1“
_. .

_ 111:1 1
w " '

131.1.

3

i1:

_ 13.1
f

,.:

"9'.

.'.'1E
. '11‘. .
1.5””

‘1
(J?
H

.121-..

 

1-14

. ! 1-D. . 1
. - . .H . . . . . . . . - 1.. .11. 1.1 .N. 9...}... 1.41.3.1... .
. .. .. . h .115 up...“ . . 1.1%»..11uﬁ11usmﬁ1#.1n.....w.ﬁu¢.u1na.5. 1.. . .51.:
. ..1%Wn . . . . 1 1.. .. . . . 1. 1..
.. .. 1 v. c 11

'L" 11?.
“R’vﬂve.
. A

11

ﬁ....~¢x.qu.._..u.1.. . 1
_.... 3.... .1
an... 3 .5..an .
333.. .1. .._. ,
wﬁa...m.m.u.ﬁ 33... 3

.I r . . 3 LA. at. 3
11%...ﬁnuﬂmﬁrau... 1.1.4.5115...

 

. s. ... . 11.13.. . .
. . .... . . _ ....-.11..1.....3..3 3.3.3.3331“...
1w»... .. . ._ . . 1. 1.3... 13333313. _ .31 3.33.13.13.31
3.11%.“? .. ... . .3 ”1.3.3.3113... 1.1.1.. 1.. z .. :33. . p

.
. . . . . ... and.” $.54...

. _. $1,... .1... .. . .. , 1 13.341. 11.11.... .1... f...

_ Hfmiwwkgfw .. .13» . . . 1. .rmmhﬁ.w.1:. «Mg...

. .H .. 1.3..” .31.... .
1 ..1 .. . .1 3. . . . . . .. 1.341....er 3 1 .. . .3 . «Judgéﬁa
1.33.13.33.41... .3. . 1...... _. 3.3. .. m. 1"... 11. .. ”.33....13... . 1. - 1.- .111
. iiwuwh... . .1. .. 1. 1. 2 . . .1... .1......1.....”.mm ..... 3.3.3.“..33 . . 1.. . . .. 1 . .. .11....1.1.r11.1.1....1..... 11.11...
. .3W1...nwu...w..1 .. .. immwﬂ... ﬂaw.” 1.3.1.33??? . . .. «€31.13.

1. .. .c .. t.. .u. 1 a. .- .oﬂlk. . 12.35%)... 11051:... .
.. . . 11133.3 2.33. ﬁﬁwﬁgﬂwﬁm
. . . . 1

..I I

 

 

.. Rm HF1... 1. .. .-..1. .n... .1...1 a... .1 n.
. 1 a. . .331...

1.
L.
0..

1. _ . . . 1 . ... . . .. . 1:...
33%“... 1.3%....3331»! $1.12.. .31- .. .. . _ .. . .
.1. .3. ..... .7. ....1. .. .. .. . . .. . 1 .. .. . . ..
1 c 1:13.311". Emutwwﬁwmno .1“ yr? 5 .: . . . . .1 .
A1 .f

a... :1 .11....1..§ .1
.3115... 111.111.. 1.1. .1». . ..
.. £“m.m1..4.mmw3....-mw.ﬁ a. .. u. u. .

1 1.11.11»... L... .1 .33». 1

1.. 11.1.... 11.33:... 1.. . 1

. . .. tr: 1.... .1. . . .. . .1 . 5.....1..1d1..1.1..... 11.1.1.1.
. 21.3.13 1 . .. .1 .m. 1 1. . 11 . 33311.1...13191 . : . .. ....1.1.. ......u.....u.1111...1:..1..1...
1.4.... .. 1... 1.1 1 .1 .. 1.... 1“. $313.1... .. . . ,. . .....h....11.uﬁ.mm.... . . .
.1... a. 1....1 35.41.41.131; Rm... 9341 .. . . .. . . .. .. .... 13.1.1.3} .. .
.. .32 .15 $91.13. ... 1 f1 , . 1
. 1 l . .

3».- x.
1. ..
.. .1. 19¢}: .

2. b1.

 

.aa

  

.01.:1111115“.

