair, Psychology Program
UNBC

Committee Member: Glen Schmidt, MSW
Assistant Professor and Chair, Social Work Program
UNBC

ExtemafExaminer: Dr. Tm s^trong
Assistant Professor, Education Program
UNBC

Date Approved:

Abstract
The visually rich facial expressions o f emotion continue to be an intriguing phenomenon
in the study o f human emotion neuropsychology. Emotions have many important functions,
some o f the most significant being their role in the dynamics of communication and behaviour
and the way in which they influence the decision making process and guide behaviour. Given
the significance of emotion within the social context, it is important to understand how its
diverse functions are mediated and coordinated effectively in facial expressions of emotion.
Moreover, it is important to understand how the adequacy of perception o f facial expressions of
emotion might contribute to aberrant behaviour, such as demonstrated by antisocial youth.
The aim o f this study was to explore if the social difficulties experienced by incarcerated
youth are related, in part, to some deficit in perceiving socially important information conveyed
by facial expressions o f emotion. Furthermore, it sought to investigate the underlying structural
or functional impairment of processing facial emotion within this subpopulation. Inasmuch as
alexithymia has been associated with functional problems in accurately detecting facial
emotion, this study sought to determine if alexithymia was a contributing factor in perceptual
performance. In phase 1, thirty-two incarcerated youth between the ages o f IS and 18 years
were asked to detect six target facial expressions o f emotion under temporal constraints. In
phase 2, the detection performance o f the incarcerated youth was tested against the detection
performance o f 31 non-incarcerated youth, aged 17 to 18 years.
In phase I, detection performance by the incarcerated participants did not differ
significantly regardless of the degree o f alexithymia (low, intermediate or high) or possible
affect problems resulting firom their incarceration. In phase 2, the incarcerated participants

ii

performed significantly poorer in accurately detecting happy, surprise, disgust and anger facial
expressions; however, alexithymia was not significantly associated with the perception
performance o f either sample.
It was discussed that the underlying basis for the poorer performance by the incarcerated
participants was likely due to a functionally deficient attentional system when subjected to time
demands. It was also discussed that alexithymia may contribute to behavioural issues
demonstrated by incarcerated youth.

m

TABLE OF CONTENTS
Abstract.................................................................................................................................... ii
Table of Contents.....................................................................................................................iv
List of Tables.............................................................................................................................vii
List of Figures........................................................................................................................... viii
Acknowledgement...................................................................................................................ix
Introduction.............................................................................................................................. 1
Chapter One

Literature Review o f Facial Emotion.

.3

Defining Emotion............................................................................ 3
The Expressive Face........................................................................ 5
Perception of Facial Em otion......................................................... 6
Attention Processes.................................................................. 6
Perception of Facial Expressions o f Emotion..........................8
Perception o f Facial Identity and Emotion: Separate Processes

11

Development o f Facial Emotion...................................................... 12
Emotional Regulation and Dysregulation........................................ 14
Hemispheric Asymmetry o f Facial Emotion...................................18
Neurological Substrates o f Facial Emotion.....................................19
Neurophysiological Associations with Facial Emotion.................. 24
Alexithymia and the Perception of Facial Emotion
and Emotional Regulation............................................................25
Literature Review o f Neuropsychological Deficits
Associated with Incarcerated Youth............................................................29

iv

Associated Neuropsychological Deficits..........................................30
Attention Deficit Hyperactivity Disorder.........................................31
Learning and Language Deficits....................................................... 34
The Biology o f Difficult Temperament: Emotion Dysregulation.... 36
Central Nervous System Anomalies.................................................39
Detecting Facial Emotion................................................................. 39
Purpose of Study and Hypotheses................................................................40
Chapter 2

Methods.........................................................................................................42
Phase 1: Incarcerated Sample....................................................................... 42
Participants....................................................................................... 42
Measures........................................................................................... 42
Materials........................................................................................... 44
Design and Procedure....................................................................... 45
Debriefing......................................................................................... 46
Phase 2: Incarcerated and Non-incarcerated Sample....................................47
Subjects............................................................................................. 47
Measures, Materials, Design and Procedure.....................................47

Chapter 3

Results...........................................................................................................48
Phase I: Analysis for Incarcerated Sample...................................................48
TAS-20 and PANAS Descriptive Statistics......................................48
Detection Index (A’) Analysis..........................................................49
Phase 2: Analysis for Between Samples....................................................... 55
TAS-20 Descriptive States............................................................... 55
V

Detection Index (A’) Analysis.......................................................... 56
Chapter 4

Discussion.....................................................................................................64
Phase 1...........................................................................................................64
Phase 2...........................................................................................................66
Biological Explanations................................................................................ 69
Theoretical Implications of a Deficit in Detection Performance................. 74
Alexithymia and Young Offenders............................................................... 77
General Theoretical Implications From The Study......................................80
Limitations O f The Study............................................................................ 81
Conclusions...................................................................................................83
Future Research.............................................................................................84

References.................................................................................................................................. 86
Appendix A - Positive and Negative Affect Schedule (PANAS) Questionnaire.................... 98
Appendix B

- Toronto Alexithymia Scale-20 (TAS-20) Questionnaire............................100

Appendix C

- Consent Form............................................................. 104

Appendix D - Table 3.6 Two Tailed T-Test for Paired Emotion Comparisons......................106

VI

List of Tables
Table 3.1

Descriptive Statistics for TAS-20 Alexithymia Groups for Incarcerated Youth

Table 3.2

Descriptive Statistics for PANAS Groups for Incarcerated Youth

Table 3.3

Descriptive Statistics for Emotion Detection Sensitivity A1 by PANAS Group for
Incarcerated Sample

Table 3.4

Descriptive Statistics for FEE Detection Sensitivity A/ for TAS Group for
Incarcerated Sample

Table 3.5

Two Tailed T-Tests for Paired Emotion Comparisons for Within-Subject Effects

Table 3.6

Descriptive Statistics for TAS-20 Alexithymia Groups by Sample

Table 3.7

Descriptive Statistics for FEE Detection Sensitivity A1 for TAS Group by
Sample

Table 3.8

Descriptive Statistics for the Between Sample Effects of Emotion (A3

Table 3.9

Descriptive Statistics for TAS-20 Alexithymia Factors by Sample

Vll

List of Figures

Figure 3.1

Mean Detection Sensitivity A1 for Samples

Figure 3.2

Within-Subject Emotion by Sample Interaction Effect for Fear, Anger, Disgust
and Sad Between Incarcerated and Non-incarcerated Samples

vni

Acknowledgements
I should like to thank ray thesis coraraittee. Dr. Glenda Prkachin, Dr. Ken Prkachin and
Dr. Glen Schmidt for their assistance in this process. I ara grateful to ray thesis supervisor. Dr.
Glenda Prkachin, for her contributions and insightful criticisms.
I also wish to acknowledge the residents o f the youth custody centre who participated
in this study.
Finally, I thank ray family for their support.

IX

Introduction
Emotion and its associated phenomena continue to be an intriguing and fascinating
aspect of scientific study in human neuropsychology. Emotions and their expressions in speech,
literature, art, and theater have been part of human experience throughout recorded history
(Ginsburg & Harrington, 1996). Whenever people interact, they experience, express, and
perceive emotion in what is a complex and often seemingly unconscious process (Strongman,
1996), providing information that requires various levels of analysis to disentangle pertinent
stimuli. The attention to emotion that is evident throughout history bespeaks something
significant with regard to the human organism. O f the potential expressive modalities o f
emotion, the visually rich facial expressions of emotion (FEE) have been the subject o f much
empirical research and are probably the best understood (Gottman, 1993; Oatley & Jenkins,
1992).
Several functions of emotion are noted in the literature. For example, emotion plays a
significant role in communication in the social constructs of society and acts as a signal o f one’s
internal state and functioning, configuring mental resources and making ready for certain kinds
of action (Damasio, 1994; Hess, Blairy, & Kleck, 1997; Singh et al., 1998; Thompson, 1988).
Emotions communicate to others, resulting in changes to the dynamics of interaction, from
cooperation to withdrawal, and conflict to deference (Oatley & Jenkins, 1996). Emotion is
instrumental in the decision making process and helps one formulate useful solutions (Damasio,
1994; Rolls, 1990; Siminov, 1997; Stein & Levine, 1989). This is particularly true in instances
where there are competing concurrent goals, emergencies and temporal demands, and the
resources o f rational thought are too limited, too slow, or too error-prone to solve problems,
prioritize and coordinate behavior in a manner that is superior to random choice (Johnson-Laird
& Oately, 1992). Moreover, emotion is instrumental in guiding behavior for self-preservation

1

(MacLean, 1993; Rolls, 1990), dealing with fundamental life-tasks, and promoting immediate
attention to important interpersonal encounters (Ekman, 1992b).
Emotion also functions in the communication o f emotional states, playing a survival
role, and in assisting in the ongoing evolution o f what is significant for the individual or group,
and the stability of social organization (Rolls, 1990; Sprengelmeyer et al., 1987). Emotion
impacts the social attachment between parents and their young that serves to increase the
likelihood of the next generation’s survival (Rolls, 1990). Furthermore, emotion can affect the
cognitive evaluation o f events and memories and their interpretations, storage, and recall (Rolls,
1990).
Given the significance of emotion to human beings, and its stature within the social
context, it is important to understand how its complex and diverse functions are mediated and
coordinated effectively in facial expressions of emotion. Moreover, it is important to
understand how the adequacy of accurate perception o f facial expressions of emotion might
contribute to aberrant behaviour, such as demonstrated by antisocial youth.
To the author’s knowledge, there has been no published research about the potential
relationship between the neuropsychology underlying the perception of FEE and the associated
neuropsychology underlying the troubled behavior o f incarcerated youth. The aim o f this study
was to explore potential neuropsychological links between these two areas.

CHAPTER ONE
Literature Review of Facial Emotion
Defining Emotion
Characterizing and defining human emotion has encompassed the evocation, modulation
and combination o f phenomena such as cognition, feelings, visceral, neurological, and
biochemical reactions; expressive displays (e.g., vocal, facial, posture and other overt behavior);
and relating internal and external stimuli (Cacioppo, Klein, Bemtson, & Hatfield, 1993). The
aggregate information in the literature provides plausible conceptual explanations and
definitions of emotion.
Facial expressions, some of the better studied expressive phenomena of emotion, have
been most often characterized and supported by research as physiologically discrete categories
(e.g., the happy feature configuration is discretely different fi’om sad) (Calder et al., 1996;
Damasio, 1994; Ekman, 1992a; Etcoff & Magee, 1992; Young et al., 1997). Some emotions
(i.e., happiness, sadness, fear, disgust, surprise, and anger) have frequently been described and
supported in the research as innate, evolutionarily prewired, basic, and universal (Damasio,
1994; Ekman, 1992a; Johnson-Laird & Oatley, 1992; McNaughton, 1989; Thompson, 1988).
Other emotions are considered more complex, such as secondary emotions (e.g., variations and
nuances o f primary or basic themes) that result from greater cognitive elaborations, more
adaptively flexible and individualized, and experientially learned (Damasio, 1994). Feelings are
another factor in defining emotion. Feelings are sometimes referred to as emotion (LeDoux,
1994; Thompson, 1988) or physiological activity relating to emotion, rather than emotion per se
(Damasio 1994; Thompson, 1988). Given the current information about emotion, it is
reasonable to consider that FEE are probably more accurately conceptualized as categorical, at

least a few of which are primary (predispositional) and universal in nature (i.e., sad, angry,
happy, surprise, disgust, and fear) and have associated feelings.
Theories of emotion differ regarding whether emotional evaluation of events is
considered to be automatic, preceding conscious evaluation, or if conscious evaluation is a
necessary antecedent of emotion (Damasio, 1994; Ekman, 1992a; Halgren, 1992; LeDoux,
1993a; Johnson-Laird & Oately, 1992; Thompson, 1988). Current neuroscientific research has
found no compelling evidence that emotion and cognition are necessarily opposed; rather, they
are likely juxtaposed or take place synchronously. Emotion and reason are at some point
interwoven and emotional evaluation occurs within the cognitive system (Damasio, 1998;
Halgren, 1992; Lane, Kivley, Du Bois, Shamasundara, & Schwartz, 1995; Rolls, 1990;
Strongman, 1996). This close consequential interaction of emotion and cognition makes it
difficult to dissociate the two, as the degradation o f one seems to affect the other (Halgren,
1992).
Emotion has been further characterized as:
.. .a collection of changes in the body state that are induced in myriad organs by nerve
cell terminals, under the control o f a dedicated brain system, which is responding to the
content of thoughts relative to a particular entity or event. Many changes in the body
state-those in skin color, body posture, facial expression, for instance-are actually
perceptible to an external observer...Other changes in body state are perceptible only
to the owner of the body in which it took place. (Damasio, 1994, p. 139).
Although clearly defining emotion continues to be the subject of considerable
discussion, a substantial broad agreement seems to exist on some of its characteristics (Izard,
1993). For instance, emotion involves particular or specialized neural processes rather than
undifferentiated generalized processes. Also, emotion is defined as an expressive or efferent
activity in the central nervous system (e.g., facial expressions), and as registering in some level
of consciousness (e.g., feelings, motivations, perceptions and attributions such as: sad, happy.

angry, etc.). Furthermore, there are likely a small set o f FEE (i.e., sadness, happiness, anger,
disgust, fear, and surprise) that are highly recognizable and appear to reflect a universal
emotional phenomenon (Ginsburg & Harrington, 1996).
The Expressive Face
Faces form the source of a multitude o f inferences. From the face we are able to
determine identity, gender, age, and what one is trying to convey. Regarding emotion, FEE in
humans has been the subject of extensive empirical research (Gottman, 1993; Oatley & Jenkins,
1992). Emotional facial expressions are effective communicators despite significant language
and cultural differences; they provide coarse evaluations of one’s underlying emotional state
(Johnson-Laird & Oatley, 1992), are highly salient and easily detected (Prkachin & Prkachin,
under review), and are able to communicate social information rapidly and with definition
(Etcoff, 1984). This is not surprising, as the human organism is highly dependent on social
interaction for survival and relies on its ability to recognize and distinguish the information that
is relayed.
Facial expressions of emotion are produced by the combinations o f varied and specific
movements of facial skin, movements of the facial musculature and so on (Damasio, 1994;
Tassinary & Cacioppo, 1992). These movements produce wrinkles, folds, and lines, and
change the position o f the brows and the comers o f the mouth. The resulting expressions are
thought to have distinct patterns and to be qualitatively different firom one another, configuring
a specific emotion category, at least for the basic emotions (Ekman, 1992b; Etcoff & Magee,
1992).
Muscle displacement involved in facial expressions has also been identified as important
support for the configuration o f categorical FEE. For example, expressions tend to be
considered smiles when the distance between the comers o f the mouth and the eyes is less than

40% of the maximum distance, a distinct and qualitative difference from other emotions (Hess
et al., 1997).
Perception o f Facial Emotion
Attentional Processes
Considering that perception of facial emotion starts with understanding the basic notion
of attentional processes, it has been suggested that attention involves a moment to moment shift
in awareness between the information provided by the external environment and one’s
experience; for example, the awareness of sensations, perceptions and conceptions o f emotion
about the self and others (Prkachin, in preparation). As the attentional system has a limited
capacity, efficient and effective functioning presumes the need for it to be selective about what
is brought into awareness (Flavell & Miller, 1998). It is well accepted in the literature that
attention is a selective rather than a random process. From birth, babies selectively attend to
their environment, displaying a preference to explore some aspects over others (Butterworth,
1998). One such early selection relates to faces. This early attention bias supports the notion
that development of specific brain circuitry acts in concert with the species’ typical environment
to bias the input o f later developing circuitry. Neurophysiological recording reveals that
individual cells code diverse specific characteristics (or ‘trigger features’) o f the environment
(Bomstein, 1992). These features, for example, wavelength, orientation o f form, movement and
so on, stimulate specific neurons in the visual system. It is this process o f selective attention
that enables the person to isolate informative aspects o f stimulus variation from the
environment, thus providing the differentiation o f perception (Butterworth, 1998).
As it relates to the external environment, attention is normally focused where one looks,
or towards the direction in which one’s eyes are fixated. Thus, attentional processes and visual
orienting most often work together to bring objects in the central foveal visual region in line for

further scrutiny (Motter, 1998). This visual scrutiny is a dynamic interaction between the visual
signals entering the brain and a variety o f internal biases functioning to ensure the processing of
behaviorally relevant rather than irrelevant stimuli. Some o f these biases are hard-wired into
the visual system (i.e., the possibility o f attentional templates) to facilitate the processing of
certain stimulus configurations (Corbetta, 1998). Given the hypothesized universality and
primacy of basic FEE, it is reasonable to consider that basic FEE fits within the notion of
attentional templates, at least to some degree, and with regard to some emotions more than
others. To accomplish all of the above requires an attentional system with at least three
operating functions: alerting or vigilance, orienting, and target detection.
Such a system has been proposed in the literature (Jackson, Marrocco, & Posner, 1994;
Posner & DiGirolamo, 1998; Swanson et al., 1988). There are two attentional subsystems, the
anterior attentional system (AAS) and posterior attentional system (PAS), comprised o f three
networks. The alerting network functions to maintain an organism’s readiness to react to
stimuli by suppressing background neural noise through the inhibition o f irrelevant mental
activity. This network involves brain regions centered in the right frontal lobe and right parietal
lobe. The orienting network functions to motorize special neuronal operations needed to bring
attention to the relevant visual location (selective visual search) and bind signals into object
perception. The brain regions involved are purported to be centered in the posterior parietal
lobe and thalamus. The executive network coordinates multiple specialized neural processes,
for example, target detection (color, form, category), to direct behavior. The brain regions
involved are purported to be those centered in the anterior cingulate, left lateral frontal lobe, and
basal ganglia. These functions o f the attentional system become associated with tasks for the
purpose of boosting signals in various brain regions to increase the probability o f relevant signal
detection. Thus, attention selects the content in one’s current awareness.

In this attentional model, the basal ganglia and anterior cingulate are significant in target
detection. The basal ganglia contribute through their manifold connections to the cortex via a
cortico-striate-thalamo-cortical circuit dedicated to common information processing (Jackson et
al., 1994). In this way the basal ganglia act as a comparator o f coded spatial representations by
the AAS and PAS; and, modulate the circuitry o f the AAS by enhancing relevant signal strength
and suppressing irrelevant signals. The anterior cingulate contributes target detection by
boosting activation and reactivation of signals for feature selection in extrastriate regions
(Posner & DiGirolamo, 1998).
Research on Attention Deficit Hyperactivity Disorder (ADHD) provides support for this
model of an attentional system (Swanson et al., 1998). Studies using event related potentials
(ERPs) suggest abnormalities in the right fi-ontal region, implicating the alerting network, and
the right parietal region, implicating the orienting network. Research using imaging technology,
such as positron emission tomography (PET), suggests abnormalities in the frontal lobe,
implicating both the alerting and executive networks. Studies using magnetic resonance
imaging (aMRl) show a smaller than normal size corpus collosum, basal ganglia, and right
frontal lobes, thus implicating all three networks (orienting, alerting, and executive
respectively). ADHD symptomatology involved with inattention suggests poor sustained
attention related to an alerting deficit, poor selective attention related to an orienting deficit, and
poor stimulus detection related to an executive function deficit. As can be gathered from the
above discussion, accurate perception requires an adequately functioning attentional system to
gather relevant task information.
Perception of Facial Expressions of Emotion
Given that the face provides a great deal o f nonverbal information, it is reasonable to
consider from an adaptive and evolutionary perspective that at least some FEE would be more

8

immediately perceptible. It has been argued that overt expressions would be o f little value if
humans failed to decode and respond to the display (Dimberg, 1997). In keeping with this, it
has been proposed that FEE perception (e.g., detection) is to some degree preattentively and
automatically accessible to the perceiver (White, 1995). In other words, humans may be
hardwired for the detection of FEE. Various expressions of emotion are thought to reflect, for
example, differing spatial relationships among facial features (Damasio, 1994; McKelvie, 1995;
Tassinary & Cacioppo, 1992). This perceptual process can be conceptualized as begitming with
the parallel scanning of the visual field for biologically relevant features (i.e., configural-feature
processing of colour, contours, size, and spatial relationships) which are then conjoined to form
the percept of the face (Coren & Ward, 1989). The determination of FEE perception accuracy
would likely depend on one’s ability to inhibit the automatic, preattentive processing of
unwanted (nontarget) encoded information.
Further to understanding the process of FEE perception, it has been suggested
(Desimone, 1991) that such perception of facial expression is a specialized neural mechanism
that evolved to facilitate social communication independent o f the larger pattern perception
process, wherein facial emotion is detected in terms o f discrete categories (Etcoff & Magee,
1992). In line with this notion, one theoretical assumption is that continuous perceptual
information is transferred fiom the nearest prototypical category to the underlying emotion
processing system until discrimination is made (Etcoff & Magee, 1992; Young et al., 1997).
Research (Prkachin & Prkachin, 1994, under review) using adaptation tests (procedures that
diminish the ability to detect FEE) has shown that adaptation selectively disrupts the perception
o f happiness, sadness, anger, disgust and surprise. Once a person has been repeatedly exposed
to a particular FEE in a short period o f time, perception o f that FEE is significantly reduced;
however, the perception of other FEE is not affected, thus providing evidence that FEE can be

dissociated as categorical. This categorization process is likely subserved by specialized
mechanisms.
Another possibility is that the perceptual processing o f FEE could be either categorical
(emphasizing configurai, or perceptual-whole, processing) or featural (emphasizing specific
compontential processing; e.g., nose, lips). A possible determinant o f the processing mode may
be spatial fi-equency. Low spatial firequencies are thought to be mediated by neurons with faster
conduction velocities, whereas the information contained in high spatial frequencies is mediated
by neurons with slower conduction velocities (Kirita & Endo, 1995). Therefore, facial
expressions o f emotion which contain many low frequency spatial qualities (less complex
encoded information) are more likely to be processed configurally (faster temporal processing),
and facial expressions of emotion which contain many high frequency spatial qualities (more
complex encoded information) are more likely to be processed componentially (slower temporal
processing).
The literature provides some support for this notion, suggesting variability in the various
aspects of perception (e.g., detection, recognition, and labeling o f FEE) (Hess et al, 1997;
Prkachin, in preparation). That is to say, we do not attend or respond to all FEE in exactly the
same way and the processes engaged by each emotion expression may be different. For
example, happy faces are recognized more quickly and accurately than any other facial
expression (McKelvie, 1995), followed by sadness suggesting, configurai processing, while
fear, surprise, anger and disgust are more difficult to detect (Prkachin & Pricachin, 1994)
suggesting componential processing. Although research continues to explore the variable
processing question, what can be gleaned fit>m the above information is that FEE are
qualitatively different fix>m one another and thus perceived at different speeds and accuracy.