 

u E
‘s..
1....

 

. . 3.11-1.11 .1011511}. . . . . 1.11.
9 11.... {.13. s 115.11.. «11.11.4115.
. .............ru.......b. ﬂawﬂfﬂdwwuma “1...... 13mg»... .. 1. a...»
.5. 1
. n Itirs‘hh

2.1.3.16...»

 

1.1” . 1W1»...
2... .13.»... 1 urv..|1
.1 . .1... 1%..111. . l... .

.. 1 .....uw.... .511...n.n1...1.
131.1»... .11. 1 . .
.. 1.... 1...... .315... 1
. 1. . 1..” 11".... 13.11.11...

 

. u .111 .1.
1.3.... 1-

v . . 1. .1 . .. £2... 1 1...
. 1.... . 1.1%. . 11.1. 1.1 11.11 .
ruwﬁauuiu3udhnﬂrn1111u: . . u. 1111.11.13! .11
.1 .1u111...11:1..11.. .. .1 :11... . . . . . . ..1..1w.1.1.1 . . . . . .
1 1 ..1I).1. 0.1 1. .11. . .1 21. 4f: 1 1. 1 . 1 1. . . ’1
11—32191! - . . 1 1 «t 1 1 1 . . 1 . . .. 1 Ikavnlhduid .. 1 ..I. .
. . . :1. 11.81.1111 . . . .. 1.5.1.. 1.11.3111... 1.1--. .111... 1.1.1.1131.
.313... 4.35.3.1... . .. . .. . . .. . 1. .nn...1.1h......1v11.1 ....1. ...........n..ﬂ1vL.1 1- 311...... .3. 3.11 .
h. 1. . . . .L
.. c .-

“£_ 1' _.

 

1‘1.

. :

 

V

 

11.1.- 1-11... 1
it. 1: . 1. . .11...- 1-. .1111).
1.11... :1 .111... 11.2.2.9. .11....1......1u91...1|. ..1....(. 1......
cl 1 . 1.3:!11: ’11:..01111‘ ... A 1 3 Svllol 1. . 1 .
11M. 1.... 14.1.1.6--. .113... 1.... . _ 1.1:; .. ... . . . 111.111.111.1112115111111111
.... .... . .1. . . .. . . . . . . . .. 31.111.110.311.-. .1111
.. ...~.v.¢1....1r..1m.m.1.1....111...1 . . . . .. . . . 1.1.1.114... «.31 x
1-0. . v6\5.f.lln .1 1 . . 1%] .
4.13. .i . 1 ..

Q

 

~i :
. 311v

 

.1111-

 

.

. . . 1..-.

1.1.1... 1... ..

.rd..1..¢§...... 13.132
.11.-. .1. .1...
. 11.1.

a“. .1...1.»....1«mm.
.11.... daﬂhnv 1.31.3”: 1 .. =1
1 .. . 1...: . .htﬁr . r1. 3.41.}...-
.. 1&1. .. 1.. .51 . $114.. .u... -11.
. 1 11-..... 3......
. .. . . . _ ..1.........%.1n1k.11..1n11..;.1.1.1_.. «.11....
. . 1.. ('1... .. .11.. 1. 14.111
. .1 .11.. 1 1. .11|.u...u....m3 1...». “913.1%...5ﬁhﬁﬁﬁ .1........1..1..1u..1..1
1A b-4161. ml? .1 1111!. .. .VAﬂn)1133 5%.‘1111.
11.1.11... 1.1.1.113unf91 ”111.112.... 9..

1.1%.“.qu 1131 .L. . IF . .ﬁuMm. .113..P.1.....M..1..111.11....u11he

 

p1

 

 

 

llvd II 110
O 1I\f‘11 . iﬁlvk 1.
6|. 1. .I 1 . e1 .3 .

.1...

n... 1). . . . .. 1 . .1 tn“... 1
. . 1. 1. 1
.2... huh... 1 1 . . . . 1.111”

.11. 1.1.1.1110
1...: 1 111.
lift

1.f.‘

I.‘

1.1.

..;1:
.1.