10

Perception of Facial Identity and Emotion: Separate Processes
There are at least two types of configurai information thought to be perceived in the face
that determine face identity (FI) recognition. These are the appropriate or expected location of
facial features and spatial or relational configuration o f features (Baenninger, 1994). Perceptual
identification involves the extrapolation of particularities such as the specific locational and
relational feature configuration. For example, the nose is always in the same place on the face.
There are also variable relationships that make a face unique from a more relative configuration
common to all faces (Sergent, Ohta, & MacDonald, 1992). Maintaining the integrity of FI,
there are many potential categorical dynamic spatial configurations (i.e., FEE) that are
perceived as information about the emotional state of the person (McKelvie, 1995). Although a
single facial component (upturned comers of the mouth) may provide some clue to a particular
FEE (happiness), the perception o f facial affect is dependent upon interaction of the moving
internal features of the face (e.g., brow, eyes, and mouth). Therefore, one could consider that
this configurai information is ultimately more important than any particular feature for both FI
and FEE perception. Furthermore, one could assume that FI and FEE perception result from the
same process.
It has been proposed that the recognition of FI and FEE share at least an early processing
stage based on configurai information (McKelvie, 1995) and are thus subserved by the same
anatomical system. However, the literature also poses a convincing argument about a double
dissociation between the processing of FI and facial affect, suggesting that the two are
subserved by two anatomically separable neural systems (Adolphs, Damasio, Tranel, &
Damasio, 1996). Research involving monkeys provides evidence that suggests that facial
expression and identity are coded by separate populations o f cells in the brain (Desimone,
1991). While identity appears to be coded by neurons in the inferior temporal gyrus, facial

11

expression is thought to be coded by neurons in the superior temporal sulcus (STS) (Desimone,
1991). Further evidence for this dissociation comes from recent PET research that found that
neurons in superior temporal sulcus (STS), in either hemisphere, do not apparently significantly
participate in the processing o f FI in humans (Sergent et al., 1992). Other recent evidence using
functional magnetic resonance imaging (fMRl) showed that FI in humans elicited a stronger
response in the fusiform gyrus then in the STS (Hoffinan & Haxby, 2000), and that dynamic
aspects o f the face such as eye and mouth movement significantly activated a region of the STS
(Puce, Allison, Bentin, Gore, & McCarthy, 1998).
Recent research demonstrated that adaptation was selective to the particular FEE
exposed, independent o f the individual face on which the expression occurred (Prkachin &
Prkachin, 1994). This result supports the physiological findings outlined above, suggesting that
neurons that respond to FEE are likely from different populations o f neurons for identity, and
that different FEE are selectively coded by different neurons in the STS. Thus one can
confidently conceive that the processing o f FEE and FI are likely parallel processes and appear
to be somewhat distinguishable.
Development of Facial Emotion
Although the specific temporal emergence of emotional displays and the specific
number of emotional displays present in young children are not exactly consistent across
studies, there is evidence that specific differentiated basic FEE, corresponding to adult
conceptions of what some specific FEE should look like, are present in infancy (Camras,
Holland, & Patterson, 1993; Fischer, Shaver, & Camochan, 1990; Lewis, 1993; Oatley &
Jenkins, 1996). At least three emotions are noticeable in the neonate (disgust, happiness,
interest) with others noticeable within the first year of life (Shaffer, 1996; Thompson, 1988) as
the facial muscle groups become more refined and coordinated (Brown, 1993). By the end o f

12

the first 3 years of life, a child has come to possess an elaborate and complex emotional system,
with the majority o f adult emotions having emerged and developed. Further emotions may
emerge and existing ones elaborate (Lewis, 1993), thus demonstrating both the innateness and
malleability of emotional development (McNaughton, 1989).
Although there is some variability across studies, basic emotions emerge in all normal
infants at roughly the same age and across cultures (Shaffer, 1996). This suggests strongly that
at least some emotions are basic and fundamentally biologically programmed rather than the
result of differentiation through experience alone (McNaughton, 1989). While exposure and
experience play an important role, this universal emergence o f basic emotions supports the
notion that humans are biologically wired to send certain emotional signals.
The ability to perceive FEE is also present in young children. Given that humans are
biologically wired to send certain facial emotional signals, it would be reasonable to consider
that a selective attention exists to detect and interpret these signals. It has been suggested that
innately available face detectors direct and control infant attention to faces (Slater &
Butterworth, 1997). Furthermore, infants appear to be attracted to faces because these visual
targets possess factors that are seen best by them (e.g., they are rich in contrast, moderate in
complexity, and have curvatures) (Shaffer, 1996). It has also been observed that by 4 months of
age infants demonstrate an ability to distinguish and respond to eye gaze, and that by no later
than 2 or 3 months of age infants display preferential head turning and eye tracking to follow
schematic versus scrambled face patterns (Johnson, 1998).
Research supports the conclusion that children also perceive specific FEE (Oatley &
Jenkins, 1992). For example, by 3 months o f age infants seem to discriminate photos of happy,
sad and angry faces and also act appropriately to natural displays of emotion, acting gleeful to
happy faces and distressed in response to angry or sad faces (Shaffer, 1996). Infants are likely

13

to turn away and play less if the mother looks sad and freeze if she looks angry, thus implying
perception and meaningful responses (Harris, 1995b). By 6 months o f age, infants discriminate
posed versions o f sad, happy and angry expressions (Haviland & Lelwica, 1987). Moreover,
the discrimination led these infants to demonstrate emotional expressions that mirrored the
emotions detected (Haviland & Lelwica, 1987). Also by 6 months o f age, infants can abstract
the invariant features of a happy expression regardless o f identity (Flavell & Miller, 1998).
Taken together, research suggests that in the first year of life children are able to at least
discriminate happy from sad and angry faces (Oatley & Jenkins, 1996).
Under the age o f 2 years, children definitely react to the emotions of others and start to
exhibit signs of empathy (Oatley & Jenkins, 1992). Children from the ages o f 4 to 6 years were
able to differentiate happy, sad, fearful and angry FEE at levels well above chance, and their
accuracy improved with age (Masters & Carlson, 1988). By age 8, children were able to make
subtle distinctions between closely related emotional states and identify the affective states of
others with some degree of accuracy (Brown, 1993). Furthermore, children at about age ten
were essentially perfect in their identification o f happiness, sadness, anger and fear (Masters &
Carlson, 1988). Collectively, the preceding discussion suggests that by adolescence a child
normally possesses a highly developed ability to express a variety o f facial emotions and a
perceptual system to reliably detect basic FEE (Brown, 1993). It also stands to reason that
flawed development may link components o f emotional development in unexpected ways so
that expected organization (e.g., expression, perception) may not occur consistently or perform
adequately (Harris, 1995b).
Emotional Regulation and Dvsreeulation
A well developed emotional system includes an ability to accomplish and experience a
balanced regulation of emotion. The implicit notion o f emotional regulation is that everyone

14

has a similar variety of emotions that function at optimal levels (i.e., intensities, duration,
patterns) (Oatley & Jenkins, 1996). From a differential perspective, basic emotions have three
fundamental, genetically determined components: conscious feeling o f emotion, FEE, and
neural substrate (Thompson, 1988). A basic feeling such as anger, the FEE o f anger, the neural
circuitry producing the physiological pattern associated with anger, and the stimulation needed
to elicit anger reflect a fundamental genetic influence. Consequently, it is reasonable to expect
that one’s emotional response to environmental stimuli can be understood at different levels.
For example, responses can be measured by way o f variable heart rate, cortisol levels, various
facial expressions and feeling states (Harris, 1995a).
Varying degrees of emotional regulation may occur within or between these factors. It
is interesting that by 2 years of age children use terms to denote emotion and that by age 3 make
causal statements of emotion (Oatley & Jenkins, 1992). As children age they use language to
understand and gain control over their emotions. Uninhibited anger and fhistration decreases
sharply in the second year of life and declines in the third as children learn to regulate their
emotion by communicating their feelings (Harris, 1995a).
The literature also notes that children 4 and 5 years o f age demonstrated appropriate
emotional regulation while at play (Oatley & Jenkins, 1996). For example, happy expressions
elicited gestures o f sharing, anger resulted in negative verbal and physical responses, and
expressions of hurt or pain garnered reassurance. From this it is evident that children respond
in different ways to different emotion displays.
The likelihood of an innate regulatory mechanism is evidenced by the capacity o f
emotions to regulate one another (Harris, 1995a). For example, anger attenuates fear and
sadness. This can influence perceptual and cognitive informational processing systems that
may in turn activate another emotion. Such emotional regulation in infants may explain some

15

o f their behavior, such as turning away from objects, self-soothing by sucking vigorously on
objects, rocking and so forth (Shaffer, 1996).
Temperament is also important in emotional regulation. There is agreement that
temperament is frmdamentally biologically based (Eme & Kavanaugh, 1995; Prior, 1992), with
constitutional differences in reactivity and self-regulation (e.g., facial expressions,
cardiovascular activity, and threshold parameters) (Derryberry & Rothbart, 1988). One’s
temperament organizes the expression of emotions and plays a role in how emotions are
expressed. Regarding FEE, children have characteristic styles or biases o f emotionality that
influence emotional patterns beginning in childhood and continuing through adulthood (Oatley
& Jenkins, 1996). Thus, the regulatory system of temperament plays a critical role in
coordinating attention and response and influences nearly every aspect of experience and
behavior (Derryberry & Rothbart, 1988).
A problem may arise when an individual’s repetitive pattern of emotional responding
becomes overly biased and the person develops a disruptive temperament. For example,
individuals who maintain a predominantly angry or fearful pattern o f responding, compared to a
flexible, adaptive emotional style, are often restricted in their emotional regulation and
maladapted in their disposition (Oatley & Jenkins, 1992).
Physiological evidence implicates the circuitry of the limbic system and reticular
activating system as involved in temperament (Derryberry & Rothbart, 1988). The feedforward
and feedback projections between these structures and the cortex accomplish multidimensional
adjustments of cortical pathways. That is, neural activity gives rise to subtle modulation
processes that adjust the target systems’ reactivity to incoming signals. The resulting pattern of
reactivity may serve to bias or narrow the flow o f information (Coren & Ward, 1989).

16

Approaching temperament in terms of the ‘systems’ reactivity’ suggests that ease and
flexibility of the attentional focus is disengaged and reoriented (Derryberry & Rothbart, 1988).
Failure at one stage of regulation development can have implications for the success of
regulating at other stages. This can result in an emotion response system that is less able to
effectively facilitate coping (Harris, 1995a) because of deficits that manifest in one or more
areas such as expression, experience, and perception of emotion (Brown, 1993). For example,
conduct disorder, and its adult counterpart, antisocial personality disorder, involve a
predominance of anger and are considered to be examples of chronic dysregulation of anger
(Oatley & Jenkins, 1992). The result is an inability to regulate emotional responses to
provocative situations and leads to social incompetence and maladaptation (Harris, 1995a).
Good examples of aberrant FEE expressiveness and recognition have been found in
studies o f depression. Individuals with depression generally display a bias toward expressing
more sadness. Furthermore, they demonstrate an impaired ability to decode facial expressions
accurately, especially sadness and happiness, resulting in a negative emotional bias both in the
perception of self and others (Bouhuys, Geerts, & Gordijn, 1999; Bradley et al, 1997; Hale,
1998; Rubinow & Post, 1992).
Studies of patients with brain injury have established the role of the orbital frontal lobes
in emotion regulation (Edwards-Lee & Saul, 1999). Lesions on the right lead to more mania
and lesions on the left to more depression. It has been suggested that relative right frontal lobe
activation reflects innate biological differences related to a vulnerability to experience certain
negative emotions in stressful situations, such as a lower threshold for experiencing sadness
(Dawson, 1994). Transcraniai magnetic stimulation studies eliciting transient hypofunctioning
showed that hypofunctioning in the left but not right prefrontal area resulted in decreased selfreported happiness and a significant increase in sadness ratings (Edwards-Lee & Saul, 1999).

17

Such findings support the paramount importance o f varied aspects of the brain with regard to
emotion.
Hemispheric Asymmetry of Facial Emotion
The right brain hemisphere has frequently been attributed as having a dominant or
special role in the perception and expression o f facial emotion (Asthana & Mandai, 1997;
Borod, Koff, Lorch, & Nicholas, 1986; Damasio, 1994). Studies o f che putative predominance
of the right hemisphere in FEE recognition tests have resulted in mixed findings (Young, 1995).
Some research has shown opposite findings wherein the left hemisphere produced results
typically anticipated from the right (Young, Newcombe, de Haan, Small, & Hay, 1993).
Furthermore, processing of facial emotions may not necessarily be the exclusive function o f the
right hemisphere (Mandai, Asthana, & Maitra, 1998).
The investigation of hemispheric asymmetry in emotion has led to the proposal o f
several models that explain how the brain mediates emotions in humans (Lane, Reiman, Ahem,
Schwartz, & Davidson. 1997). One model suggests that the right hemisphere mediates all basic
emotions. Another proposes that emotions are organized differently accordingly to valence,
with positive emotions mediated by the left hemisphere and negative emotions by the right. A
third posits that emotions are lateralized in regard to the associated motor responses, for
example approach emotions (e.g., happy) are lateralized to the left hemisphere and withdrawal
emotions (e.g., fear) to the right.
However, a recent study showed that FEE information for most basic emotions is
processed more accurately when both the componential (relatively more left hemisphere) and
configurai (relatively more right hemisphere) processes function in parallel (Mandai et al.,
1998). ADHD research suggests that these individuals make more FEE perception errors than
normal because they only attend to the face as a whole (Oades, 1998; Singh et al., 1998). To

18

accurately identify emotion one must pay attention to the individual parts o f the face that
provide differentiating cues. This requires an appropriate level and focus of attention. The
right hemisphere shows some superiority in processing complex FEE that requires greater
visuo-spatial (parietal lobe, dorsal visual stream function) processing in addition to emotional
processing. This suggests that the right hemisphere plays a role in face perception in addition to
its involvement in basic visual abilities.
Neurological Substrates o f Facial Emotion
Since Papez proposed his theory o f the ‘limbic lobe’ as the anatomical basis for emotion
and the discovery that temporal lobe damage produces several behavior changes including flat
affect (Kolb & Whishaw, 1996), notions related to the ‘emotional brain’ have become more
evident in the literature.
One such notion questions whether emotion is mediated by separate neural substrates,
sharing different regions within the same structures, or whether there is a single system for
processing all emotion. Although the current map o f the emotion-CNS relationship remains
crude (Thompson, 1988), there is increasing evidence about selective neuronal responses to
various facial stimuli (Prkachin & Prkachin, 1994, under review; Streit et al., 1999). The
argument is made that different sets of components affect different emotions (Adolphs et al,
1996); however, consistent findings about the specific brain areas involved in emotion remain
to be clearly established (George et al., 1998). To the extent that there are neural substrates of
emotion, one might expect these substrates to include templates for specific emotions, or some
of the so-called basic emotions.
Research supports at least a partially distinguishable emotion substrate mediating system
(Lane, Reiman, & Bradley et al., 1997). Participants viewing FEE in neuroimaging studies
showed differential activation o f the left medial temporal lobe, bilateral occipital-temporal

19

cortex, and cerebellum depending on whether the expression was unpleasant, neutral, or
pleasant. Activation of the caudate nucleus, head of the basal ganglia, occurred when pleasant
emotions were viewed. This did not occur when subjects viewed unpleasant and neutral
emotions. These results suggest that the aforementioned structures are fundamental in the
emotion mediation network with differential involvement dependent on the emotion.
Support for an emotion substrate model comes from consistent findings o f uneven
impairment of FEE perception concerning fear, anger, and especially disgust in studies of
Huntington’s disease (Broks et al., 1998; Gray, Young, Barker, Curtis, & Gibson, 1997;
Sprengelmeyer et al., 1997). Neuroimaging has revealed that Huntington’s disease is associated
with atrophy of the parietal, frontal, temporal and occipital regions, and the caudate nuclei
(Sprengelmeyer et al., 1997). Impairment of disgust perception has been associated with
caudate nuclei damage while impairment o f fear and anger perception has been attributed to
bilateral amygdala damage and atrophy o f the temporal lobe caused by the disease. Findings
related to FEE recognition impairments associated with Huntington’s disease support the
possibility that some of the basic emotions may have dedicated neural substrates (Gray et al.,
1997).
Neuroanatomical correlates have been identified, using PET, not only for disgust, but
also for happiness and sadness (Lane, Reiman, & Ahem et al., 1997). The medial prefrontal
cortex and thalamus seem to be activated in all three emotions regardless o f induction method.
The anterior and posterior temporal cortex appear to participate in all aspects o f these emotions
when they are film-generated. The ventral medial frontal cortex seems to be differentially
activated depending on the valence of the emotion. Furthermore, the anterior insular cortex was
differentially involved in certain aspects o f negative emotion only. In short, prefrontal cortex
and thalamus activation appear to be common to all three emotions, the caudate nucleus and
20

putamen are associated with sadness and disgust, and the anterior insular cortex appears to be
sensitive to the type o f emotion and the nature of the emotional stimulus.
Use of functional magnetic resonance imaging (fMRI) has also highlighted a possible
neural substrate for the perception of happy facial expressions (Phillips et al., 1998). This
technology detects signal increases predominantly in the left anterior cingulate gyrus, bilateral
posterior cingulate gyri, medial frontal cortex and right supramarginal gyrus. These limbic
structures were previously thought to be involved in visuospatial and emotion processing tasks.
However, fMRI studies that include other FEE are required to confirm these findings and clarify
the nature of the neural substrates in perception of distinct facial emotion.
Another example of the emotion substrate hypothesis comes from the study o f fear.
Fear is possibly the most basic of cross-species emotions, evident in almost all animal groups as
a common life experience. As a result, it is probably the best understood and most extensively
and successfully researched emotion in relation to associated neurology (LeDoux, 1994). Much
o f our current understanding o f the potential relationship of neuronal substrates to emotion
relies on studies of fear (Prkachin, in preparation).
What is known about fear may serve as a conceptual model o f how other emotions are
associated with neurological and psychological functions. It is likely that other emotional
expressions involve somewhat different neuronal processes (Prkachin, under review). The most
consistent findings of a neural substrate for fear indicate an association between fear and the
amygdala (LeDoux, 1992), areas o f the superior temporal gyrus, inferior posterior temporal
gyrus, middle temporal gyrus, and the medial frontal cortex (Phillips et al., 1998). Central to
this association is whether the amygdala is involved in emotions other than fear and whether
uni- or bilateral damage is necessary to impair the ability to recognize facial expressions o f fear
(Calder et al., 1996; Gottman, 1993).
21

Similar to impairment in recognizing fear, problems with recognition o f other emotions,
specifically anger and disgust, appear to result from damage to the amygdala, at least in cases
where there has been extensive temporal lobe damage (Broks et al., 1998; Calder et al., 1996)
that is sometimes associated with right parietal cortices (Adolphs et al., 1996). However,
decreased recognition of anger and disgust from facial expressions is usually only evidenced in
comparison to a differential impairment in recognizing fear. This is interesting in that difficulty
identifying the same three emotions has been observed in cases of Huntington’s disease,
implying damage beyond the temporal lobe (Gray et al., 1997; Rolls, 1990; Sprengelmeyer et
al., 1997). This may suggest that FEE perception deficits found in amygdala damage
(differentially impairing fear recognition) and Huntington’s disease (differentially impairing
disgust recognition) share some neurological overlap, or, that cases with extensive temporal
lobe damage and cases of Huntington’s disease interconnect through some other neurological
structure, resulting in common FEE recognition deficits that are differential in severity.
Studies o f the amygdala have shown this structure to have complex neural inputs from
and outputs to numerous cortical and subcortical areas, such as the thalamus and STS
(Aggleton, 1993; LeDoux, 1992,1993b; Rolls, 1995; Young, Hellawell, van de Wal, &
Johnson, 1996). In light of such interconnections, these structures are likely key in mediating
fear, anger and disgust and are co-associated in the recognition impairment o f these emotions.
However, research still needs to establish if any discrete cells within these areas are linked to
selective FEE perception impairments (Pricachin, in preparation).
Although the amygdala has been portrayed as a key component (Rolls, 1995), it would
be inaccurate to consider it the centerpiece’ o f the brain’s emotional system, especially
concerning emotions other than fear (LeDoux, 1995; Thompson, 1988). The amygdala’s
mediation capacity is limited by its crude structure and function (LeDoux, 1993b) and it is
22

essentially one part in a larger neurological organization (Thompson, 1988). Furthermore, the
literature includes cases where the amygdala was not damaged and fear recognition impairment
was caused by bilateral neocortical damage (Broks et al., 1998) as well as reports of bilateral
amygdala damage where fear recognition impairment was not conspicuous in the testing
(Hamann et al., 1996). These mixed findings indicate that bilateral amygdala damage does not
inevitably or invariably interfere with the recognition o f FEE. They also imply that amygdala
damage may not be a required or sufficient condition to impair facial recognition o f fear or
other emotions. This suggests that emotion is not necessarily mediated by a general system but
rather by different substrates for different emotions or clusters o f emotions (Thompson, 1988).
It also implies that emotions are likely distributed throughout reciprocally acting subcortical and
cortical regions (Dawson, 1994; Streit et al., 1999). However, the amygdala seems to be more
specific to stimulus-reward association formations related to conditioned fear, at least in
animals (LeDoux, 1993b).
Neurophvsioloeical Associations with Facial Emotion
Components o f emotion in addition to the particular hypothesized central nervous
system substrate structures have been proposed. Nearly all twentieth-century theories o f human
emotion postulate some type of connection between emotion and physiological activity
(Thompson, 1988). Physiological activation is present, to some degree, in response to internal
(e.g., imagined) experience or external stimuli (e.g., sighting a bear). Therefore, body responses
are associated with emotion in some manner (Strongman, 1996; Thompson, 1988), with FEE
often inducing an emotional state and a response FEE by the observer (Prkachin & Pricachin,
under review).
It has been demonstrated that, for example, voluntarily posed FEE resulted in distinct
stereotyped pattems of involuntary autonomic nervous system (ANS) and

23

electroencephalograph (EEG) activity patterns for some of the universal basic emotions,
specifically anger, fear, disgust, and possibly sadness. The same distinct pattems also were
found across cultures, thus establishing emotion-specific ANS activity associations (Camras et
al., 1993; Ekman, 1992a, 1992b; Thompson, 1988).
Other research found cardiovascular activity differences between the basic emotions in a
directed facial action task (Levenson, 1992). For example, anger, fear and sadness were
associated with greater heart rate acceleration than disgust, and anger produced greater finger
temperature increases than produced by fear. Different heart rates result between positive and
negative emotions in that greater acceleration occurred in relation to fear and anger than
happiness. Vascular differences were also observed. For example, fear showed lower diastolic
blood measures, cooler skin surface temperatures, and greater vascular constriction.
Interestingly, the closer the approximation to the prototypical FEE, the more pronounced the
vascular differences. Regardless of whether these changes fiilly discriminate the emotion,
autonomic activity can lead to different emotional percepts, primarily dependent on the
elaboration of cognitive evaluations (Cacioppo et al., 1993).
It has been purported that body states are an important emotion component. The relative
combination o f ANS, viscera, motor system, and biochemical arousal and activation places the
body in a state commonly associated with a triggering event (Damasio, 1994). As FEE
recognition likely leads to retrieval of relevant information firom the diverse neural system that
serves emotion, past body states have been identified as an important information component
(Adolphs et al., 1996). These states likely rely on associated autonomic and somatovisceral
activity information. Relative inability to access this information would likely diminish or
impair FEE recognition performance.

24

Another interesting postulation has been that ‘feelings’ are what one experiences, the
existential nexus, in the course o f forming a body state to a stimulus (Damasio, 1994). That is,
feeling the emotion as described in the aforementioned somatovisceral activity. A feeling can
be a motivator in guiding behavior as well as providing a somatovisceral frame o f reference
(e.g., “somatic marker.” Damasio, 1994, p. 173) for future behavior (Levenson, 1992). This
occurs regardless of whether feelings are considered a discrete autonomic-somatovisceral
profile o f the emotion or an undifferentiated arousal experience that serves an attentional
purpose for the broader cognitive system.
Alexithvmia and the Perception of Facial Expressions and Emotional Regulation
Alexithymia presents an interesting neuropsychological phenomenon that adds to the
understanding o f emotion. Persons with alexithymia demonstrate a deficit in their ability to
verbally describe or label emotion. They also appear to have difficulty experiencing emotion,
or at least they appear as emotionally blunted or flat (Haviland & Reise, 1996; Lane, Ahem,
Schwartz, & Kaszniak, 1997; Lane et al., 1996; Roedema & Simons, 1999). In alexithymia,
there seems to be an apparent disconnectivity between the labeled emotion, the somatovisceral
and autonomic components o f the emotion, and the triggering event; that is, those with this
disorder lack a cohesive, conscious emotional experience or awareness o f emotional activity
(Lane, Ahem, & Schwartz et al., 1997; Lane et al., 1996). Thus the impact o f this deficit
reaches beyond vocabulary impairment.
It has been suggested that the anterior cingulate cortex contributes to the orchestration
and modulation o f motor, neuroendocrine, and autonomic activities related to emotion through
its feedforward and feedback connections. The anterior cingulate cortex has complex
interactive neural cormections with the somatosensory cortex, basal ganglia, temporal lobe, and
subcortical structures involved in bodystate activities (e.g., with the amygdala, hippocampus,

25

thalamus, hypothalamus). The ‘attentional spotlight’ for these stimuli (e.g., somato visceral and
autonomic) could be a function of the anterior cingulate cortex, helping to select emotion
relevant body state information, and providing a frame o f reference to enhance transference of
this information to explicit conscious awareness processes (e.g., frontal lobe) (Lane, Ahem, &
Schwartz et al., 1997).
The degree o f functional impairment of the anterior cingulate cortex may be related to a
deficit in one’s capacity for interoceptive perception o f emotion relevant information. This
could lead to the breakdown in connecting one’s body sensation to an emotional state or feeling
emotions to a lesser degree. In alexithymia, the proposed attentional function o f the anterior
cingulate cortex may fail to contribute to the conjunction o f labeled emotion and the body state
of arousal. Beyond a perceptual problem, alexithymia may be an affective impairment o f the
degree to which one processes (i.e., experiences) emotion, thus resulting in a deficient cognitive
appreciation of FEE.
Research suggests that physiological activity associated with basic emotions provides at
least an immediate and spontaneous awareness, and often an indelible and enhanced intuitive
sensory memory about the relationship between oneself and some stimulus. This is similar to
what has been proposed as ‘somatic markers’ (Damasio, 1998). These are images that result
from somatosensory patterns and mark the related stimulus situation as good or bad, thus
metaphorically creating a decision-making space that is constrained, manageable and cost
effective.
The labeling o f emotion deficit experienced by persons with alexithymia presents a
different kind o f challenge in FEE recognition. Although these individuals exhibit difficulty in
labeling emotion (nouns) they do not demonstrate a deficit in using nonemotion nouns. Why?
The argument that right hemisphere dysregulation is associated with alexithymia appears to be

26

widely accepted in the literature (Jessimer & Maricham, 1997). Moreover, there is likely a
dysfunction of the complex neural network (crossroads) between the anterior cingulate cortex,
temporal (STS), parietal (PG) and occipital lobes, (initiating the dorsal and ventral visual
streams) that subserves the perception of FEE.
Given that the visual system is functioning normally, from retina through tectum and
thalamus, the problem of labeling emotions must originate ‘downstream.’ Area PG (spatial) of
parietal lobe (located by the supramarginal gyrus) responds to input from visual (occipital lobe)
and somatosensory areas and the anterior cingulate cortex. Spatial capacity is likely involved
with FEE perception through the feature spatial configuration and relationship. Furthermore,
cells in the temporal lobe are sensitive for FEE recognition. The STS (in the posterior temporal
cortex) is a key site for polymodal input and intermodal combining from verbal, auditory, visual
and somatosensory systems. This likely underlies stimulus categorization and assignment of
phonemic tags to emotion categories.
The above activity, largely implicit, is likely part of the discrimination of emotions.
However, its effectiveness is greatly enhanced by language, which transforms knowledge from
implicit to explicit, thus modifying the allocation o f attentional resources in the anterior
cingulate cortex. Words denoting emotion create cognitive schemata that further determine
how emotion relevant information is processed. Failure to provide effective neuronal
differentiation for competing stimuli (emotion relevant visual and language information)
required for further ‘conscious’ processing (e.g., prefrontal cortex feedforward and feedback
processes) may diminish one’s capacity to describe FEE. This also confounds the conscious
experience of what the FEE represents.
Individuals with right frontal lobe damage exhibit impaired verbal expression o f
emotion (Edwards-Lee & Saul, 1999). They use less appropriate emotional words and words

27

with lower emotional intensity when describing emotional situations. Furthermore, emotional
words were processed more accurately when presented in the left visual field, supporting the
relationship between alexithymia and neurologically based emotional impairments.
Those with alexithymia also present a picture o f emotional dysregulation. They have
difficulty in accurately perceiving FEE (Duchesneau, 1996; Jessimer & Markham, 1997), and
demonstrate an impaired capacity to empathize with others (Parker et al., 1993). These
individuals experience difficulty with affect verbalization as well as restricted physiological
arousal and sensations (Roedema & Simons, 1999). There is also an indication that they
experience violent outbursts (Kroner & Forth, 1995). Such dysregulation may result in failure
to consistently recognize specific emotions (Brown, 1993).

It also appears that alexithymia

traits are similar to those o f anorexia nervosa, bulimia, substance abuse, and post traumatic
stress disorder (Salminen, Saarijarvi, & Aarela, 1995). As such, alexithymia may be a
secondary phenomenon resulting from significant psychological trauma during critical
developmental periods in childhood or major catastrophes in adult life (Duchesneau, 1996;
Salminen et al., 1995).
Research has also shown that the construct of (high) alexithymia is related to the
emotionally based aggression of violent offenders and dimensions of psychopathology (Kroner
& Forth, 1995). There seems to be a relationship between the inability to monitor one’s level of
emotional excitement during interactions and physical aggression, causing these individuals to
fail to withdraw firom conflict in social situations. This suggests that those who are unaware of
expressed hostility will maintain hostile feelings whereas those experiencing an adequate level
of awareness will modulate hostile feelings and reduce aggressive outbursts.

28

Literature Review of Neuropsychological Deficits Associated
with Incarcerated Youth
Serious conduct problems are the most common reason for child referral to inpatient and
outpatient psychiatric treatment facilitates (Eme & Kavanaugh, 1995), accounting for up to onehalf of all child and adolescent clinic referrals (Webster-Stratton, 1993). Children with serious
conduct problems are extremely disruptive. They exhibit high rates of antisocial behavior such
as noncompliance and defiance, aggression and cruelty towards people and animals.
Furthermore, they engage in destructive acts, lying, stealing, running away and cheating, that are
clinically significant and clearly beyond the realm of ‘normal’ functioning. These behaviors
bring many such individuals into contact with the criminal justice system (Kazdin, 1997).
Moreover, these behaviors are not mitigated by maturity but are highly predictive of significant
behavior problems in adolescence and adulthood (Phelps & McClintock, 1994), remaining
stable over time and across situations (Moffitt, 1993b). This behavioral profile is common in
incarcerated youth. Hence, they present a significant general concern and a potentially
interesting neuropsychology. Unfortunately, very limited attention has been focused on mental
health problems experienced by incarcerated youth (Ulzen & Hamilton, 1998).
The prevalence of the extreme form o f problem behavior, conduct disorder (CD), in
children aged 4 to 18 years has recently been estimated to range from 4% to 16% in boys and
2% to 9% in girls. It has been suggested that many of these youth are responsible for
approximately 50% to 60% of known crimes (Lynam, 1996; Moffitt, 1993b). It also estimated
that up to 87% o f incarcerated youth (aged 11 to 17) meet the diagnosis criteria for CD
(Eppright, Kashani, Robinson, & Reid, 1993). Although the professional nomenclature may
change, the faces remain the same as these youth drift through successive systems aimed at
addressing their deviance. As such, there is growing interest about neuropsychological factors

29

that appear to put children at risk for serious problem, delinquent behavior (Eme & Kavanaugh,
1995; Lyons et al., 1995; Moffitt, 1997; Webster-Stratton, 1993).
The literature shows that up to 85% o f convicted felons are eligible for antisocial
personality disorder (APD) diagnosis (Lynam, 1996). Among professionals it is generally
accepted that 40% to 50% of children with CD become recidivist criminals and/or adults with
APD (Werry, 1997). However, recent argument has been made that this percentage is too
conservative given that all adults with APD have a history o f CD. Moreover, further study of
children with CD is likely to confirm that a greater percentage than previously estimated will
develop APD in adulthood (Werry, 1997). In addition, when all diagnoses are considered in
conjunction with CD, 84% of the full sample o f subjects received a psychiatric diagnosis as
adults and demonstrated continued dyshmction, criminal behavior and social maladjustment
(Kazdin, 1997).
Associated Neuroosvchological Deficits
One of the most robust findings related to antisocial behavior is that children who
persistently exhibit these behaviors suffer from neuropsychological deficits, especially in
relation to executive control, verbal, visual spatial functioning and other learning disabilities
(Eme & Kavanaugh, 1995; Moffitt, 1993a; Moffitt, 1997). These deficits have statistical
variance that is independent of social class, race, test motivation, and academic attainment.
Furthermore, these deficits are clearly related to underlying neuropsychological functioning, by
definition. Specifically, these impairments have been hypothesized as etiological factors related
to developmental problems, such as ADHD and learning difficulties, that may put children at
risk o f antisocial and delinquent behavior.
At the basic level, biological processes are involved in the way children and adolescents
learn, remember, think, make choices and so on. Biological factors that threaten emotional,

30

behavioral and cognitive functions, reduce overall competence or exacerbate behaviour
problems are the most significant in creating a risk o f delinquency (Conger & Simons, 1997).
In fact, it has been estimated that children with central nervous system (CNS) damage have five
times the incidence of CD (Mrazek & Haggerty, 1994).
Attention Deficit Hvneractivitv Disorder
Attention Deficit Hyperactivity Disorder (ADHD) is a very significant predictor and
comorbid condition of serious conduct problems, including criminal behavior (Barkley, 1990;
Eme & Kavanaugh, 1995; Fomess, Kavale, King, & Kasari, 1994; Hinshaw, 1994; Hinshaw,
Lahey, & Hart, 1993; Horacek, 1998; Kazdin, 1997; Loeber & Keenan, 1994; Mandel, 1997). It
is noteworthy that a large percentage of children (up to 75%) diagnosed with significant conduct
problems typically met the diagnostic criteria for ADHD (Fomess et al., 1994; Kazdin, 1997;
Zubieta & Alessi, 1993). Moreover, reports also show that up to 75% of children diagnosed
with ADHD have been diagnosed with serious conduct problems (Webster-Stratton, 1993).
Exclusion o f those with ADHD from the research would significantly impact what is known
about the neuropsychological features o f children who exhibit serious behavior problems
(Hooper & Tramontana, 1997). Children with comorbid ADHD and conduct problems exhibit,
for example, more physical aggression, more persistent and varied antisocial activity, increased
underachievement and learning problems. Furthermore, they have higher rejection rates from
peers, and increased police contacts, criminal offences and incarcerations (Lynam, 1996)
relative to children who exhibit comorbidity of conduct problems with other symptom clusters
(Hinshaw et al., 1993).
These challenging and problematic behavior pattems are prominent and persistent in
incarcerated youth. In addition, it appears that affect discrimination may be dysfunctional in
ADHD children and adolescents, evidenced in deficits related to the perception, interpretation

31

and labeling o f FEE (Allison, 1997; Corbett, 1998; Harris, 1995a; Ingram, 1996; Singh et al.,
1998). It is conceivable that such deficits could exacerbate the risk o f conduct problems by
maintaining and intensifying a dysfunctional social awareness that contributes to subsequent
externalized misbehavior. Furthermore, dysfunctional social awareness appears to contribute to
disinhibited behavior pattems in youth, increasing the frequency and severity o f their delinquent
behavior, and strengthening the likelihood that they will maintain a pattern o f conduct problems
(Eme & Kavanaugh, 1995).
A hypofrontality associated with ADHD has been confirmed, suggesting reduced global
glucose metabolism (gCBF) and regional glucose metabolism (rCBF) (Harris, 1995a; Horacek,
1998). One of the most significant reductions is in the superior prefrontal cortical areas,
important in attentional processes. In ADHD adolescents, rCBF is decreased bilaterally in the
striatal region as well as in the frontal, temporal, and thalamic areas (Harris, 1995a).
Furthermore, hypoactivity is evidenced in the right medial fi-ontal cortex, right inferior
prefrontal cortex, left caudate nucleus and cingulate area (Rubia et al., 1999). Other research
revealed that the genu of the corpus collosum has been observed (via MRI) to be smaller in
persons with ADHD, potentially limiting relaying and integration of information between both
the left and right frontal cortical areas (Horacek, 1998). ADHD is also associated with an
asymmetry in the cortico*striatal network and atrophy in the right hemisphere (Oades, 1998;
Papa, Berger, Sagvolden, Sergeant & Sadile, 1994). Animal research supports human ADHD
studies, implicating the abnormal functioning o f the basal ganglia and their connections to the
orbito-frontal and limbic structures in attention deficits (Papa et al., 1994). It is within reason to
suspect that brain physiological and functional anomalies experienced by persons with ADHD
could also affect the perception o f FEE, as implied earlier.

32

Lower than normal blood flow in the basal ganglia has been observed in children and
adolescents with ADHD; this may be associated with some o f the brain function anomalies of
the disorder (Copeland, 1991; Rubia et al., 1999). Some functions o f the basal ganglia (e.g.,
integration, coordination, and transmission o f stimuli) are not only implicated as involved in the
ADHD phenomenon (Copeland, 1991) but are also a neurological piece in the emotion system
o f FEE impairments (Kolb & Whishaw, 1996; Lane, Ahem, & Schwartz et al., 1997).
The degree of developmental functioning (e.g., developmental lag) o f the reticular
activation system (RAS) may also contribute to some aspects o f ADHD (Copeland, 1991). This
subcortical structure is best understood for its arousal or modification o f brain functioning. It
has a diffuse neural system that extends to the thalamus and gives rise to important ascending
and descending systems (Kolb & Whishaw, 1996). The reticular system is not fully connected
to the limbic system until adolescence. This connection completes later for persons with
ADHD than it does for their peers (Copeland, 1991) and potentially contributes to a
hypofunctioning affective system. It may also result in a particular reactivity pattern that limits
or biases the flow of information and affects the perception of FEE (Coren & Ward, 1989).
Persons with ADHD experience increased difficulty in scanning and selecting relevant
aspects of stimuli (Oie & Rund, 1999). They notice fewer targets, make more perceptual errors,
and have reduced reaction times, factors that have been associated with at least fronto-striatal
dysfunction (Oades, 1998). Therefore, at minimum, persons with ADHD exhibit
developmental anomalies in the fironto-striatal neural networks associated with varying degrees
of attending problems. One o f the problems has been the correct perception o f FEE (Singh et
al., 1998).

33

Learning and Language Deficits
Other well researched factors associated with antisocial behavior are skills involving
expressive and receptive language, problem solving, and visual spatial ability. It is commonly
understood that the ability to communicate effectively facilitates social behavior. Research has
shown that up to 75% o f children with language deficits (Copeland, 1991) and up to 80% o f
children with learning disorders (Sprouse, Hall, Webster, & Bolen, 1998) referred for clinical
services also had ADHD. While little is known about how deficits and disorders are
intertwined, it is clearly indicated that a significant number of children struggle with
overlapping language and learning deficits and behavior disorders (Copeland, 1991 ; Donahue,
Cole, & Hartas, 1994). The data suggests that the linkages between these problems are more
specifically associated with attention deficits (Hinshaw, 1994). This link between
neuropsychological impairment and antisocial behaviors is a very robust finding. Furthermore,
there is strong evidence that such deficits in early childhood are linked to persistent and extreme
antisocial behavior (Moffitt, 1993a, 1997). For example, lesions in the left firontal lobe and the
connection to the limbic system may be causal in verbal skill deficits commonly exhibited by
children with serious conduct problems (Mandel, 1997).
Research consistently shows that at least 50% o f children with behavioral problems also
exhibit significant language deficits (Donahue et al., 1994). Moreover, it suggests that language
disorders were found in 80% of antisocial boys in residential treatment, likely representative o f
the most severe antisocial cases. Difficulty in social communication with one’s peers, parents,
teachers and others has been noted as one o f the most critical risk factors for conduct problems
that could ultimately develop into persistent antisocial behaviors in adulthood (Moffitt, Lynam,
& Silva, 1994).