 

L1...1.13}.1w113....a§3.$37¢¢. .1. g .. $313.3 Eunuu.r...._.,..._1:n.1Wm... ... 1.1. .1 , - , .. .mwww , 1 r 1:. . 1- 1 11...-.. 11.. .. 1 - ..

 

lMC CH IGAN STATE

mass IIIIIII III IIIIIIIIIIIIIIIII IIIII

1301570 3220 I

3LZQIBRAFIY

MlchiganState
University

 

 

 

 

 

 

 

 

 

 

 

 

 

 

This is to certify that the

thesis entitled

THE EFFECTS OF UNILATERAL FACIAL MANIPULATION AND
SEX 0F SUBJECT ON THE EXPERIENCE OF EMOTIONS

presented by

Joshua Marcus Cohen

has been accepted towards fulﬁllment
of the requirements for

M.A. degree in Psycholoqy

Mmm

ajor professor

Date 6/3”;

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

_ _ v v v - — a

H v v Q .-

v - O O - 0 w o

 

PLACE u RETURN 30x36 mm“; {mi checkout iron your record.
TO AVOID FINES mum on or More 6% drama I

DATE DUE ""BATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU Is An Afﬁrmative ActioNEqunl Opportunity Institution
WM‘

 

 

, —————

THE EFFECTS OF UNILATERAL FACIAL MANIPULATION AND SEX OF
SUBJECT ON THE EXPERIENCE OF EMOTIONS

By

Joshua Marcus Cohen

A THESIS

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

MASTER OF ARTS

Department of Psychology

1 997

ABSTRACT

THE EFFECTS OF UNILATERAL FACIAL MANIPULATION AND SEX OF
SUBJECT ON THE EXPERIENCE OF EMOTIONS

By

Joshua Marcus Cohen

Facial manipulations were used as a mood induction technique to examine the
Valence Hypothesis of Emotion and possible sex-related diﬁ‘erences in cerebral
lateralization for emotion. Two primary predictions were that 1) subjects would make
more negative responses following left-face manipulations (right hemisphere) and more
positive-like responses following right-face manipulations (left hemisphere) and 2) that
men would be more clearly lateralized than women on the mood induction task. Baseline
data was gathered from 40 male and 40 female subjects using the DACL and POMS-BI.
Subjects manipulated each side of their face (alternating sides each time), completed
parallel forms of the DACL following each manipulation, and completed a second POMS-
BI following the last manipulation. Data were tested using W
and ﬁndings did not support either prediction. Possible explanations are given for the
unsupported hypotheses, clinical implications of this research are discussed, and future

research possibilities are suggested.

DEDICATION

I dedicate this Thesis to my wife Sandy. It represents not only my hard work but
also Sandy’s unwavering support and timely encouragement. Without Sandy I would not
have reached this seminal point in my academic career. I am both deeply in love and
forever grateful.

ACKNOWLEDGMENTS

First, I must thank the chairperson of my committee for his tireless efforts at
reading and revising my manuscripts; thank you Professor Lauren Harris. Next, I would
like to thank Professor Anne Bogat who offered a fresh perspective on the manuscript at
those times when I could not see beyond the written page. Professor Bogat also guided
me through the statistical areas in which I needed assistance and for that I am very
thankful. Next, I would like to thank Professor Alytia Levendowsky who approached this
project patiently, always offered words of encouragement, and read the text thoughtfully.
Her suggestions helped me to strengthen sections that were previously incomplete and
helped make the ﬁnal product a better product. Finally, although not a formal committee
member, Professor Gersh Kauﬁnan offered his support and advice when I needed to
progress through difﬁcult times. The phone calls he made to my home offering his
support, and our many discussions about the most eﬁcient ways to accomplish my goals
were tremendous factors in my ﬁnishing this project. Thank you from the bottom of my
heart.

Furthermore, I would like to thank Travis Fogel for all of his eﬂ‘orts. In addition
to sharing the task of data collection and data entry, Travis often played the role of a

“thoughtful devil’s advocate,” ﬂushing-out complex ideas and making diﬁicult decisions

iv

easier.

Also, I would like to thank Ken Brown, a colleague who gave of his time simply
because he chose to be helpful. Ken’s organizational skills and research experience were
clearly demonstrated when he assisted in setting up data ﬁles and offered statistical advice.