34

The co-occurrence of these deficits with conduct problems seems to signify an important
and complex developmental course for childhood onset antisocial behavior, likely having a
fundamental underlying neuropsychological explanation. Cognitive impairments, especially
those related to verbal skills, increase the probability o f children acting impulsively as they
struggle to emotionally regulate stressful or provocative situations (Hooper & Tramontana,
1997). It has been found that delinquents are less accurate than nondelinquents in applying
verbal labels to the emotional states of others when asked to identify expressed emotions
displayed in videotaped vignettes (Savitsky & Czyzewski, 1978). Furthermore, these labeling
problems have been associated with alexithymia in serious adult criminal offenders (Roedema
& Simons, 1999). Persons with alexithymia demonstrate difficulty in accurately perceiving
FEE (Duchesneau, 1996; Jessimer & Markham, 1997; Parker et al., 1993) and impaired
capacity to empathize with others (Parker et al., 1993). They also experience restricted
physiological arousal and are prone to violent outbursts (Kroner & Forth, 1995). While no
research on alexithymia in populations o f incarcerated youth has been published to date, such a
disorder, including difficulties in accurately perceiving FEE, may be a risk factor for
delinquency.
Social skills depend not only on one’s language development and learning capacity but
also involve the ability to discriminate nonverbal social cues. For example, social-emotive
signals for accurate social understanding involve specific pattems o f facial movements thought
to require specialized brain processing mechanisms that reside in the right hemisphere (Harris,
1995a). Children with learning disorders exhibit deficits in decoding nonverbal cues, a
purported relative right hemisphere pathology. It is not surprising that they experience more
relationship and teacher interaction problems, have fewer fiiends, and exhibit other
interpersonal difficulties that persist into adulthood (Sprouse et al., 1998). The ability o f

35

children with nonverbal learning disabilities to accurately perceive the emotions and feelings of
others from facial expressions is diminished compared to those without learning disabilities
(Harris, 1995a; Holder & Kirkpatrick, 1991). Furthermore, reduced attention to evaluating
displayed expressions resulted in greater difficulty in the perception of FEE (Harris, 1995a;
Holder & Kirkpatrick, 1991). Although a smirk and a smile are similar in appearance, they
deliver different information. The ability to distinguish between the two is essential to making
correct inferences and responding appropriately. However, with inadequate social perception,
self-correction is limited, resulting in disturbed and problematic social interactions.
Learning disabilities, language delays, and attention deficits appear to be significantly
related to conduct problems (Webster-Stratton, 1993). The specific association between these
problems is thought to be linked through ADHD (Hinshaw et al., 1993; Hinshaw, 1994). This
key role is supported in the literature and relates to the history o f neurological developmental
problems that increase a child’s vulnerability to the development o f serious conduct problems
(Eme & Kavanaugh, 1995; Mandel, 1997; Moffitt, 1993b). Such findings suggest that
observable behavior is linked to the physical health o f the brain (Moffitt et al., 1994).
The Bioloev of Difficult Temperament: Emotional Dvsreeulation
As previously mentioned, temperament has also been implicated in the etiology of
serious conduct problems, including dysfunctional emotional processes. Biological variations
contribute to difficult temperament and are responsible for the increased prevalence of serious
conduct problems. This ‘child deficit’ hypothesis argues that an abnormal aspect o f the child’s
internal organization at the neurophysiological and neuropsychological level is at least partially
responsible for the development of externalized problem behavior.
Research in temperamental characteristics foretells specific troublesome attributes that
are highly associated with behavior problems. One such attribute has been described as

36

insufficient control (Caspi, Henry, McGee, Moffitt, & Silva, 1995). Children with insufficient
control reflect an inability to modulate impulsive expression, are impersistent in problem
solving, and react with negatively charged emotion. This characteristic is significantly
associated with reports o f antisocial behavior in late childhood.
Recent research supports high impulsivity as the strongest personality predictor of
delinquent behavior (Tremblay, Pihl, Vitaro, & Dobkin, 1994). This presents a clear overlap
with findings previously noted relative to ADHD and CD. More specifically, it suggests that
high impulsivity in kindergarten aged boys was highly predictive of later childhood antisocial
behavior. It also suggests that impulsivity is linked to an abnormally functioning behavior
activation system relating to temperament factors in CD. Furthermore, this research confirmed
the prediction that young persons high in impulsivity, low in anxiety, and low in reward
dependence were at the greatest risk for antisocial behavior.
Impulsivity is not the only link to the greater prevalence o f serious conduct problems.
Autonomic underarousal, perhaps linked to low anxiety, is characteristically exhibited by
children who express extreme antisocial, violent behavior and later psychopathology (ZahnWaxier, Cole, Welsh, & Fox, 1995). Of interest is how such children act towards, feel, think
about, and viscerally experience distressing emotions in others. Low heart rate, decreased skin
conductance response (SCRs) amplitudes and recovery time to base line readings are sensitive
indices of ANS anomalies (i.e., underarousal) evidenced by individuals with CD and APD, and
to a lesser degree by criminals in general (Ellis, 1987; Lynam, 1996; Mrazek & Haggerty, 1994;
Scarpa & Raine, 1997; Zhan-Waxler et al., 1995). Thus, it may be posited that such ANS
anomalies underlie antisocial pattems in youth and adult males. This condition is expressed as
a lack of empathy and concern toward others. The findings revealed that lower heart rate and
lower SCRs were linked with externalizing behaviors, co-occurring with lower empathetic and

37

prosocial behaviors. Furthermore, the most anomalous SCRs have been found in children with
CD-ADHD comorbidity.
Skin conductance responses are a very sensitive index of the orienting response to
stimuli (Scarpa & Raine, 1997). This indicates an allocation o f attentional resources to the
processing of stimuli. As previously noted, the orienting network functions to motorize special
neuronal operations that are required to bring attention to the relevant visual location (selective
visual search) and bind signals into object perception (Jackson et al., 1994: Posner &
DiGirolamo, 1998; Swanson et al., 1998). The brain regions purported to be involved are the
posterior parietal lobe and thalamus. The orienting network acts together with other aspects of
the attentional system to increase relevant signal detection, resulting in the content o f one’s
awareness. As previously noted, such a system has been implicated as important for the
accurate perception of FEE.
It is reasonable to expect that inherited temperamental differences due to various brain
structures and circuits will affect a child’s reactivity to situations. Children who embark on the
conduct problem pathway in their early years, and present as persistent antisocial and violent
individuals, may have inherited a brain physiology that raises the threshold of consciously
experienced anticipatory anxiety, fear and guilt about what violates community standards
(Kagan, 1997). It is thought that most children and adults experience these feelings at an
intensity that moderates their behavior as a result o f adequately functioning somatovisceral
reality checks. However, conduct disordered children possess less brain sensitivity or
responsiveness from the amygdala and ventromedial prefrontal surfaces. As a result, they are
less inhibited than the majority with regard to significant misbehavior. As discussed, FEE
perception likely leads to retrieval o f relevant information, i.e., body states, fi:om the diverse
neural system that serves emotion. Past body states have been identified as an important

38

information component that relies on associated ANS and somatovisceral activity information,
among other factors (Adolphs et al., 1996). Relative inability to access this information or other
reality based information would likely diminish or impair accurate FEE perception.
Central Nervous Svstem Anomalies
Although the literature regarding brain physiology in antisocial children, youth and
adults lacks definitive findings, there are a few broad implications. For example, abnormal
functioning o f the right hemisphere has been implicated in delinquent and criminal conduct
(Mrazek & Haggerty, 1994). The proposed connection is in the bias of negative emotionality,
which is more relatively associated with the functioning o f the right hemisphere. Furthermore,
the right hemisphere is associated in the processing of non-verbal stimuli. Negative
emotionality is characteristic of individuals who tend to be less dependent on language to guide
their behavior.
The frontal lobes have also been implicated in conduct problems. Adolescents
diagnosed with CD manifest impairments that are characteristic of dysfunction found in adults
with frontal lobe damage, such as failure to use feedback to correct responses (Lueger & Gill,
1990), problems in attention and concentration (Mrazek & Haggerty, 1994), and diminished
selective attention. That is, they respond to irrelevant stimuli in the same manner as relevant
stimuli (Pincus, 1999). Furthermore, it is suggested that such problems would likely diminish
the accurate perception of FEE.
Detecting Facial Expressions of Emotion
The limited focus of the literature on FEE recognition and conduct disordered children,
excluding incarcerated youth, has found that children with CD have an impaired ability to
accurately recognize FEE in general, regardless o f CD severity (Greer, 1997). Furthermore,
children who demonstrate these behavioral problems early also demonstrate significant

39

impairment compared to children who develop CD later. Although masked by more apparent
behavior problems, the perception process is definitely abnormal in those with CD (Meloy &
Gacono, 1998). It is expected that this abnormality is also experienced by incarcerated youth.

Purpose of Study and Hypotheses
Given that incarcerated youth appear to represent a very troubled population that
demonstrates significant behavioral, social, cognitive and emotional deficits and sequelae, it
would not be surprising to discover some level of impaired ability to accurately perceive FEE.
As noted, many common problems appear to have, in part, definite neuropsychological
foundations that are linked conceptually with the neuropsychology of emotion and potentially
extend to the perception of facial expressions of emotion. The construct of alexithymia was
included in the study due to its association with FEE perception in the literature (Parker et al.,
1993; Prkachin & Prkachin, 2001, March) and its purported neuropsychology.
Given the significant pressure on community and government resources to develop
meaningful interventions and corrective strategies for behaviorally challenged youth, it would
be useful to determine if social difficulties experienced by this population are related, in part, to
a deficit in perceiving socially important information, especially under demanding temporal
circumstances. The study explored this question in two phases. The first phase investigated
whether there was a difference among incarcerated youth in their adequacy o f accurately
perceiving socially relevant information conveyed by the face, specifically facial expressions of
emotion, under time constraints. It was hypothesized that youth with high alexithymia would
perform less accurately at detecting facial expressions o f emotion than non-alexithymic youth,
under time constraints.
The second phase investigated whether there was a difference between incarcerated
youth and non-incarcerated youth in their ability to accurately perceive socially relevant

40

information, specifically facial expressions o f emotion, under time constraints. Two hypotheses
were explored. The first hypothesis was that incarcerated youth would perform less accurately
at perceiving facial expressions of emotion, under time demands, as compared to non­
incarcerated youth. The second hypothesis was that incarcerated youth are more significantly
affected by alexithymia, contributing to a diminished ability to accurately perceive facial
expressions of emotion, as compared to non-incarcerated youth.

41

CHAPTER TWO
Methods
Phase 1: Incarcerated Sample
Participants
Thirty-four participants, ranging in age from 15 to 18 years, were recruited from youth
confined to a Canadian youth custody centre. The sample consisted o f 23 males (M_ age =
16.56) and 11 females (M age = 16.36) with an overall mean age of 16.5. Typically, the youth
were incarcerated for criminal behaviour that included a variety of summary to indictable
(including capital) offences. Participation in the study was solely on a voluntary basis, with
youth indicating their interest by printing their name on a sign-up poster. The poster introduced
the study as one concerned with facial expressions of emotion and suggested that participants
may find such a study interesting. As no incentive was offered for participating in the study it
was assumed that youth who volunteered did so for personal interest. All spoke English as their
primary language. Although not a factor in this study, those of Aboriginal descent typically
account for between 35% to 55% of the centre’s population, with the remaining population
typically made up o f Caucasians.
Measures
All of the incarcerated participants completed two self-report measures: the Positive and
Negative Affect Schedule (PANAS, Watson, Clark, & Tellegen, 1988, see Appendix A) and the
Toronto Alexithymia Scale-20 (TAS-20, Bagby, Taylor, & Parker, 1994, see Appendix B). The
TAS-20 questionnaire was used to determine if the hypothesized inadequacy to detect FEE was
associated with the proposed neuropsychology of alexithymia. As such, alexithymia would
seem to be one reasonable construct to question in this study. The PANAS questionnaire was
used to assess the potential presence of positive or negative affectivity that could confound the

42

results o f the study. The TAS-20 self-report measure involves rating 20 items on a 5-point
Likert scale for a maximum score o f 100. The 20 items are divided into three sub-scales or
factors: 1) difficulty identifying feelings; 2) difficulty describing feelings; and 3) externally
oriented thinking. Although alexithymia is a dimensional construct, cutoff scores have been
established on the TAS-20 to enable the division of the subject sample into 3 alexithymia (TAS)
groups: high alexithymie (>61), intermediate (52-60), and nonalexithymic (< 51) (Lane et al.,
1996).
The TAS-20 has demonstrated internal consistency (Cronbach’s alpha = .81) and good
test-retest reliability (.77) in both clinical and non-clinical populations and reported high
convergent and concurrent validity and moderate discriminate scale validity (Bagby et al.,
1994).
The PANAS self-report measure uses a 5 point Likert scale to identity participants who
may be high or low in negative affectivity (NA) and positive affectivity (PA). It is comprised of
two 10 item mood scales, one for NA and one for PA, that are useful for identifying possible
anxiety, depression, and general psychological distress (Watson et al., 1988). The PANAS
scales provide reliable, precise and largely independent measures of positive and negative affect
regardless of the subject population or the time frame and response format used. Positive affect
reflects the extent to which a person feels enthusiastic, active, and alert, with high PA scores
characterizing high energy, full concentration, and pleasurable engagement and low PA scores
characterizing sadness and lethargy. Negative affect reflects subjective distress and unpleasant
engagement that subsumes a variety o f aversive moods like anger, contempt, disgust, guilt, fear
and nervousness, with low NA scores characterizing a state of cahnness and serenity. These
scales may also indicate individual differences concerning positive and negative emotional
reactivity. The unique circumstances o f youth custody made it advisable to assess potential

43

situational effects that could impact the participant’s ability to detect FEE. This measure has
been reported to have good internal consistency, high test-retest reliability, and good convergent
and discriminant validity (Watson et al., 1988). However, there is no information relative to its
age group or cross-cultural validity.
Materials
A videotape o f 90 stimulus faces composed from Ekman and Friesen’s (1975) Pictures
of Facial Affect of happiness, sadness, fear, anger, disgust and surprise were used in this study.
The face stimuli (images) were arranged on videotape with video editing equipment (Fast FM
Studio). Each image was captured as a digital graphic image and presented as a single fhune for
about the duration o f an eye blink (33 ms.) with a 2 second interstimulus interval filled by a
black screen. The video was presented on the 32 cm X 42 cm screen o f a video monitor. The
stimulus faces were 23 cm X 18 cm on the monitor screen (in the range o f actual head size).
The video tape consisted of six sets of 90 FEE images (i.e., trials), one set of images for each
target FEE under study (happiness, sadness, fear, anger, disgust and surprise). O f the 90 FEE
images in each set, 15 were the predetermined target FEE, with one target FEE for each set.
The 90 images, or trials, were presented in random order in each FEE set. The 15 images of the
target FEE (happiness, sadness, fear, anger, disgust and surprise) were presented in random
order across the 90 trials, with the restriction that the same facial expression could not be
repeated in succession.
For each of the six target FEE sets in the video tape, a sample display of each target
stimulus (e.g., happy face) was presented first for five seconds, followed by a neutral image (a
statue o f a cat) indicating to the participant that their 90 trial detection procedure was about to
begin. A voice stated the trial image number (sequentially numbered fix)m 1 to 90) immediately
prior to each stimulus presentation.

44

There were four differently ordered versions o f the video tape set presentations, videos
numbered one to four. Each had happiness as the first target FEE set and surprise as the last, as
these emotions are most consistently correctly detected by participants. The remaining four
target FEE sets (i.e., fear, sad, anger and disgust) were presented in a different order between
the happiness and surprise target FEE set presentations. The purpose of this arrangement was to
minimize order effects.
Design and Procedure
This study utilized a mixed design and took approximately 60 minutes for each
participant to complete. The between-subject design involved half o f the participants
completing the TAS-20 and half the PANAS questionnaire prior to viewing the video series
followed by completion of the remaining questionnaire after the viewing. Participants were
randomly allocated to complete either the before or after questionnaire. The within-subject
design involved a comparison of the participants’ ability to detect the various facial expressions.
Participants went through the entire study procedure separately and sequentially to minimize the
potential for interpersonal problem behaviors that are common in custodial institutional
settings. The time of the youths’ participation in the study varied according to the operational
requirements o f the institution—late morning to mid-aftemoon; however, the location within
the institution was the same for each participant. The sample data were collected over a five
month period from late 1999 to early 2000.
All components of the process occurred in the same room and in one sitting, one
participant at a time. Firstly, each participant was required to read along with the tester and sign
a consent form (see Appendix C). This was followed by their completion o f the randomly
assigned questionnaire. Next, each participant was seated in fix)nt o f the video monitor and the
procedure for the remainder of the process was explained.

45

Each participant viewed only one o f the four versions o f the video tape, starting with
tape one for the first participant, followed by each subsequent participant viewing the next
successive video in sequence, always in order fi-om one to four.
This study used the ‘yes only’ form of the detection test (i.e., versus yes-no) in which the
participant’s task was to decide whether the target FEE was present. Prior to each participant
viewing a video, instructions were read aloud indicating that they were to identify a target FEE
(i.e., happiness, sadness, anger, fear, disgust, or surprise) in each numbered face on the video
and respond with ‘yes’ if they detected the emotion or silence if they did not. Furthermore, it
was explained that for each target FEE set a sample display would appear for 5 seconds,
followed by an image of a cat statue indicating that the detection procedure was about to begin.
The experimenter stood to the side and slightly behind the subject’s chair for the purpose of
manually recording their responses. Participants then viewed a video tape with the content as
described in the materials section. Upon completion, subjects completed the remaining
questionnaire.
Debriefing
Youth Custody Centre participants were accompanied individually to the specified
testing room by a youth custody supervisor. Most appeared cautious when entering the room
and avoided eye contact initially. The experimenter was able to quickly establish rapport with
all but one of the participants.
The testing process, consent procedure and steps taken to maintain the anonymity of
participants were explained to each subject. It was also noted that if they felt uncomfortable at
any time they could stop the process and return to their unit. The participant with whom the
experimenter was unable to establish rapport completed one questionnaire and proceeded
through approximately half of the videotape before advising that he was leaving. One other

46

participant ended the session early because he was in pain due to a recent injury incurred while
playing a sport in the gymnasium. The balance of the participants engaged willingly in the
study-albeit some more enthusiastically than others.
All of the participants left the testing room immediately upon completion o f the final
questionnaire and only one asked questions about the potential outcome o f the study.
Phase 2: Incarcerated and Non-incarcerated Sample
Participants
Phase two of this study compared data between the incarcerated sample (fi’om phase 1)
and a community sample, herein referred to as the non-incarcerated sample. The nonincarcerated sample data were thirty-one 17 to 18 year-old volunteers from a group of
introductory psychology students used in a previous study from the University o f Northern
British Columbia (Prkachin & Prkachin, 2001, March). The 31 participants consisted of 14
males (M age = 17.9) and 17 females (M age = 17.9), with an overall mean age of 17.9. As
referenced, these 31 participants were part o f a larger study o f 127 participants, ranging in age
firom 17 to 35 years (M = 22.5), assumed to be a homogenous group.
Measures. Materials. Design and Procedure
The measures, materials, design and procedure used in the Prkachin & Prakchin (2001,
March) study were the same as outlined in phase one o f this present study. The use o f the TAS20 questionnaire was alternated to either before or after viewing the video tape; the PANAS
questionnaire was not used in the Prkachin & Pricachin (2001, March) study. This study used a
research lab at the University of Northern British Columbia to study their participants, one
person at a time.

47

CHAPTER THREE
Results
Phase 1: Analysis for Incarcerated Sample
TAS-20 and PANAS Descriptive Statistics
The total score on the TAS-20 and its three subscales were determined. The TAS-20
scores were converted to categories o f low (total scores less than 52), intermediate (scores from
52 to 60) and high alexithymia (scores greater than 60) based on the criteria provided by Lane et
al. (1996). There were 28.1 % incarcerated participants who scored above 60 on the TAS-20,
21.9% who had scores from 52 to 60, and 50% who had scored below 52. Descriptive statistics
for the TAS groups are shown in Table 3.1.
Table 3.1
Descriptive Statistics for TAS-20 Alexithvmia Groups for
Incarcerated Youth

TAS Group

M

SD

n

Min

Max

High

67.22

4.97

9

61

76

Intermediate

55.71

3.73

7

52

60

Non

45.75

4.60

16

35

51

Note: High alexithymia were scores > 61, intermediate from 52 to
60, and low <51. Non refers to non-alexithymic.

The total score on the PANAS and its two component factors for the incarcerated
sample were determined. The PANAS scores were converted to categories o f low and high NA
( X critical > 23.19) and low and high PA ( x critical < 29.72) (Watson et al, 1988), resulting in
the division of the participants into four sub-groups according to their combined PANAS

48

scores. Group one had low NA and high PA comprising 40.6 % o f the sample, group two had
high NA and high PA comprising 28.1 % of the sample, group three had low NA and low PA
comprising 18.8 % of the sample, and group four had both high NA and low PA comprising
12.5 % o f the sample. It was anticipated that the group having the combination o f critical
means for NA (high) and PA (low), in this case group four, would possibly indicate the greatest
effect on the participants’ sensitivity toward the perception of FEE in the subsequent analysis.
The descriptive statistics for PANAS groups is shown in Table 3.2.
Table 3.2
Descriptive Statistics for PANAS Groups for Incarcerated Youth

Negative Affectivity

Positive Affectivity

Group

n

M

SD

M

SD

One

13

16.54

3.62

35.85

4.10

Two

9

30.00

4.66

35.00

3.04

Three

6

18.83

3.31

26.83

2.40

Four

4

32.00

7.62

25.75

3.95

Note: NA critical mean > 23.19, PA critical mean < 29.72.
Detection Index f A’) Analvsis
An alpha level o f .05 was used for all statistical tests. Perception o f the FEE data was
assessed by a signal detection procedure, the A1 (A-prime) index, a nonparametric model to
determine a good estimate o f forced-choice stimulus detection performance as recommended by
Snodgrass & Corwin (1988). The calculation of their performance was determined first by the

49

frequency of hits and false alarms. A hit (H) was defined by the subject saying the correct
response ‘yes’ to the trial target FEE (e.g., saying yes to fear) in a set o f 90 trials when it was
actually presented; a false alarm (FA) was defined by the participant saying the incorrect
response ‘yes’ when the non-target FEE was presented (e.g., saying ‘yes’ when happiness,
sadness, surprise, disgust, or anger were presented rather than the target fear). The frequency of
hits and false alarms were subsequently converted into probabilities (PH & PEA) which were
used to calculate measures of the participants’ ability to detect the target FEE, A1 = (0.5 +
((PH-PFA) * (1 + PH - PFA)) / ((4* PH) * (1 - PFA)). ^ is an estimate of the area under the
receiver operating characteristic curve defined by a single pair of hit and false alarm
probabilities. On the occasions that false alarms were greater than hits the formula A1 = 0.5 ((FA-H)(1+FA-H)) / 4FA(1-H)) was used (Snodgrass & Corwin, 1988). This procedure was
performed for each o f the six target emotions, comprising six sets o f 90 FEE trial presentations.
For the main study test, the A1 results for each participant were then entered into a repeated
measures analysis o f variance (ANOVA). The sensitivity to the six expressions of emotion was
the repeated measure. The PANAS groups and the TAS-20 subgroups were the betweensubject factors.
Initial examination of the data revealed that two o f the cases had one or more A’
standardized scores in excess of 3.29, indicating potential outliers (Tabachnick & Fidell, 1996).
Consequently, prior to entering the incarcerated sample into further analyses, these two cases
were removed from the data and omitted from the study, reducing the incarcerated sample size
from 34 to 32.
Prior to the main analysis, several preliminary analyses were performed. Because the
selection of age could not be controlled, significant unequal age representation resulted. The
incarcerated sample consisted o f 15 to 18 year-olds, with five 15 year-olds, three 16 year-olds,

50

fourteen 17 year-olds, and six 18 year-olds. In contrast, the comparison sample ages ranged
from 17 to 18 years, with three 17 year-olds and twenty-eight 18 year-olds. As the anticipated
result would likely be confounding effects, thereby distracting from the main interest o f the
study, age was added as a covariate in certain analyses and ignored in others as was appropriate
based on the following tests. First, the results of a two-tailed Pearson correlation analysis
confirmed that age was not significantly related (p > .05) to any of the emotion A1 scores,
indicating that age was not correlated with detection performance. Second, the results o f a
three-way univariate ANOVA testing age (dependent variable) and the order of questionnaire,
order of tape and gender (independent variables) revealed that age was not significantly
different between these factors. Also, research has demonstrated (Masters & Carlson, 1988)
that at about age 10 children are as accurate at detecting basic FEE as adults. Consequently,
age was excluded from further analysis.
Another preliminary test was conducted to analyze whether gender, order of
questionnaire and order of tape had any effect on FEE detection A1 results. It has been
identified in the literature (Guerrero & Reiter, 1998) that subtle differences exist in how males
and females respond to a number of discrete emotions and that females seem to have the
advantage in decoding emotional information. Consequently, gender was included in the
subsequent preliminary analysis. The form of the test was a 2 (gender) x 2 (questionnaire order)
X 4 (tape order) x 6 (emotion) repeated measures ANOVA, with emotion as the repeated

measure. There were no between-subject main effects for gender, questionnaire order or tape
order; furthermore, there were no statistically significant interactions between these variables.
However, there was a within-subject main effect o f emotion, F (3.27,5) = 21.43, p = .000, r\~ =
.56, but no significant within-subject interactions for these variables. The lack o f effect by
questionnaire order, tape order and gender on emotion detection performance has been
51

confirmed elsewhere (Prkachin & Prkachin, 2001, March). Other research (Holder &
Kirkpatrick, 1991; Parker et al., 1993) also confirms that gender has not been a significant
factor in FEE detection tests. Consequently, to simplify matters, these factors were not included
in the subsequent analysis.
The final preliminary analysis was a two-way univariate ANOVA that tested age
(dependant variable), PANAS groups and TAS groups (independent variables). There were no
main effects nor did these variables significantly interact. This indicated that age was not
significantly different between these variables and as a result was excluded from the subsequent
analysis.
The ultimate phase 1 analysis was a 4 (PANAS groups) x 3 (TAS groups) x 6 (emotion)
repeated measures ANOVA, with emotion as the repeated measure. There was no betweensubject main effect for either the PANAS groups or TAS groups, nor was there significant
interaction. There was a significant within-subjects main effect for emotion, GreenhouseGeisser F (5,3.10) = 11.01, g = .000, q“ = .34, but no statistically significant within-subject
interaction between these three variables. Descriptive statistics for PANAS group A1 are shown
in Table 3.3.

52

Table 3.3
Descriptive Statistics for Emotion Detection Sensitivity A’ bv PANAS Group
for Incarcerated Sample

Group 3 n = 6

Group 4 n = 4

Group I n = 13

Group 2 n = 9

Emotion

M

SD

M

SD

M

SD

M

SD

Happy

.94

.03

.93

.04

.94

.02

.94

.03

Surprise

.92

.04

.90

.08

.89

.05

.90

03

Sad

.90

.04

.87

.12

.88

.04

.91

.04

Disgust

.86

.08

.82

.09

.76

.17

.92

.02

Anger

.80

.11

.80

.09

.76

.16

.86

.03

Fear

.74

.05

.69

.09

.66

.07

.71

.07

Note: Group I = low NA and high PA, Group 2 = high NA and high FA, Group 3 = low
NA and low PA, and Group 4 = high NA and low PA.

Descriptive statistics for TAS group A’ performance are shown in Table 3.4.

53

Table 3.4

for Incarcerated Samole

Group I n = 9

Group 2 n = 12

Group 3 n = 10

Emotion

M

SD

M

SD

M

SD

Happy

.94

.02

.93

.03

.94

.03

Surprise

.91

.03

.89

.06

.90

.06

Sad

.89

.11

.89

.03

.89

.06

Disgust

.84

.11

.80

.11

.86

.12

Anger

.82

.08

.79

.08

.79

.14

Fear

.70

.11

.72

.04

.71

.06

Note: Group 1 refers to high alexithymia with scores >61, Group 2 refers
to intermediate alexithymia having scores from 52 to 60, and Group 3
refers to non-alexithymia having scores <51.
To investigate the within-subjects main effect o f emotion, a series o f paired t-tests were
conducted (see Table 3.5 in Appendix D). The results showed that the incarcerated sample
detected happiness better than any other FEE; surprise was second, and was detected better than
anger, disgust, and fear; sadness was third, and was detected better than anger, fear, and disgust
but not significantly different from surprise; disgust was fourth, and was detected better than
fear but not significantly different fix>m anger; anger was fifth, and was detected better than fear
but not significantly different from disgust; fear was last, detected more poorly than any other

54

but not significantly difTerent fi’om surprise; disgust was fourth, and was detected better than
fear but not significantly different fi’om anger; anger was fifth, and was detected better than fear
but not significantly different fi’om disgust; fear was last, detected more poorly than any other
FEE. In short, detection accuracy fi’om best to poorest was happiness, surprise (marginally
better than sadness), disgust (marginally better than anger), with fear last.

Phase 2: Analvsis for Between Samples
TAS-20 Descriptive Statistics
The subsequent analyses compared FEE detection between the incarcerated and the nonincarcerated sample. The total score on the TAS-20 and its three sub-groups were determined
for the non-incarcerated sample, and are presented along with the incarcerated sample in Table
3.6 below. In the non-incarcerated sample, 29% scored above 60,38.7% had scores from 52 to
60, and 32.3% scored below 52.
Table 3.6
Descriptive Statistics for TAS-20 Alexithvmia Groups bv Sample

Incarcerated

Non-incarcerated

TAS Group

M

SD

n

Min

Max

M

SD

n

Min

Max

High

67.22 4.97

9

61

76

66.44

5.90

9

61

79

Intermediate 55.71

3.73

7

52

60

55.25

2.41

12

52

59

Non

4.60

16

35

51

43.20 6.36

10

32

49

45.75

Note: High alexithymia were scores >61, intermediate scores firom 52 to 60, and nonalexithymia <51.

55

Detection Index (A l Analvsis
The signal detection process described in phase 1 was also used in phase 2 for the
preliminary and the main analysis. Because the selection of age could not be controlled,
unequal and dissimilar age representation resulted between the samples. As the anticipated
result would likely be confounding effects, thereby distracting firom the main interest o f this
study, age was added as a covariate in certain analyses and ignored in others as was appropriate
based on the following tests. The first test, a three-way univariate ANOVA, testing age
(dependant variable), questionnaire order, tape order and gender (independent variables) found
no main effects or significant interactions for these factors. This indicated that age was not
significantly different between these factors. However, a two-tailed Pearson correlation
analysis revealed that age was positively and significantly related to the ability to detect
happiness (r = .49, g = .000) and with surprise (r = .41, g = .001). This indicates that the higher
the age the better the detection performance o f happiness and surprise among the 63
participants.
Preliminary tests were conducted to analyze the effects of questionnaire order, tape order
and gender on the detection of emotion fi-om the data o f all 63 subjects. The first preliminary
test involved happiness and surprise covaried with age as these emotions were significantly
correlated with age. The form of the test was a 2 (questionnaire order) x 4 (tape order) x 2
(gender) x 2 (emotion) repeated measures ANCOVA, with emotion as the repeated measure.
The regression for age was significant, F (1,46) = 15.37, p = .000, u ' = .25. The adjusted age
mean was 17.2. After adjustment by the covariate, the between-subject test showed no main
effects for or significant interactions between questionnaire order, tape order and gender;

56

however, there was a within-subject main effect for emotion, Greenhouse-Geisser F (3.15, 5) =
95.33, E = .000,

= .67, but no significant within-subject interactions.

The next preliminary analysis tested for effects o f questionnaire order, tape order and
gender on emotion detection for fear, sadness, disgust and anger without age as a covariate,
because age was not significantly correlated with these emotions. The form o f the test was a 2
(questionnaire order) x 4 (tape order) x 2 (gender) x 4 (emotion) repeated measures ANOVA
with emotion as the repeated measure. There were no between-subject main effects or
significant interactions between the variables; however, there was a within-subject main effect
for emotion, Greenhouse-Geisser F (2.34,3) = 57.99, p = .000, r\~ = .55 but no significant
within-subject interactions. Consequently, to simplify matters, gender, questionnaire order, tape
order, and gender were omitted fi’om further analysis in this study.
The ultimate phase 2 tests were comprised of one repeated measure ANOVA and one
repeated measure ANCOVA. Prior to running these tests, a two-way univariate ANOVA was
conducted with age as the dependant variable and TAS-20 groups and samples as the
independent variable. The results showed that the only significant relationship between these
variables was age and sample, F (1,62) = 35.09, g = .000, q" = .38, confirming that age was
significantly different between the two samples. The subsequent analysis involved happiness
and fear with age as the covariate, as they were significantly correlated to age. The form o f the
test was a 2 (sample) x 3 (TAS group) x 2 (emotion) repeated measures ANCOVA, with
emotion as the repeated measure. The regression for age was significant, F (1,56) = 5.63, g =
.021, q" = .09. The adjusted mean age was 17.2. After adjustment by the covariate, there was
no between-subject main effect for TAS- 20 groups but there was for sample, F (1,56) = 4.63, g
= .036, q^ = .08. See Table 3.7 for descriptive statistics for A1 means for TAS-20 groups.

57

Table 3.7
Descriptive Statistics for FEE Detection Sensitivity A’ for TAS Group bv Sample

Incarcerated

Non-incarcerated

Group 1

Group 2

Group 3

Group 1

Group 2

Group 3

Emotion

M

M

M

M

M

M

Happy

.94“ .00

.94“ .00 .95“ .00

.97“ .00

.95“ .00

.97“ .00

Surprise

.92“ .00

.91“ .00 .91“ .00

.93“ .00

.92* .00

.95“ .00

Sad

.89

.89

.03

.89

.06

.92

.03

.90

.05

.92

.01

Disgust

.84 .105 .80

.11

.86

.12

.93

.06

.90

.07

.92

.05

Anger

.82

.08

.79

.08

.79

.14

.85

.11

.85

.07

.85

.05

Fear

.70

.11

.72

.04

.71

.06

.72

.15

.70

.10

.66

.17

SD

.11

SD

SD

SD

SD

SD

Note: Group 1 refers to high alexithymia with scores >61, Group 2 refers to intermediate
alexithymia with scores from 52 - 60, and Group 3 refers to non-alexithymia with scores
<51. Superscript “ refers to statistical mean after the adjustment for age.
As shown in Table 3.8 and Figure 3.1 below, the incarcerated sample performed poorest
at detecting both happy and surprise.

58

Table 3.8
Descriptive Statistics for the Between Sample Effects of Emotion (A’)

Incarcerated

Non-incarcerated

Emotion

M

SD

MIN

MAX

M

SD

MIN MAX

Happy*

.95“

.00

.87

.98

.97“

.00

.92

1.00

Surprise*

.91“

.00

.72

.98

.93“

.00

.85

.99

Sad

.89

.07

.62

.98

.91

.03

.82

.98

Disgust*

.83

.11

.47

.96

.92

.06

.76

1.00

Anger*

.80

.11

.47

.83

.85

.08

.62

.95

Fear

.71

.07

.48

.84

.69

.14

.24

.91

Note: Means and standard deviation for emotion

Asterisks (*) indicate emotions that

were statistically significantly different. The superscript “ refers to statistical mean after
adjusting for age.

59

Happy Surprise

Sad

Disgust

Anger

Fear

Emotion

Figure 3.1. Mean Detection Sensitivity ^ for samples. Incar refers to incarcerated
sample. Non refers to non-incarcerated sample.

60

The second main test involved the four emotions not significantly correlated with age:
fear, sadness, disgust and anger. The form o f the test was a 2 (sample) x 3 (TAS group) x 4
(emotion) repeated measures ANOVA with emotion as the repeated measure. There was no
main effect for TAS-20 group; however, there was for sample, F (1, 57) = 4.86, p = .032,

=

.08. There was also a within-subject main effect for emotion, Greenhouse-Geisser F (2.64,3) =
66.38, B = .000, i\~ = .24 and a significant interaction for emotion by sample, GreenhouseGeisser F (2.64,3) = 3.43, g = .023, r|^ = .06 (see Figure 3.2 below). An independent samples
post hoc test showed that the incarcerated sample performed significantly poorer at detecting
anger, t (56.03) = -2.03, g = .048 and disgust, t (47.09) = -3.41, g = 001. Descriptive statistics
for A1 performance are shown in Table 3.8.

61

<

i

s
0.7 •

0.6
Fear

Anger Disgust

Sad

Emotion

Figure 3.2. Within-subject emotion by sample interaction effect
for fear, anger, disgust and sad between samples. Incar refers to
incarcerated sample. Non refers to non-incarcerated sample.

62

To compare the linear order o f accuracy of detecting FEE, a series of paired t- tests were
conducted (see Table 3.5 in Appendix D). For the incarcerated sample, their detection accuracy
from best to poorest was happiness, surprise (marginally better than sadness), disgust
(marginally better than anger), and fear last. The non-incarcerated sample performed similarly
except they detected surprise better than sadness, but sadness not significantly different from
disgust. Thus the resulting detection sensitivity for this sample from best to poorest was from
happiness, surprise, disgust (marginally better than sadness), anger, and fear last.
The differences o f the TAS-20 factor scores between the samples were analyzed. The
analysis was a one-way univariate ANOVA, testing sample as the dependent variable and the
TAS-20 factors as the independent variables. The results showed that the samples did not
significantly differ between their factors 1,2 or 3 score means. Means are shown in Table 3.9.
Table 3.9
Descriptive Statistics for TAS-20 Alexithvmia Factors bv Sample

Non-lncarcerated

Incarcerated
TAS Factor

M

SD

Min

Max

M

SD

Min

Max

DIF

17.13

6.48

7

31

16.84

4.69

9

31

DDF

13.91

4.87

6

23

14.94

4.19

8

24

EOT

22.94

4.50

12

31

23.19

5.54

10

32

Note. DIF refers to alexithymia factor 1 or Difficulty Identifying Feelings, DDF refers to
factor 2 or Difficulty Describing Feelings, and EOT refers to factor 3 or Externally
Oriented Thinking.

63

CHAPTER FOUR
Discussion
The underlying purpose o f this study was to explore whether incarcerated youth
experience an inadequacy in the processing of emotion stimuli. Also o f interest was whether
the inadequacy could be conceptually related to neuropsychological factors underlying the FEE
detection process and deficits associated with antisocial behaviour. Given the restrictions of
accessing medical, mental health, behavioural and education information and history o f the
participants, this study used only external sources of information. The data under study,
obtained from subjects completing two short mental health questionnaires (TAS-20 and
PANAS) and viewing videotaped FEE, were used to explore the relationship between the
proposed neuropsychology underlying FEE detection and the proposed neuropsychology of
deficits associated with antisocial youth. Exploration o f this relationship occurred in two
phases.
Phase 1
The first phase of this study investigated whether there were alexithymia-associated
differences among incarcerated youth in their ability to accurately perceive socially relevant
information conveyed by the face, specifically facial expressions of emotion, under time
constraints. It was hypothesized that youth with high alexithymia would perform less accurately
at detecting facial expressions of emotion than non-alexithymic youth, under time constraints.
As the results showed, the incarcerated participants’ TAS-20 sub-groups did not
significantly differ in their detection performance. Individuals with high alexithymia performed
marginally poorer at detecting happiness, disgust and fear and those who did not have
alexithymia did marginally poorer at detecting surprise and anger. The intermediate

64

alexithymie group did not perform better or poorer on any of the expressions compared to the
high and low alexithymie groups.
An essential question was whether environmental stressors, and associated emotional
problems, related to residence in a custodial setting would have a negative effect on detection
performance. Another important consideration was pre-existing emotional problems caused by
disruptive family histories and unstable life-styles common in young offenders. The PANAS
questionnaire was used to examine this issue. As the results showed, existing mood problems
did not significantly affect the incarcerated participants’ detection performance.
However, that is not to say that potential mood problems were not evident. In fact, the
questionnaire revealed that 59 % of the incarcerated sample reported high negative affect, low
positive affect, or a combination of both. In terms of the questionnaire, those who reported high
negative affect were likely experiencing a high degree of subjective distress and unpleasurable
engagement, possibly translated into elevated anger, contempt, disgust, guilt, fear and
nervousness. Furthermore, low reports of positive affect could indicate that the participant was
experiencing sadness and lethargy. Moreover, the combination of these emotional states would
indicate a potentially noteworthy mood problem. For example, research (Bouhuys et al., 1999;
Bradley et al., 1997; Hale, 1998; Rubinow & Post, 1992) has demonstrated that depression can
diminish one’s ability to decode facial expressions accurately, especially sadness and happiness.
As shown in Table 3.3, group 3 of the incarcerated sample, which self-reported as having low
positive affect, showed a trend toward lower A1 means for surprise (.89), disgust (.76), anger
(.76) and fear (.66). However, inasmuch as the PANAS questionnaire reflected the presence of
affect issues, they did not significantly affect the participants’ detection performance.
The results, then, indicate no significant differences associated with alexithymia,
negative affectivity or positive affectivity among the incarcerated participants in their ability to

65

accurately perceive facial emotion as investigated by the PANAS and TAS-20 questionnaires.
Consequently, the discussion turns to phase 2 o f the study, the comparison between the
incarcerated and the non-incarcerated samples.
Phase 2
The second phase of the study investigated whether there was a difference between
incarcerated youth and non-incarcerated youth in their adequacy to accurately perceive facial
expressions of emotion under time constraints. Two hypotheses were explored: that the ability
of incarcerated youth to detect facial expressions of emotion under temporal constraints would
be less than that of non-incarcerated youth; and, that this poorer performance would be
associated with a greater prevalence o f alexithymia among incarcerated youth than nonincarcerated youth.
As the results showed, the incarcerated sample was significantly poorer than the nonincarcerated sample at detecting 4 o f the 6 FEE, specifically happiness, surprise, disgust and
anger. Although detection of sadness was not significantly different between the two samples,
the incarcerated sample did not perform as well at detecting this facial expression, suggesting a
possible trend toward poorer detection. The reasonfs) for the difference between the two
samples raises some interesting questions. As shown in phase 1, if the incarcerated sample
experienced more mood problems during the study, as measured by the PANAS questionnaire,
their detection performance was not significantly affected. Therefore, the incarcerated sample’s
significantly poorer detection of the majority o f FEE is likely related to some other factorfs).
This leads to the discussion of the second hypothesis o f phase 2.
It was anticipated that because alexithymia has been significantly related to poorer
detection performance, and the incarcerated sample would have a higher prevalence of
alexithymia, their poorer performance as a group would be due to that higher prevalence.