I also want to thank my parents, Bill and Ricki Cohen for the unyielding support
they gave me not only during graduate school but throughout my life. They laid the
ground that enabled me to achieve goals like this one. Thank you both.

Finally, to the most important person in my life, my wife Sandy. During the good
times of this process Sandy was my bes. ﬁiend and my most enthusiastic supporter.
Likewise, during the diﬁcult times she was my best friend as well as my shoulder to cry
on. She never failed to oﬁ‘er me words of encouragement and she inspired me always to
perform at the best of my ability. If her support was not so strong, such heights may never
have been reached. Sandy has gone through this process with me and for me. For all of
the reasons I have just stated, and all of the reasons that I chose to make Sandy my partner

for life, I dedicate this massive project to her.

TABLE OF CONTENTS

LIST OF TABLES ........................................................................................................... x
LIST OF FIGURES ....................................................................................................... xii
CHAPTER 1: INTRODUCTION .................................................................................... 1
Sex-Related Diﬁ‘erences in Cerebral Organization for Language and
Visuo-Spatial-Perceptual Functions ................................................................... 1
Sex-Related Diﬂ‘erences in Cerebral Organization for Emotion? ............................ 3
Neuropsychological Theories of Emotional Processing ......................................... 8
Evidence for the Right Hemisphere Hypothesis ......................................... 9
Evidence for the Valence Hypothesis ....................................................... 10
Mood-Induction Studies .......................................................................... 13
Sex-Related Differences in Laterality Studies of Emotion .................................... 20
Studies of Perception of Emotion ............................................................ 20
Studies of Expression and Experience of Emotion ................................... 22
Rationale ............................................................................................................ 25
Why Such Inconsistent Findings? ............................................................ 25
Clinical Implications ................................................................................ 27
CHAPTER 2: CURRENT EXPERIMENT .................................................................... 30
Primary Predictions ............................................................................................. 30

Right Hemisphere Hypothesis vs. Valence Hypothesis ............................. 30

Sex-Related Diﬁ‘erences .......................................................................... 31
Sex-Related Response Patterns ............................................................... 31
Secondary Predictions ........................................................................................ 32
Timing of Manipulation (part 1) .............................................................. 32
Timing of Manipulation (part 2) .............................................................. 32
Timing of Manipulation and Sex of Subject ............................................. 32
CHAPTER 3: METHOD .............................................................................................. 33
Subjects .............................................................................................................. 33
Design ................................................................................................................ 33
Materials ............................................................................................................ 34
Procedure ........................................................................................................... 39
Statistics ............................................................................................................. 42
CHAPTER 4: RESULTS .............................................................................................. 43
POMS-BI Factor Analysis .................................................................................. 43
Demographic Information ................................................................................... 45
Results of 8-item Handedness Questionnaire ....................................................... 45
Results of Primary Predictions ............................................................................ 45
Valence Hypothesis ................................................................................. 46
Sex-related diﬁ‘erences ............................................................................ 47
POMS-BI .................................................................................... 47

Sex-Related Response Patterns ............................................................... 48

Results from Secondary Predictions .................................................................... 48

Timing of Manipulation (part 1) .............................................................. 48
Timing of Manipulation (part 2) .............................................................. 49
Timing of Manipulation and Sex of Subject ............................................. 50
CHAPTER 5: DISCUSSION ........................................................................................ 51
Discussion of Primary Predictions ....................................................................... 51
Valence Hypothesis ................................................................................. 51
Sex-Related Differences .......................................................................... 54
POMS-BI .................................................................................... 57

Sex-Related Response Patterns ............................................................... 58
Discussion of Secondary Predictions ................................................................... 59
Timing of Manipulation (part 1 and 2) ..................................................... 59
Timing of Manipulation and Sex of Subject ............................................. 61
Clinical Implications ........................................................................................... 62
Cognitive-Behavioral Approach .............................................................. 63

Aﬁ‘ect Theory Approach ......................................................................... 63

Future Research Directions ................................................................................. 65
APPENDICES ............................................................................................................... 69
Appendix A Informed Consent F orm ................................................................ 69
Appendix B Subject Identiﬁcation Form ........................................................... 71
Appendix C Medical History Questionnaire ...................................................... 72
Appendix D Personal Information Questionnaire .............................................. 73

Appendix E Laterality Questionnaire A & B ..................................................... 74

Appendix F POMS-BI Form ............................................................................. 76
Appendix G DACL lists A, B, C, D, G .............................................................. 77
REFERENCES .............................................................................................................. 95

Table 1a.