66

However, this result was not realized in the study. That is, the incarcerated youth did poorer at
detecting FEE regardless of the severity of their alexithymia. This outcome was a surprise for
two reasons. The first was that the incarcerated youth did not have a higher prevalence of
alexithymia. In fact, although not significantly different, the non-incarcerated sample had
somewhat more persons with alexithymia, when intermediate and high scores were combined
for both samples. The second reason was that although both samples reported a rather large
representation o f high alexithymia (28 % for the incarcerated sample and 29 % for the nonincarcerated sample), there was no significant affect o f alexithymia on detection performance
for either sample. This is interesting in light of other research using a sample o f university
students (Prkachin & Prkachin, 2001; Parker et al., 1993) which demonstrates that alexithymia
subgroups performed significantly differently on the detection of sadness, anger and fear. This
raises the question as to why alexithymia was not a significant factor in this present study.
There are five possible explanations for this outcome. It may be that alexithymia does
not necessarily affect the detection process when the participant is required to acknowledge the
detection o f the target expression by saying ‘yes’ and by remaining silent if the target is not
detected. Although a temporal constraint was used in this study, there was no language demand
such as having to say the label of the expression, for example ‘happy.’ Perhaps the effects of
alexithymia are more often experienced down line in the perception process when one is
required to discriminate between two presented expressions or when labelling demands are
imposed. For example, research using the TAS-20 and Wechsler Adult Intelligence ScaleRevised found that persons with alexithymia showed deficits on the Digit Symbol subtest
(Duchesneau, 1996). Unlike having to detect a familiar expression such as a happy face, the
Digit Symbol presents more perceptual demands by requiring the subject to associate certain
symbols with certain other symbols, thus demanding more concentration. This test is

67

considered to be highly sensitive to brain damage and the most affected by minimal brain
damage of WAIS-R subscales. The results of this test indicate that persons with alexithymia
experience increased difficulty with visual recognition, visual associative learning and attention
to detail. It would be interesting to know if the detection performance scores o f these
individuals in the present study would have been significantly affected if more demands were
incorporated, such as labeling the target expressions or discriminating between expressions.
It could also be that alexithymia, either as a construct or as measured by the current
version o f the TAS-20, is not valid in measuring FEE detection performance in the particular
age range tested, which was from IS to 18 years. A recent study with a large sample of normal
participants found that alexithymia was only related to higher ages (Salimen et al, 1995). If this
is the case, it would suggest that alexithymia may not be an appropriate construct to use with
adolescents in a performance test for FEE detection.
It is possible that the incarcerated sample is a very unique grouping o f youth, with
emotional developments not experienced by a more normative population. This could explain
why the reported presence of alexithymia did not affect their detection performance. There is
support for this notion in the literature (Kroner & Forth, 1995) in that the TAS-20 may not be
sufficiently sensitive to measure the nuances o f emotional capacities o f offenders. If this is the
case, the TAS-20 may not tap into the critical information required to assess this population
sample’s FEE detection performance.
There is also some question as to whether the factors that make up the TAS-20 measure
the same thing in every person. For example, the factor o f difficulty describing feelings may
actually be describing something else, such as a primary measure o f shyness, social anxiety or
shame, or a fear o f disclosing inner feelings (Suslow, 1998). Such things may not necessarily
impact FEE detection test performance. It is interesting to note that many incarcerated youth do

68

not readily share their inner feelings. This may be related to, for example, tragic relationships
with significant family members, high incidence of abuse, poorly developed social skills,
underachieving and poor self-esteem (Moffitt, 1997). It is worth considering that the TAS-20
factor of Difficulty Describing Feelings may not necessarily be testing for emotion word
poverty but rather a learned cautiousness in exposing one’s inner feelings.
Lastly, the non-significant effect for the TAS-20 scores could be due to the rather small
sample size of 63 participants used in this study, one half of which were incarcerated youth and
the other a community comparison. According to Cohen (1992) a two group study would
require at least 64 participants in each group, double that available for this study, to detect a
medium effect size. If that were true with this study, the problem may be insufficient power
through sample size to expose a subtle relationship between alexithymia and detection
performance.
Although not significantly different, it is interesting to note that the incarcerated sample
had marginally higher scores in all three TAS-20 subgroups (see Table 3.6), with the greatest
importance related to the low or non-alexithymic group. The marginally higher scores o f the
incarcerated sample suggest a possible trend of higher alexithymia within that sub-population,
possibly accounting for some of the differences in detection performance between the samples.
Biological Explanations
Inasmuch as neither the PANAS nor the TAS-20 scores were significant determinants in
accounting for the test results between the samples, the proposed neurology underlying FEE and
proposed neurology linked to deficits associated with young offenders offer some conceptual
explanations. Given the complexity o f both areas o f study, only more general information will
be considered in the creation of a conceptual frameworic to examine potential correlates that
may be construed as rising fix>m structural anomalies or functional deficits.

69

The most important interconnection may be the attendonal system. It appears that
attentional processes play a primary role in detecting emotional stimuli, not only because of
emotion’s significant import in social communication, but also because adequately functioning
attention is crucial in situations involving significant time constraints, such as in this study. If
one is to process information, such as FEE, one would need to attend to a stimulus to a
sufficient degree to detect it accurately. It is therefore possible that the process o f attention
could be the key factor for the results of this study.
One important factor that has been associated with problematic attention is ADHD. Its
relevance to this current study is the significant prevalence of ADHD among antisocial youth,
including incarcerated youth, and the fact that this disorder decreases performance in FEE
detection tests. Those with ADHD tend to notice fewer targets and make more perceptual
errors. Interestingly, this disorder has been related to a global hypofrontality with large
reductions of neural activity in different prefrontal cortical areas, right medial frontal cortex,
right inferior prefrontal cortex and areas o f the basal ganglia (i.e., caudate nucleus and cingulate
area) and its connections to the orbito-frontal, temporal and thalamic areas (Harris, 1995a;
Rubia et al., 1999). An attentional system that includes some of the aforementioned structures
has been proposed in the literature (Jackson et al, 1994; Posner & DiGirolamo, 1998; Swanson
et al., 1998). This attentional system is thought to be made up of two subsystems, the anterior
attentional system (AAS) and the posterior attentional system (PAS), and have three networks:
the alerting network (organism readiness) centering on the right frontal lobe and right parietal
lobe; the orienting network (motorizes attention to relevant visual location and object percept),
centering on the posterior parietal lobe and thalamus; and, the executive network (e.g., target
detection: form and category) centering on anterior cingulate, left lateral frontal lobe and basal
ganglia.

70

Research with ADHD patients provides support for its connection to the attentional
system (Swanson et al., 1998). Studies using event related potentials (ERPs) suggest
abnormalities in the right frontal region, implicating the alerting network; and the right parietal
region, implicating the orienting network. Research using imaging technology such as positron
emission tomography suggests abnormalities in the frontal lobe, implicating both the alerting
and executive networks. Studies using anatomical magnetic resonance imaging show smaller
than normal size corpus collosum, basal ganglia, and right frontal lobes, thus implicating all
three networks (orienting, alerting, and executive networks respectively). The ADHD
symptomatology of inattention describes poor sustained attention related to an alerting deficit,
poor selective attention related to an orienting deficit, and poor stimulus detection related to an
executive function deficit. As can be gathered, accurate perception requires an adequately
functioning attentional system to gather relevant task information.
Given that attention must be present to detect a stimulus, it would be reasonable to
expect that the neurology of attention would be connected, whether directly or working in
parallel through networks, with activity related to the neurology involved in FEE detection. For
example, a recent study (Streit et al., 1999) using magnetoencephalography revealed a spatial
neural activity that is crucially involved in the decoding o f basic FEE (happiness, sadness, fear,
surprise, disgust, and anger). This activity was noted in the posterior superior right temporal
cortex, the inferior occipitotemporal cortex (areas o f the early visual processing stream), the
middle sector of temporal cortex (having numerous anatomical connections to the dorsal and
ventral visual stream), the right amygdala, right anterior cingulate, left inferior prefrontal cortex,
and the left inferior frontal structures for all FEE. In short, the process activated the inferior
frontal cortex, amygdala and different parts o f the temporal cortex. As reviewed earlier,
neuroimaging techniques have indicated medial prefrontal cortex, thalamus and cingulate gyrus

71

activity for happiness, sadness and disgust; parts of the basal ganglia active for sadness and
disgust; amygdala, temporal gyrus and medial frontal cortex active for fear and anger; and,
temporal lobe damage associated with anger and disgust FEE.
It is interesting to note that most of the structures associated with the functioning of
ADHD, attention and FEE detection overlap and are connected by various complex neural
networks and loops, contributing to the function at hand. For example, the basal ganglia are
indicated in all three areas and are significant in target detection; furthermore, the basal ganglia,
through their manifold connections to the cortex via the cortico-striate-thalamo-cortico circuit,
contribute to the circuit’s common information processing and act as a comparator o f spatial
representations. The temporal lobe is also involved; for example, through its many parts that
relate to sensory integration, feature selection, object recognition, and perception o f eye and
mouth movement (primary in emotion identification). Furthermore, the temporal lobe is active
in the flow of information within the visual stream with its interconnections to the frontal,
occipital and parietal lobes. Parts of the cingulate are also implicated as contributing toward
target detection by boosting activation and reactivation of extrastriate regions o f selection
expectations in feature detection. The prefrontal cortex is also implicated and is considered as
the end point of sensory integration for the dorsal and ventral visual streams through its many
interconnections to, for example, visual regions of the temporal lobe, superior temporal sulcus
and amygdala, posterior parietal areas, the cingulate and the basal ganglia. With such complex
neuronal interconnections, the prefrontal cortex is a significant contributor in the processing of
social and contextual infonnation and behavior responses. Furthermore, the prefrontal areas
influence the autonomic system which are implicated as important factors in emotional
responses. Taken together, such neural and anatomical connections and activity could be a
significant factor in the incarcerated sample’s poor performance in FEE detection, and may also

72

have implications beyond the immediate situation to the larger behaviour issues associated with
this subpopulation.
The hypoactivity of the reticular activating system (RAS) has also been implicated as
contributing to some aspects of ADHD. As indicated in the literature review, ADHD is
frequently associated with antisocial youth. The RAS is best understood for its arousal or
modification of brain fimctioning, having a diffuse neural system that extends to the thalamus
and hypothalamus and likely into parts of the cerebral cortex, giving rise to ascending and
descending systems (Kalat, 1988; Kolb & Whishaw, 1996). Normally, the RAS is not
completely connected to the limbic system until adolescence, and even later for persons with
ADHD. This hypofimctioning of the RAS may contribute to a more global hypofunctioning
associated with ADHD, especially related to the attentional system and poorer perception of
FEE.
Other deficits associated with antisocial youth include learning problems and emotional
dysregulation or difficult temperament. Difficult temperament may contribute to reduced
detection performance inasmuch as autonomic anomalies are thought to be associated with
problems in the attention orienting processes. With diminished orienting, the binding of
stimulus signals in object perception, critical in the performance o f FEE detection, will likely be
hampered. This deficit has been associated with the attentional system and linked to the
posterior parietal lobe and thalamus.
Children with learning challenges evidence less accuracy in identifying and decoding
non-verbal cues such as social-emotive signals or expressed emotions. Recent research has
shown that children with learning disabilities are less proficient at interpreting emotions from
facial expressions (Holder & Kirkpatrick, 1991). This reduced accuracy in decoding socialemotive signals is thought to be more o f a right hemisphere pathology. This is interesting in
73

light of arguments in the literature that the right hemisphere plays a major role in emotion and
FEE detection and is thought to be associated with delinquent and criminal behaviour. A
mutual problem that might overlap between these areas could be a failure o f the frontal lobe to
successfully integrate sensory information important for attentional processes and behaviour
correction
Taken together, the interrelated neurology and brain anatomy linked with deficits
associated with antisocial youth and underlying FEE processes provide a conceptual framework
that suggests the incarcerated sample’s poorer performance in the detection test may have been
due to a less adequate attentional system, either directly or from interconnections. However,
more research needs to synthesize the literature in both areas to verify this hypothesis.
Theoretical Implications of a Deficit in Detection Performance
The results of this study indicate that incarcerated youth performed significantly less
accurately at detecting facial expressions o f emotion when given a very limited period of time
(the speed of an eye blink) to make the observation. How might this reduced adequacy impact
their behavior? As social creatures, humans rapidly observe numerous facial expressions with
little conscious thought or recollection. These quick observations may give rise to first
impressions that are difficult to dispel. Perhaps these glimpses provide us with information that
serves to weigh or bias our decision to engage in social interaction, and how we will engage. It
could be that the degree of accuracy in detecting facial expressions contributes to and reinforce
one’s style of social engagement and adaptiveness.
The notion o f an emotional fast-track system related to quick detection and reaction to a
threatening stimulus has been proposed in the literature (LeDoux, 1994). The idea stems from
research on fear. From an evolutionary standpoint, such a system would be advantageous or
even essential to avoid situations that present potential threats to safety, well-being and, more

74

importantly, to survival. If such a reaction system exists for fear, other basic emotions may
have their own version of a fast-track type system for different reasons and purposes. For
example, detecting a happy expression may trigger a body state, an internal reaction, which
relays the sense that the immediate experience is potentially good. In this case, rather than
avoiding a potential threat, the person relaxes, ‘feels good’ about the stimulus and is available
to engage positively. At the moment when the happy expression was detected, the attentional
system became conservative by orienting its energy and processes to that stimulus. What
happens, then, if the detection accuracy is poor?
The impact of a poor detection system likely depends on whether it resulted from a
simple failure to detect (i.e., neglect), was caused by a perceptual bias in what was thought to be
the expressed stimulus, or was the outcome o f frequent random error. If detection accuracy has
any weighting on behavior, one could speculate that merely neglecting to detect an expression
would be less serious than erroneously indicating a certain expression’s presence, indicating
low discrimination, or poor accuracy due to frequent random error. In neglecting to detect, one
might be characterized as slow in recognizing cues and engaging in social interaction. In the
case o f a biased response pattern, one might be characterized as being more prone to a certain
style o f interacting, be it positive or negative. Finally, a person demonstrating frequent random
error might be characterized as disorganized, frustrated and less predictable.
Young offenders’ interactions within the confines o f a custody facility are fraught with
frequent outbursts of aggression that at times appear to be without cause. The preceding
discussion and the results of this study suggest that the problem, in part, maybe associated with
a deficit in accurately recognizing facial expressions, which may subtly impact further decoding
and subsequent behavior. Although this study did not explore the differences between the
samples related to bias and error analysis, it may be speculated that incarcerated youth are
75

frequently victims o f aggression in part because they fail to accurately detect important facial
cues. For example, detecting angry expressions too late may interfere with anticipating a
problem in time to divert the situation. Or, a youth with a particular detection bias, say thinking
a disgust expression is directed towards him when in fact the expression was sadness, might
trigger an erroneous somatovisceral state that directs him towards an altercation. Furthermore,
a person who makes frequent errors in detecting facial cues may be more unpredictable and
volatile (i.e., heightened somatovisceral state), their frustration resulting from being
perceptually disorganized leading to aggression with minimal provocation. Although
interesting for discussion, such speculation has not been addressed in the literature, to the
author’s knowledge, and requires further study to establish whether data can support such
hypotheses.
One of the expressions detected less accurately by the incarcerated sample was disgust.
This may have some important implications about the ability of young offenders to self-correct
socially aberrant behaviour. An important function o f disgust expressions is to convey
disapproval of what is occurring or has transpired and to indicate that it is not socially
acceptable. A normal reaction to being the object of an expression of disgust would be a feeling
of anxiety that should ultimately result in some effort to make the appropriate behavioural
changes. Many people are discouraged from engaging in criminal activities by intense feelings
of shame, guilt and anxiety that appear to be effective restraints. It has been suggested (Kagan,
1997) that some offenders continue to engage in criminal behaviour because they lack intense
anticipatory anxiety or guilt related to the consequence o f conunitting crime. A deficit in these
crucial feelings is thought to be linked to reduced responsiveness in the orbitofrontal cortex or
amygdala to sensory information from areas like the stomach, heart, and muscles that
accompany anxiety and guilt. It is interesting to note that these anatomical areas are also

76

associated with the emotion o f disgust. Consequently, the offender may not attend adequately
to disgust expressions and/or to the weakened signals of anxiety, and thus fail to censor their
behaviour.
One of the main concerns related to incarcerated youth is an apparent lack o f empathy
and remorse sufficient to effect change in their behavior. If poor detection o f emotional
expressions is associated with the capacity for empathy, given that to be empathetic one needs
to accurately detect another’s emotional state, the results o f this study may provide crucial
information for intervention strategies. For example, in addition to the cognitive skills model of
therapy offered in custody centres, it would be important to incorporate ways to assist young
offenders to become sensitive to facial expressions of emotion, thereby increasing the
possibility of making the critical connections between behaviour, facial expressions, empathy,
self-monitoring and self-correction.
Alexithvmia and Young Offenders
Although the TAS-20 scores were not significantly related to the detection performance
differences between the samples, the topic of alexithymia as it relates to young offenders
warrants some discussion. To the author’s knowledge, the construct of alexithymia, which may
offer some insight into behavioural issues, has not been directly applied to the problematic
behaviour associated with young offenders. It is interesting to note that alexithymia traits have
been associated with traits of substance abuse and post traumatic stress disorder (Salminen et
al, 1995). The association between delinquency and substance abuse is well established in the
literature (Moffitt, 1993a, 1993b), as is its association with post traumatic stress disorder (Eth,
1990). Therefore, a discussion of alexithymia is warranted regardless o f non-significant study
results.

77

The essential factor in alexithymia is a general impairment in the capacity for encoding
and transforming emotional information; that is to say a deficit in emotional information
processing. This problem is demonstrated by presenting as emotionally blunt and having
difficulty in putting emotions into words, indicating a general lack o f emotional awareness. In
this instance, emotional awareness means the capacity to transform implicit knowledge of
emotion into explicit knowledge through some linguistic system. Part o f the problem may be
that persons with alexithymia consciously experience the visceral concomitants that comprise
‘feeling’ an emotion in a disorganized or confused way, or minimally, or not at all. As a result,
these individuals tend to express themselves through non-verbal behaviour.
O f the 32 young offenders under study, 50 % could be considered to have mild to high
alexithymia. This is particularly concerning given that many young offenders are generally
academic underachievers, demonstrate poor language skill development and have learning
difficulties. Given such concurrent challenges, it would be reasonable to expect that the
probability of young offenders successfully self-correcting, or being motivated to correct
aberrant social behaviour, would be low. Furthermore, it is likely that these conditions would
be highly resistant to intervention. Young offenders who fit this profile would likely
demonstrate little empathy, take little pleasure in social interaction, and experience a high
degree o f strained interpersonal behaviour, rejection and isolation. These individuals may
appear to be unconcerned about such dynamics and could be characterized by others as ‘not
getting it’ or disinterested in what is happening around them.
One study of serious adult offenders (Kroner & Forth, 1995) suggests that the TAS-20
might be useful in assessing the ability o f offenders to identify and communicate emotions.
Such infonnation may assist in exploring the relationship between emotional arousal and
physical aggression. Given that social conflicts are inevitable, it would be advantageous to be

78

able to monitor one’s level of excitement so as to know when to withdraw in order to prevent
hostile engagements. However, persons with alexithymia, at least high those with high degrees
o f alexithymia, tend not to monitor internal emotional states and thus ignore feedback from
emotions in making decisions to avoid or minimize the potential for aggression. In other words,
without an adequate level o f interoceptive awareness, persons with alexithymia will do poorly
at assessing hostile situations and avoiding aggressive interactions. As such interactions are not
uncommon in youth custody facilities, it would not be surprising to discover that those with
alexithymia who participated in this study were also involved in frequent altercations while
incarcerated. If this were true, information about individuals who potentially have alexithymia
would have implications for prevention and intervention strategies aimed at reducing such
behaviour.
If the incarcerated sample studied is at all representative of the percentage o f individuals
with alexithymia present in custody facilitates, it would be important to recognize this special
subpopulation in the development of effective behavioural management strategies and therapy
interventions. For example, because they are less aware of emotional feedback, incarcerated
youth with alexithymia require a behaviour response system that provides immediate cues, such
as direct informative feedback, about what is transpiring in order to avoid or mitigate hostile
interactions. The success of such a strategy would require custodial staff to be knowledgeable
in how to assess and respond to such individuals. By way o f therapy, it would be important to
teach these alexithymie youth emotional awareness and feeling recognition and interpretation
through frequent debriefing, provide assistance in appraising their social interactions and the
subsequent ramifications, and assist them to use emotionally descriptive language to raise
awareness. Although insight-oriented approaches are difficult with persons with alexithymia,
the therapeutic relationship is critical to gradually preparing these individuals to accept
79

interventions (Salminen et al., 1995). However, more research needs to be conducted on
strategies that work best with those who have alexithymia.
General Theoretical Implications of the Studv
This research is important because it studied young offenders relative to both the
constructs of alexithymia and the perceptual detection o f FEE. This study supports other
findings that indicate a similar linear order of FEE detection accuracy for basic emotions
(Prkachin & Prkachin, 2001). The results showed that happy expressions are detected best,
usually followed by surprise and disgust, with a general trend of sadness and anger, followed by
fear as the emotion least accurately detected. Although research has yet to clearly establish why
different emotions are detected better than others, the results o f the present study support the
general notion that each emotional expression is likely to have different qualitative aspects that
result in different performance demands on the perceptual system.
The results o f this study support not only the linear accuracy o f detecting different FEE
but indicate that incarcerated youth perform with less accuracy than more normative
adolescents. This could be significant in addressing the remedial needs of young offenders.
Although this study examined only a small set o f emotional expressions, the apparent deficit
demonstrated by the incarcerated sample may bespeak a potentially more pervasive deficit in
accurately perceiving other socially important information. It stands to reason that if perception
detection accuracy, the essential first step in the more complex stimulus appraisal and response
system, is poor, behaviour could be affected to some degree. This is a very important
consideration given that, generally speaking, laws and the threat of sanctions for criminal
behaviour appear to be relatively ineffective in many cases. The apparent failure to perceive
consequences that is demonstrated by many young offenders creates frustration within society;
however, increasing something that is ineffective (longer sentences and harsher conditions)

80

does not increase its effectiveness. It is crucial that the developmental issue o f possessing a less
adequate perception system and its implications on behavior be recognized in strategies aimed
at addressing the offending behaviors o f incarcerated youth.
Another general theoretical implication of this study is whether the construct o f
alexithymia (and the TAS-20 subgroups and factors) is useful in researching the ability of
incarcerated youth to detect emotional expressions in conjunction with extreme time
constraints. Although alexithymia was indicated through the self-report questionnaire,
alexithymia did not statistically affect the detection performance in a significant manner. To the
author’s knowledge, the validity and applicability o f the alexithymia construct in a normative
adolescent population has not been established in the literature. Therefore, an examination of
the relationship between adolescent development, alexithymia and the TAS-20 measure is
required.
Limitations of Studv
Caution should be exercised in interpreting and generalizing the results o f this study for
several reasons. Firstly, this study measured perception performance related to a restricted
number o f stimuli (six emotional expressions) within a specific temporal constraint (33 ms). In
real life, people deal with enonnous amounts of socially important information within a variety
of constraints. Therefore, further theoretical development is required to link the importance of
accurate detection of stimuli within extremely fast time constraints to the importance o f such
functioning in social interactions.
The next limitation is the matter of the comparison sample. The data from this
particular group came &om first year undergraduate students who may not be representative o f a
normative sample o f adolescents from the general population. Future studies should include a

81

more general youth population, for example a public high school group, as the comparison
sample to improve study generalizability.
The samples also had some age-range dissimilarities and fairly inequitable numbers
within ages. Such disparity made it impossible to include age as a study factor without the
likelihood of confounding effects. To achieve more generalizable results, subsequent studies on
this topic should match both the age range and numbers within each age category between the
samples.
Another, and probably the most important limitation, was the small sample size.
According to Cohen (1992), the sample size should be large enough to detect small differences
between two independent samples and not run the risk of mistakenly rejecting the null
hypothesis. This may have been the case with the expression of sadness and the rejection that
the reported presence of alexithymia did not significantly affect detection performance.
According to Cohen, a two group study, such as this, should have 64 participants in each
sample, half of what was available for analysis in this instance. One way to increase the
incarcerated sample would be to include youth fiom other provincial custody facilities.
Furthermore, this strategy would also provide more participants in all o f the age categories and
improve overall representation of youth in custody.
Also, to support the theoretical behaviour implications presented, data need to be
correlated with current behaviour and mental health records to examine what relationship, if
any, detection performance has to behaviour performance. Further studies on this topic should
include sufficient information to determine if the data support the detection performancebehaviour performance hypothesis presented. If this hypothesis is supported by data, research
can help to develop a more comprehensive theory about the behavioural implications o f having
less adequate perceptual sensitivity in detecting emotional expressions.

82

Lastly, the study results are also limited by the use of two self-report questionnaires, the
PANAS and TAS-20. Because questionnaires are very subjective, there is always the concern
that respondents have not answered truthfully or have misunderstood what they were being
asked. Given that incarcerated youth are often academic underachievers, with significant
deficits in language and learning, the results of the PANAS and TAS-20 may underestimate or
overestimate some dimensions of alexithymia and therefore be o f limited value.
Conclusions
Although not without limitations, this study was important because it included
adolescent offenders in the research concerning the perceptual performance of detecting facial
emotion in conjunction with the construct o f alexithymia. This study confirmed, as
hypothesized, that incarcerated youth would perform less accurately than non-incarcerated
youth in detecting facial expressions o f basic emotions. Specifically, the incarcerated sample
performed significantly more poorly at detecting happiness, surprise, disgust and anger, with a
trend for sadness. Although neither alexithymia nor a generalized mood disorder provided an
explanation for the poorer performance by the incarcerated sample, the neurology associated
with FEE detection and the neurology linked to deficits often associated with antisocial and
criminal behaviour provided some conceptual explanations as well as a focus on an inadequate
attention-perception system. Given this result, it would be important to explore if data can
support the notion that FEE detection performance by incarcerated youth is associated with the
neuropsychological factors as suggested.
Regardless of the basis for the problem, this study provides valuable information as it
highlights a potential deficit in the adequacy o f incarcerated youth to perceive important social
information, which may in part be responsible for some o f the antisocial behaviour
demonstrated by this subpopulation.

83

Future Research
Because the present study was not able to examine age as a factor, it would be important
to investigate if the poorer perceptual performance by the incarcerated sample is associated with
an anomaly in the perception of facial emotion between the ages as compared to a well matched
community sample. Such a study would help us to understand if incarcerated youth have a
perceptual problem that is developmentally different between the age groups as compared to
non-incarcerated youth.
Another interesting study would be to examine if post traumatic stress disorder (PTSD)
is a significant contributor to poorer detection performance. Many antisocial youth come from
adverse home conditions and criminogenic environments (Mofütt, 1993a) and have antisocial
parents, at least one of whom expresses exceptionally violent behaviour (Eme & Kavanaugh,
1995). In such homes, these children have been targets of harsh and abusive treatment. Some
of these youngsters were removed from their families by government officials because of
neglect and abuse. It is reasonable to expect that some young offenders have been traumatized
in these environments. The literature indicates that traumatized youths may act out with
problem behaviours such as truancy, substance abuse, delinquency and aggression (Eth, 1990).
Furthermore, these youths may adopt a rebellious attitude that seems impervious to
intervention. Given the problem behaviour associated with incarcerated youth, it is likely that
some are suffering from PTSD. O f the cluster o f symptoms associated with PTSD, the
problems o f irritability and poor concentration could be associated with poorer detection of
facial emotion.
Although it would be difficult to gain approval for such research, it would be interesting
to include psychophysiologic infonnation, such as skin conductance and heart rate, collected
while the youth are viewing presentations o f facial emotion. Because autonomic dysregulation

84

has been associated with antisocial behaviour, such data could be examined to see if it is
associated with perceptual differences the present study found between incarcerated and nonincarcerated youth. If differences were found, it would provide further support for the
biological hypothesis that inadequate neurological functioning could contribute to poorer
perceptual performance, such as the orienting function o f the attentional system discussed
earlier.

85

References
Adolphs, R., Damasio, H., Tranel, D., & Damasio, A. R. (1996). Cortical systems for
the recognition of emotion in facial expressions. The Journal o f Neuroscience. 16 (23), 76787687.
Aggleton, J. P. (1993). The contribution o f the amygdala to normal and abnormal
emotional states. Trends in Neuroscience. 16 (8), 328-333.
Allison, M. L. (1997). Affective processing and frontal lobe dysfunction in adolescents
with attention deficit hyperactivity disorder. Sciences and Engineering. 57 (9), 5904.
Asthana, H. S., & Mandai, M. K. (1997). Hemiregional variations in facial expression
o f emotions. British Journal of Psvcholoev. 8 8 .519-525.
Bagby, R. M., Taylor, G. J., & Parker, J. D. A. (1994). The twenty-item Toronto
Alexithymia Scale-Q; Convergent, discriminant, and concurrent validity. Journal o f
Psvchosomatic Research. 38 (1), 33-40.
Barkley, R. (1990). Attention deficit disorders; History, definition and diagnosis. In M.
Lewis, & S. Miller (Eds.), Handbook o f developmental psvchooathologv fpo. 65-75). New
York, NY: Plenum Press.
Baenninger, M. (1994). The development of face recognition: Featural or
configurational processing? Journal o f Experimental Child Psvcholoev. 5 7 .377 - 396.
Bomstein, M. H. (1992). Perception across the life span. In M. H. Bomstein, & M. E.
Lamb (Eds.), Developmental psvcholoev. (pp. 155-209). Hillsdale, NJ: Lawrence Erlbaum.
Borod, J. C., Koff, E., Lorch, M. P., & Nicholas, M. (1986). The expression and
perception of facial emotion in brain-damaged patients. Neuroosvchologia, 24 (2). 169-180.
Bouhuys, A. L., Geerts, E., & Gordijn, M. C. M. (1999). Gender-specific mechanisms
associated with outcome of depression: Perception o f emotions, coping and interpersonal
functioning. Psvchiatrv Research. 8 5 .247-261.
Bradley, B. P., Mogg, K., Millar, N., Bonham-Carter, C., Fergusson, E., Jenkins, J., &
Parr, M. (1997). Attentional biases for emotional faces. Cognition and Emotion, 11 (1), 25-42.
Broks, P., Young, A. W., Maratos, E. J., Coffey, P. J., Calder, A. J., Isaac, C. L., Mayes,
A. R., Hodges, J. R., Montaldi, D., Cezayirli, E., Roberts, N., & Hadley, D. (1998). Face
processing impairments after encephalitis: Amygdala damage and recognition o f fear.
Neuropsvchologia. 36 (1), 59-70.
Brown, D. (1993). Affective development, psychopathology, and adaptation. In S.
Ablon, 0 . Brown, E. Khmtzian & J. Mack ^ d s.). Human feelings: Explorations in affect
development and meaning (pp. 5-65). Hillsdale, NJ: Analytic Press.

86

Butterworth, G. (1998). Perceptual and motor development. In A. Demetriou, W.
Doise, & C. F. M. Van Lieshout (Eds.), Life-span developmental psvcholoev (pp. 101-135).
Toronto, ON: John Wiley.
Cacioppo, J. T., Klein, D. J., Bemtson, G. C., & Hatfield, E. (1993). The
psychophysiology of emotion. In M. Lewis & J. M. Haviland (Eds.), Handbook of emotions
(pp. 119-142). New York, NY: Guilford Press.
Calder, A. J., Young, A. W., Rowland, D., Perrett, D. 1., Hodges, J. R., & Etcoff, N. L.
(1996). Facial emotion recognition after bilateral amygdala damage: Differentially severe
impairment of fear. Cognitive Neuroosvcholoev. 13 (5), 699-745.
Camras, L. A., Holland, E. A., & Patterson, M. J. (1993). Facial expression. In M.
Lewis & J. M. Haviland (Eds.), Handbook of emotions (pp. 199-208). New York, NY:
Guilford Press.
Caspi, A., Henry, B., McGee, R. O., Moffitt, T. E., & Silva, P. A. (1995).
Temperamental origins of child and adolescent behavior problems: From age three to age
fifteen. Child Development. 6 6 .55-68.
Cohen, J.( 1992). (Quantitative methods in psychology: A power primer. Psychological
Bulletin. 112111.155-159.
Conger, R. D., & Simons, R. L. (1997). Life-course contingencies in the development
of adolescent antisocial behavior: A matching law approach. In T. P. Thomberry (Ed.),
Advances in criminoloev theorv: vol. 7. Developmental theories o f crime and delinouencv ( d p .
55-99). New Brunswick, NJ: Transaction.
Copeland, E. (1991). Medications for attention disorders and related medical problems:
A comprehensive handbook (pp. 17-68). Atlanta, GA: SPl Press.
Corbett, B. A. (1998). Perception of affect in children with attention deficit
hyperactivity disorder. Sciences and Engineering. 59 (2), 868.
Corbetta, M. (1998). Functional anatomy of visual attention in the human brain: Studies
with positron emission tomography. In R. Parasuraman (Ed.), The attentive brain (pp. 95-122).
Cambridge, MA: MIT Press.
Coren, S., & Ward, L. M. (1989). Sensation and perception (3"* ed ). Toronto, On:
Harcourt Brace Jovanovich.
Dawson, G. (1994). Development o f emotional expression and emotion regulation in
infancy: Contributions o f the fimntal lobe. In G. Dawson & K. W. Fischer (Eds.), Human
behavior and the developing brain (pp. 346-379). New York. NY: Guilford Press.
Derr>’berr>', D., & Rothbart, M. K. (1988). Emotion, attention, and temperament. In C.
E. Izard, J. Kagan & R. B. Zajonc (Eds.), Emotions, cognition, and behavior (pp. 132-166).
New York, NY: Cambridge University Press.
87

Desimone, R. (1991). Face-selective cells in the temporal cortex of monkeys. Journal
of Cognitive Neuroscience. 3 (1). 1-8.
Dimberg, U. (1997). Psychophysiological reactions to facial expressions. In U.
Segerstrale & P. Molnar (Eds ), NonveAal communication: Where nature meets culture (pp. 47
- 60). Mahwah, NJ: Lawrence Erlbaum Associates.
Damasio, A. R. (1994). Descartes’ error. New York, NY: Grosset/Putman Books.
Damasio, A. R. (1998). Emotion and reason in the future of human life. In B. Cartledge
(Ed.), Mind, brain, and the environment (pp. 75-921. New York, NY: Oxford University Press.
Donahue, M., Cole, D., & Hartas, D. (1994). Links between language and emotional/
behavioral disorders. Education and Treatment of Children. 17.244 - 254.
Duchesneau, A. (1996). Alexithymia and visual perception. Perceptual and Motor
Skills. 8 3.291-298.
Edwards-Lee, T. A, & Saul, R. E. (1999). Neuropsychiatry of the right frontal lobe. In
B. L. Miller & J. L. Cummings (Eds.), The human frontal lobes: Functions and disorders (pp.
304-320). New York, NY: Guilford Press.
Ekman, P. (1992a). An argument for basic emotions. Coalition and Emotion. 6 (3/4),
169-200.
Ekman, P. (1992b). Facial expressions o f emotion: New findings, new questions.
Psvcholoeical Science. 3 (1), 34-38.
Ekman, P., & Friesen, W. V. (1975). Pictures of facial affect. Palo Alto, CA:
Consulting Psychologists Press.
Ellis, L. (1987). Neurohormonal bases o f varying tendencies to learn delinquent and
criminal behavior. In E. K. Morris & C. J. Braukmann (Eds.), Behavioral approaches to crime
and delinauencv (pp. 499-518). New York, NY: Plenum Press.
Eme, R. F., & Kavanaugh, L. (1995). Sex differences in conduct disorder. Journal of
Clinical Child Psvcholoev. 24 (4), 406-426.
Eppright, T. D., Kashani, J. H., Robinson, B. D., & Reid, J. C. (1993). Comorbidity o f
conduct disorder and personality disorders in an incarcerated juvenile population. American
Journal of Psvchiatrv. 150 (8), 1233-1236.
Etcoff, N. L. (1984). Selective attention to facial identity and facial emotion.
Neuropsvcholoeia. 22 (3), 281-295.
Etcoff, N. L., & Magee, J. J. (1992). Categorical perception o f facial expressions.
Cognition. 4 4 .227-240.
88

Eth, S. (1990). Post-traumatic stress disorder. In M. Hersen & C. Last (Eds.),
Handbook o f child and adult psychopathology (pp. 263-274). Toronto, ON: Pergamon Press.
Fischer, K. W., Shaver, P. R., & Camochan, P. (1990). How emotions develop and how
they organise development. Cognition and Emotion. 4 (2), 81-127.
Flavell, J., & Miller, P. (1998). Social cognition. In W. Damon, D. Kuhn & R. S.
Siegler (Eds.), Handbook o f child psvchologv (pp. 851-898). New York, NY: John Wiley.
Fomess, S. R., Kavale, K. A., King, B. H., & Kasari, C. (1994). Simple versus complex
conduct disorders: Identification and phenomenology. Behavioral Disorders. 19 (4), 306-312.
George, M. S., Huggins, T., McDermut, W., Parekh, P. 1., Rubinow, D., & Post, R. M.
(1998). Abnormal facial emotion recognition in depression: Serial testing in an ultra-rapidcycling patient. Behavior Modification. 2 2 12). 192-204.
Ginsburg, G., & Harrington, M. (1996). Bodily states and context in situated lines o f
action. In R. Harre & W. Parrott (Eds.), The emotions: Social, cultural and biological
dimensions (pp. 227-258). Thousand Oaks: Sage.
Gottman, J. M. (1993). Studying emotion in social interaction. In M. Lewis & J. M.
Haviland (Eds.), Handbook o f emotions (pp. 475-487). New York, NY: Guilford Press.
Gray, J. M., Young, A. W., Barker, W. A., Curtis, A., & Gibson, D. (1997). Impaired
recognition of disgust in Huntington’s disease gene carriers. Brain. 120.2029-2038.
Greer, A. E. (1997). Emotional development in children with child-onset and
adolescent-onset conduct disorder. Sciences and Engineering. 58 (5), 2675.
Guerrero, L. K., & Reiter, R. L. (1998). Expressing emotion: Sex differences in social
skills and communicative responses to anger, sadness, and jealousy. In D. J. Canary & K.
Dindia (Eds.), Sex differences and similarities in communication (pp. 321-350). Mahwah, NJ:
Lawrence Erlbaum.
Hale, W. W. (1998). Judgment of facial expressions and depression persistence.
Psvchiatrv Research. 8 0 .265-274.
Halgren, E. (1992). Emotional neurophysiology o f the amygdala within the context o f
human cognition. In J. P. Aggleton (Ed.), The Amvgdala: Neurobiological aspects o f emotion,
memorv. and mental dvsfunction (no. 191-228). Toronto, ON: Wiley-Liss.
Hamann, S. B., Stefani, L., Squire, L. R., Adolphs, R., Tranel, D., Damasio, H., &
Damasio, A. (1996). Recognizing facial emotion. Nature. 379 (8), 497.
Harris, J. C. (1995a). Developmental neuropsvchiatrv: Assessment diagnosis, and
treatment o f developmental disorders (pp. 175-403). New Yoik, NY : Oxford University Press.

89

Harris, J. C. (1995b). Developmental neuropsvchiatrv: The fundamentals (pp. 121-218).
New York, NY: Oxford University Press.
Haviland, J. M., & Lelwica, M. (1987). The induced affect response: 10-week-old
infants’ responses to three emotion expressions. Developmental Psvcholoev. 23 (1), 97-104.
Haviland, M. G., & Reise, S. P. (1996). A California Q-set alexithymia prototype and
its relationship to ego-control and ego-resiliency. Journal of Psvchosomatic Research. 41 (6),
597-608.
Hess, U., Blairy, S., & Kleck, R. E. (1997). The intensity o f emotional facial
expressions and decoding accuracy. Journal o f Nonverbal Behavior. 21 (4), 241-257.
Hinshaw, S. P. (1994). Attention deficits and hyperactivity in children. In A. E.
Kazdin (Ed.), Developmental clinical psvcholoev and psvchiatrv (Vol. 29). London: Sage.
Hinshaw, S. P., Lahey, B. B., & Hart, E. L. (1993). Issues o f taxonomy and comorhidity
in the development of conduct disorder. Development and Psvchopatholoev. 5 .31-49.
Hofhnan, E. A., & Haxhy, J. V. (2000). Distinct representations o f eye gaze and
identity in the distributed human neural system for face perception. Nature Neuroscience. 3 (1),
81-84.
Holder, H. B., & Kirkpatrick, S. W. (1991). Interpretation o f emotion from facial
expressions in children with and without learning disabilities. Journal of Learning Disabilities.
24(3), 170-177.
Hooper, S. R., & Tramontana, M. G. (1997). Advances in the neuropsychological bases
of child and adolescent psychopathology: Proposed models, findings, and ongoing issues. In T.
H. Ollendick & R. J. Prinz (Eds.), Advances in clinical child psvcholoev (pp. 133-175). New
York, NY: Plenum Press.
Horacek, H. J. (1998). Brainstorms: Understanding and treating the emotional storms of
attention deficit hvperactivitv disorder from childhood through adulthood. New Jersey: Jason
Aronson.
Ingram, N. F. (1996). The role o f affect decoding and encoding in the social
relationships of children with attention deficit hyperactivity disorder. Humanities and Social
Sciences. 56 (7), 2610.
Izard, C. E. (1993). Four systems for emotion activation: Cognitive and noncognitive
processes. Psvcholoeical Review. 100 ID. 68-90.
Jackson, S. R., Marrocco, R., & Posner, M. I. (1994). Networks o f anatomical areas
controlling visuospatial attention. Neural Networks. 7 (6), 925-944.
Jessimer, M., & Markham, R. (1997). Alexithymia: A right hemisphere dysfunction
specific to recognition of certain facial expressions? Brain and Copnition. 3 4 .246-258.
90

Johnson, M. (1998). The neural basis o f cognitive development. InW. Damon, D.
Kuhn & R. S. Siegler (Eds.), Handbook of child psvcholoev (pp. 851-898). New York, NY:
John Wiley.
Johnson-Laird, P. N., & Oatley, K. (1992). Basic emotions, rationality, and folk theory.
Cognition and Emotion. 6 (3/4), 201-223.
Kagan, J. (1997). Temperament and the reactions to unfamiliarity. Child Development.
^ ( l ) , 139-143.
Kalat, J. W. (1988). Biological psvchologv (3"* ed ). Belmont, CA: Wadsworth.
Kazdin, A. E. (1997). Conduct disorder across the life-span. In S. S. Luther, J. A.
Burack, D. Cicchetti & J. R. Weisz (Eds.), Developmental psvchologv: Perspectives on
adiustment. risk and disorder (pp. 248-272). New York, NY: Cambridge University Press.
Kirita, T., & Endo, M. (1995). Happy face advantage in recognizing facial expressions.
Acta Psvchologica. 8 9 .149-163.
Kolb, B., & Wishaw, 1. (1996). Fundamentals of human neuropsvchologv (4th ed ).
New York, NY: W. H. Freeman.
Kroner, D. G., & Forth, A. E. (1995). The Toronto alexithymia scale with incarcerated
offenders. Personalitv and Individual Differences. 19151.625-634.
Lane, R. D., Ahem, G. L., Schwartz, G. E., & Kaszniak, A. W. (1997). Is alexithymia
the emotional equivalent of blindsight? Biological Psvchiatrv. 4 2 .834-844.
Lane, R. D., Kivley, L. S., Du Bois, M. A., Shamasundara, P., & Schwartz, G. E. (1995).
Levels of emotional awareness and the degree of right hemispheric dominance in the perception
of facial emotion. Neuropsvchologia. 33 (5), 525-538.
Lane, R. D., Reiman, E. M., Ahem, G. L., Schwartz, G. E., & Davidson, R. J. (1997).
Neuroanatomical correlates of happiness, sadness, and disgust. American Joumal of Psvchiatrv,
154(7), 926-933.
Lane, R. D., Reiman, E. M., Bradley, M. M., Lang, P. J., Ahem, G. L., Davidson, R. J.,
& Schwartz, G. E. (1997). Neuroanatomical correlates of pleasant and unpleasant emotion.
Neuropsvchologia. 35 t i l l . 1437-1444.
Lane, R. D., Sechrest, L., Reidel, R., Weldon, V., Kaszniak, A., & Schwartz, G. E.
(1996). Impaired verbal and nonverbal emotion recognition in alexithymia. Psvchosomatic
Medicine. 58.203-210.
LeDoux, J. E. (1992). Emotion and the amygdala. In J. P. Aggleton (Ed ), The
Amvpdala: Neurobiological aspects of emotion, memorv. and mental dvsfunction (pp. 339351). Toronto, ON: Wiley-Liss.
91

LeDoux, J. E. (1993a). Emotional memory systems in the brain. Behavioural Brain
Research. 5 8.69-79.
LeDoux, J. E. (1993b). Emotional networks in the brain. In M. Lewis & J. M.
Haviland (Eds.), Handbook o f emotions (pp. 109-118). New York, NY: Guilford Press.
LeDoux, J. E. (1994). Emotion, memory and the brain. Scientific American. 50-57.
LeDoux, J. E. (1995). Emotion: Clues from the brain. In J. T. Spence, J. M. Darley &
D. J. Foss (Eds.). Annual review o f psvcholoev (pp. 209-235). Palo Alto, CA: Annual Review
Inc.
Levenson, R. W. (1992). Autonomic nervous system differences among emotions.
Psvcholoeical Science. 3 (1), 23-27.
Lewis, M. (1993). The emergence of human emotions. In M. Lewis & J. M. Haviland
(Eds.), Handbook of emotions (pp. 223-235). New York, NY: Guilford Press.
Loeber, R., & Keenan, K. (1994). Interaction between conduct disorder and its
comorbid conditions: Effects of age and gender. Clinical Psvcholoev Review. 14 (6), 497-523.
Lueger, R. J., & Gill, K. J. (1990). Frontal-lobe cognitive dysfunction in conduct
disorder adolescents. Joumal o f Clinical Psvcholoev. 46 (6), 696-706.
Lynam, D. R. (1996). Early identification o f chronic offenders: Who is the fledging
psychopath? Psvcholoeical Bulletin. 120 (2), 209-234.
Lyons, M. J., True, W. R., Eisen, S. A., Goldberg, J., Meyer, J. M., Faraone, S. V.,
Eaves, L. J., & Tsuang, M. T. (1995). Differential heritability in adult and juvenile antisocial
traits. Archives of General Psvchiatrv. 5 2 .906-915.
MacLean, P. D. (1993). Cerebral evolution o f emotion. In M. Lewis & J. M. Haviland
(Eds.), Handbook of emotions (pp. 67-831. New York, NY: Guilford Press.
Masters, J. C., & Carlson, C. R. (1988). Children’s and adults’ understanding o f the
causes and consequences of emotional states. In C. E. Izard, J. Kagan & R. B. Zajonc (Eds.),
Emotion, cognition and behavior (pp. 438-463). New York, NY: Cambridge University Press.
Mandai, M. K., Asthana, H. S., & Maitra, S. (1998). Right hemisphere damage impairs
the ability to process emotional expressions o f unusual faces. Behavior Modification, 22 (2),
167-176.
Mandel, H. P. (1997). Conduct disorder and underachievement. Toronto, ON: Wiley.
McKeh'ie, S. J. (1995). Emotional expression in upside-down faces: Evidence for
configurational and componential processing. British Joumal of Social Psvchologv. 3 4 ,325334.
92

McNaughton, N. (1989), Biology and emotion. New York, NY: Cambridge University
Press.
Meloy, R., & Gacano, C. (1998). Understanding, managing and treating conduct
disorder and antosicial personality.
Moffitt, T. E. (1993a). Adolescence-limited and life-course-persistent antisocial
behavior: A developmental taxonomy. Psychological Review. 100 (4), 674-701.
Moffitt, T. E. (1993b). The neuropsychology of conduct disorder. Development and
Psychopathology. 5 . 135-151.
Moffitt, T. E. (1997). Adolescence-limited and life-course-persistent offending: A
complementary pair of developmental theories. In I . P. Thomberry (Ed.), Advances in
criminology theory: vol. 7. Developmental theories of crime and delinquency (pp. 11-55).
New Brunswick, NJ: Transaction.
Moffitt, T. E., Lynam, D. R., & Silva, P. A. (1994). Neuropsychological tests predicting
persistent male delinquency. Criminology. 32 (2), 277-300.
Motter, B. C. (1998). Neurophysiology o f visual attention. In R. Parasuraman (Ed.),
The attentive brain (pp. 51.70). Cambridge, MA: MIT Press.
Mrazek, P. J., & Haggerty, R. J. (1994). Reducing risk for mental disorders: Frontiers
for preventive intervention research. Washington, DC: National Academy Press.
Oades, R. (1998). Frontal, temporal and lateraiized brain function in children with
attention-deficit hyperactivity disorder: a psychophysiological and neuropsychological
viewpoint on development. Behavioral Brain Research. 9 4 .83-95.
Oatley, K., & Jenkins, J. M. (1992). Human emotions: Function and dysfunction. In
M. Rosenzweig & L. Porter (Eds.), Annual review of psychology (pp.55-85). Palo Alto, CA:
Annual Review.
Oatley, K., & Jenkins, J. M. (1996). Understanding emotions. Toronto, ON: Blackwell.
Oie, M., & Rund, B. R.. (1999). Neuropyschological deficits in adolescent-onset
schizophrenia compared with attention deficit hyperactivity disorder. American Joumal of
Psvchiatrv. 15618). 1216-1222.
Papa, M., Berger, D. F., Sagvolden, T., Sergeant, J. A., & Sadile, A. G. (1994). A
quantitative cytochrome oxidase mapping study, cross-regional and neurobehavioural
correlations in the anterior forebrain of an animal model o f attention deficit hyperactivity
disorder. Behavioral Brain Research. 94. 197-211
Paiker, J. D. A, Taylor, G. J., & Bagby, R. M. (1993). Alexithymia and the recognition
o f facial expressions of emotion. Psychotherapy and Psvchosomatics. 5 9 .197-202.
93

Phelps, L , & McClintock, K. (1994). Papa and peers; A biosocial approach to conduct
disorder. Joumal of Psychopathology and Behavioral Assessment. 16(1153-67.
Phillips, M. L., Young, A. W., Scott, S. K., Calder, A. J., Andrew, C., Giampietro, V.,
Williams, S. C. R., Bullmore, E. T., Brammer, M., & Gray, J. A. (1998). Neural responses to
facial and vocal expressions of fear and disgust. Proceedings of the Roval Society o f London:
Biological Sciences. 2 6 5 .1809-1817.
Pincus, J. H. (1999). Aggression, criminality, and the frontal lobes. In B. L. Miller & J.
L. Cummings (Eds ), The human frontal lobes: Functions and disorders (pp. 547-603). New
York, NY: Guilford Press.
Posner, M. I., & DiGirolamo, G. J. (1998). Executive attention: Conflict, target
detection, and cognitive control. In R. Parasuraman (Ed.), The attentive brain (pp. 401-424).
Cambridge, MA: MIT Press.
Prior, M. (1992). Childhood temperament. Joumal o f Child Psychology and Psvchiatrv.
and Allied Disciplines. 33 (1), 249-279,
Prkachin, G. C. (under review). The detection and identification of facial expressions of
emotion: The effects o f orientation and practice. British Joumal of Psychology.
Prkachin, G. C. (in preparation). The perception o f facial expressions: Social and
emotional attention. Manuscript in preparation. University o f Northem British Columbia at
Prince George.
Prkachin, G. C., & Prkachin, K. M. (March 9,2001). Alexithymia and detection of
facial expressions of emotion. Paper presented at the meeting of the American Psychosomatic
Association, Monterey, CA.
Prkachin, G. C., & Prkachin, K. M. (under review). Detection o f facial expressions of
emotion following adaptation. Canadian Joumal of Psychology.
Prkachin, G. C., & Prkachin, K. M. (1994). Adaptation to facial expressions of emotion.
Cognition and Emotion. 8 (1), 55-64.
Puce, A., Allison, T., Bentin, S., Gore, J. C., & McCarthy, G. (1998). Temporal cortex
activation in humans viewing eye and mouth movements. The Joumal of Neuroscience. 18 (6),
2188-2199.
Roedema, T. M., & Simons, R. F. (1999). Emotion-processing deficit in alexithymia.
Psychophysiology. 3 6 .379-387.
Rolls, E. T. (1990). A theory of emotion and its application to understanding the neural
basis o f emotion. Cognition and Emotion. 4 (31.161-190.

94

Rolls, E. T. (1995). A theory o f emotion and consciousness, and its application to
understanding the neural basis of emotion. In M. Gazzaniga, E. Bizzi, 1. B. Black, C.
Blakemore, L. Cosmides, S. M. Kosslyn, J. E. LeDoux, J. A. Movshon, S. Pinker, M. 1. Posner,
P. Rakic, D. L. Schacter, J. Tooby & E. Tulving (Eds.), The cognitive neurosciences (pp. 10911106). Cambridge, MA: MIT Press.
Rubia, K., Overmeyer, S., Taylor, E., Brammer, M., Williams, S. C. R., Simmons, A., &
Bullmore, E. T. (1999). Hypofrontality in attention deficit hyperactivity disorder during higherorder motor control: A study with functional MRI. American Joumal o f Psvchiatrv. 156 (6),
891-896.
Rubinow, D. R., & Post, R. M. (1992). Impaired recognition of affect in facial
expression in depressed patients. Biological Psvchiatrv. 3 1 .947-953.
Salminen, J. K.., Saarijarvi, S., & Aarela, E. (1995). Two decades o f alexithymia.
Joumal of Psvchosomatic Research. 39 (7), 803-807.
Savitsky, J. C., & Czyzewski, D. (1978). The reaction o f adolescent offenders and
nonoffenders to nonverbal emotion displays. Joumal of Abnormal Child Psvchologv. 6 (1), 8996.
Scarpa, A., & Raine, A. (1997). Psychophysiology o f anger and violent behavior.
Psvchiatric Clinics of North America. 20 (2), 375-394.
Sergent, J., Ohta, S., & MacDonald, B. (1992). Functional neuroanatomy o f face and
object processing. Brain. 115.15-36.
Shaffer, D. R. (1996). Developmental psvchologv: Childhood and adolescence (4th
ed.). Toronto, ON: Brooks/Cole.
Siminov, P. V. (1997). Brain mechanisms of emotions. Neuroscience and Behavioral
Phvsiologv. 27 (4), 405-413.
Singh, S. D., Ellis, C. R., Winton, A. S. W., Singh, N. N., Leung, J. P., & Oswald, D. P.
( 1998). Recognition of facial expressions o f emotion by children with attention-deficit
hyperactivity disorder. Behavior Modification. 2 2 12). 128-142.
Slater, A., & Butterworth, G. (1997). Perception o f social stimuli: Face perception and
imitations. In G. Bremner & A. Slater (Eds.), Infant development: Recent advances (pp. 223245). Hove, England: Psychology Press.
Snodgrass, J. G., & Corwin, J. (1988). Pragmatics of measuring recognition memory:
Applications to dementia and amnesia. Joumal of Experimental Psvchologv: General, 117(1),
34-50.
Sprcngelmeyer, R., Young, A. W., Sprengelmeyer, A., Calder, A. J., Rowland, D.,
Perrett, D., Homberg, V., & Lange, H. (1997). Recognition o f facial expressions: Selective
95

impairment of specific emotions in Huntington’s disease. Cognitive Neuropsychology. 14 (6),
839-879.
Sprouse, C. A., Hall, C. W., Webster, R. E., & Bolen, L. M. (1998). Social perception
in students with learning disabilities and attention-deficit/hyperactivity disorder. Joumal o f
Nonverbal Behavior. 22 (2). 125-134.
Stein, N. L., & Levine, L. J. (1989). The causal organization o f emotional knowledge:
A developmental study. Cognition and Emotion. 3 (4), 343-378.
Strongman, K. T. (1996). The psychology o f emotion: Theories of emotion in
perspective (4th ed.). Toronto, ON: Wiley.
Streit, M., loannides, A. A., Liu, L., Wolwer, W., Dammers, J., Gross, J., Gaebel, W., &
Muller-Gartner, H. W. (1999). Neurophysiological correlates of the recognition o f facial
expressions of emotion as revealed by magnetoencephalography. Cognitive Brain Research. 7.
481-491.
Suslow, T. (1998). Alexithymia and automatic affective processing. European Joumal
of Personalitv. 12.433-443.
Swanson, J., Posner, M. I., Cantwell, D., Wigal, S., Crinella, F., Filipek, P., Emerson, J.,
Tucker, D., & Nalcioglu, O. (1998). Attention-deficit/hyperactivity disorder: Symptom
domains, cognitive processes, and neural networks. In R. Parasuraman (Ed.), The attentive
brain (pp. 445-460). Cambridge, MA: MIT Press.
Tabachnick, B. G., & Fidell, L. S. (1996). Using multivariate statistics (3"* ed.).
Northride, CA: Harper Collins.
Tassinary, L. G., & Cacioppo, J. T. (1992). Unobservable facial actions and emotion.
Psychological Science. 3 Ü). 28-33.
Thompson, J. G. (1988). The psychobiology o f emotions. New York, NY: Plenum
Press.
Tremblay, R. E., Pihl, R. O., Vitaro, F., & Dobkin, P. L. (1994). Predicting early onset
o f male antisocial behavior firom preschool behavior. Archives o f General Psvchiatrv. 5 1 .732739.
Ulzen, T. P. M., & Hamilton, H. (1998). The nature and characteristics o f psychiatric
comorbidity in incarcerated adolescents. Canadian Joumal of Psvchiatrv. 4 3 .57-63.
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation o f brief
measures o f positive and negative affect: The PANAS scales. Joumal o f Personalitv and Social
Psvchologv. 54 (6). 1063-1070.

96

Webster-Stratton, C. (1993). Strategies for helping early school-aged children with
oppositional defiant and conduct disorders: The importance of home-school partnerships.
School Psvchologv Review. 22 (3), 437-457.
Werry, J. S. (1997). Severe conduct disorder-some key issues. Canadian Joumal o f
Psvchiatrv. 4 2 .577-583.
White, M. (1995). Preattentive analysis o f facial expressions of emotion. Cognition and
Emotion. 9 (5), 439-460.
Young, A. (1995). Perceiving social and physical environments. In B. Cartledge
Œd.l. Mind, brain, and the environment (pp. 72-136). New York, NY: Oxford University
Press.
Young, A. W., Hellawell, D. J., van De Wal, C., & Johnson, M. (1996). Facial
expression processing after amygdalotomy. Neuropsvchologia. 34 (1), 31-39.
Young, A. W., Newcombe, P., de Haan, E. H. P., Small, M., & Hay, D. C. (1993). Pace
perception after brain injury. Brain. 116.941-959.
Young, A. W., Rowland, D., Calder, A. J., Etcoff, N. L , Seth, A., & Perrett, D. 1.
(1997). Facial expressions megamix: Tests of dimensional and category accounts o f emotion
recognition. Cognition. 6 3 .271-313.
Zahn-Waxler, C., Cole, P. M., Welsh, J. D., & Pox, N. A. (1995). Psychophysiological
correlates of empathy and prosocial behaviors in preschool children with behavior problems.
Development and Psvchonathologv. 7 .27-48.
Zubieta, J. K., & Alessi, N. E. (1993). Is there a role o f serotonin in the disruptive
behavior disorders? Joumal of Child and Adolescent Psvchopharmacologv. 3 (1). 11-35.

97

Appendix A

The Positive and Negative Affect Schedule (PANAS)

98

THE PANAS
The following list of words describes different feelings and emotions. Read each item, and then
mark the appropriate answer in the space next to that word. Indicate to what extent you
generally feel this way, that is, how you feel on the average.

1
very slightly
or not at all

2
a little

3
moderately

interested
distressed
excited
upset
strong
guilty
scared
hostile
enthusiastic
proud
irritable
alert
ashamed
inspired
nervous
determined
attentive
jittery
active
afraid

99

4
quite a bit

5
extremely

Appendix B

The Toronto Alexithymia Scale-20 (TAS-20)

100

TAS-20

Indicate how much you agree or disagree with each o f the following statements by marking the
appropriate space with a check mark. Give only one answer for each statement.

1
strongly
disagree

2
moderately
disagree

1. I am often confused
about what emotion I am
feeling.
2. It is difficult for me to
find the right words for
my feelings.
3. I have physical
sensations that even
doctors don’t understand.
4. I am able to describe my
feelings easily.

5. I prefer to analyze
problems rather than just
describe them.
6. When I am upset, I don’t
know iflamsad,
lightened, or angry.
7. I am often puzzled by
sensations in my body.

8. I prefer to just let things
happen rather than to
understand why they
turned out that way.

1
101

3
neither
disagree or
agree

4
moderately
agree

5
strongly
agree

1
strongly
disagree

2
moderately
disagree

9. I have feelings that I
can’t quite identify.

10. Being in touch with
emotions is essential.

11. 1 find it hard to describe
how I feel about people.
12. People tell me to
describe my feelings
more.

13. 1don’t know what’s
going on inside me.

14. I often don’t know why 1
am angry.

15. I prefer talking to people
about their daily
activities rather than
their feelings.
16. I prefer to watch “light”
entertainment shows
rather than psychological
dramas.
17. It is difficult for me to
reveal my iimennost
feelings, even to close
friends.

102

3
neither
disagree or
agree

4
moderately
agree

5
strongly
agree

1
strongly
disagree

2
moderately
disagree

18. I can feel close to
someone, even in
moments of silence.
19. I find examination of my
feelings useful in solving
personal problems.
20. Looking for hidden
meanings in movies or
plays distracts from their
enjoyment.

103

3
neither
disagree or
agree

4
moderateiy
agree

5
strongiy
agree

Appendix C

Consent Form

104

Informed Consent Form
Note: All research involving human participants at UNBC falls under the authority o f the
Human Research Committee. The University and those conducting this research subscribe to
the ethical conduct of research and to the protection at all times o f the interests, confidentiality,
comfort and safety of all participants.
Present Study: Incarcerated Youth and the Identification of Posed Facial Expressions of
Emotion.
Researcher’s Personnel: If you have any questions or concerns regarding study procedures and
questionnaires please feel free to contact Dr. Prkachin at 960-6632 or Ms. Fenner at 565-4153.
Purpose: The purpose of this study is to determine if there is any relationship between
incarcerated youth and the ability to identify posed facial expressions o f emotion under fast
conditions.
Task Requirements: All participants will be required to 1) watch videos o f various facial
expressions o f emotion, 2) fill out two brief questionnaires, consisting o f 20 short questions
each, one questionnaire before watching the video and one after the video. Both questionnaires
regard feeling and emotions.
Duration: To complete this study will take participants approximately seventy minutes (forty
minutes for watching the videos, ten minutes to fill out each of the questionnaires.
Potential Risks: There are no known risks associated with participation in this study.
Anonvmitv/Confldentialitv: The data collected in this study will remain anonymous, and
individual data will be available only to project staff. No individual data will be released to
anyone.
Right to Withdraw: If at any time during the experiment you should feel uncomfortable you
may withdraw without consequence.
Participant Restrictions: Participants must be sentenced. Those who are in remand, facing
other charges or appealing their sentence cannot participate.
1have read the above description and I understand the conditions o f my participation. My
signature indicates that I agree to participate in this study.

Signature

Date

Witness

Date

105

Appendix D

Table 3.6 Two-Tailed T-Tests for Paired Emotion Comparisons for Within-Subject
Effects

106

Table 3.5
Two-Tailed T-Test for Paired Emotion Comparisons for Within-Subiect Effects

Incarcerated

Non-incarcerated

Emotion Pairs

t Value

D.F.

Sig.

t Value

D.F.

Sig.

Happy*-Anger

-7.49

31

.000

-8.99

30

.000

Happy*-Disgust

-5.18

31

.000

-5.93

30

.000

Happy*-Fear

-16.23

31

.000

-11.75

30

.000

Happy*-Sad

3.83

31

.001

9.86

30

.000

Happy*-Surprise

4.00

31

.000

4.83

30

.000

Surprise*-Anger

-5.12

31

.000

-7.16

30

.000

Surprise*-Disgust

-3.10

31

.004

-3.21

30

.003

Surprise*-Fear

-13.56

31

.000

-10.38

30

.000

Surprise-Sad**

-1.18

31

.246

-4.67

30

.000

Sad*-Anger

-4.48

31

.000

-4.75

30

.000

Sad*-Disgust

-2.39

31

.023

.27

30

.789

Sad*-Fear

-14.46

31

.000

-9.02

30

.000

Disgust*-Fear

6.22

31

.000

9.05

30

.000

Disgust-Anger**

-1.55

31

.132

-4.56

30

.000

Anger*-Fear

4.30

31

.000

5.75

30

.000

Note: Astéries (*) indicates the significantly better mean A1 for paired comparison for both
samples. Double astéries (**) indicates the significantly better mean A’ for the paired
comparisons for the comparison sample only. Significance (Sig.) at alpha .05.

107