Table 1b.

Table 2.

Table 3.

Table 4.

Table 5.

Table 6.

Table 7.

Table 8.

Table 9.

Table 10.

Table 11.

LIST OF TABLES

Right Hemisphere Hypothesis .................................................................. 82
Valence Hypothesis ................................................................................. 82
Correlations among ﬁve state lists for male and female normals ............... 83
Correlations among mood scales from Proﬁle of Mood States-

Bipolar .................................................................................................... 83
POMS-BI retest correlations and mean scores ......................................... 84

Results of Conﬁrmatory Factor Analysis for a six-factor model of
the POMS-BI .......................................................................................... 84

Correlation matrix for positive and negative adjectives selected for the
Factor Analysis on the POMS-BI ............................................................ 85

Factor loadings, reliability analysis, and alpha variance if item deleted
for positive and negative factors on the POMS-BI ................................... 86 ~

Results of Repeated Measures Analysis of Co-Variance for Condition
(RLRL, LRLR), Time (DACL B, C, D, G), and Sex (male, female) ......... 87

Mean DACL scores and SD for men and women in the RLRL and
LRLR conditions ..................................................................................... 87

Results of Simple Factorial ANOVAs on the second administration
of the POMS-BI, using negative and positive adjectives, and
co-varying the ﬁrst administration ........................................................... 88

Mean number of adjectives endorsed on each DACL (A, B, C, D, G)
and results ﬁom a Simple Factorial ANOVA comparing response
rates on each DACL (A, B, C, D, G) for men and women ....................... 89

Table 12. Mean and SD of absolute diﬁ‘erence scores for all subjects in either the
RLRL or LRLR condition ................................................................. 89

Figure 1.

Figure 2.

Figure 3.

Figure 4.

Figure 5.

LIST OF FIGURES

Percent of male and female subjects in each age category, year in
school, and ethnicity/race category .......................................................... 90

Mean responses from men and women for hand-preferences
on eight tasks .......................................................................................... 91

Interaction between Time (DACL B, C, D, G) and Condition
(RLRL, LRLR), co-varying pre-measure DACL-A .................................. 92

Mean DACL scores for left-face manipulations summed across
conditions and mean DACL scores for right-face manipulations

summed across conditions for both men and women ................................ 93

Mean DACL scores for men and women in the RLRL and LRLR
conditions ............................................................................................... 94

xii

CHAPTER 1
INTRODUCTION

After more than a century of research dating back to the 18603, there is now
abundant evidence that the two cerebral hemispheres of the human brain play diﬂ‘erent
roles in the initiation, regulation, and execution of certain higher ﬁinctions (Benson &
Zaidel, 1985). The general conclusions ﬁom this research credit the left hemisphere with
primary control over language and the right hemisphere with primary control over visuo-
spatial and perceptual functions. Language and spatial perception are conventionally
regarded as mgnitixe ﬁrnctions. In the case of ﬁrnctions that are not so regarded, namely
feelings and emotions, laterality studies do not allow such clear distinctions. Some
evidence points to right hemisphere dominance for all emotional behaviors; other evidence
suggests that both hemispheres contribute in different ways depending on the valence of
the emotion, that is, whether it is predominantly positive or negative. One of the aims of

this study was to help elucidate this point.

 

For language and visuo-spatial-perceptual functions, the question also has been
raised whether generalizations about cerebral organization apply equally to males and
females. One reason for raising this possibility are the well—documented sex-related
diﬂ‘erences in performance on verbal and visuo-spatial tasks. For instance, males typically

1

2

perform at a higher level than females on tasks with a strong visuo-spatial component.
These include tasks requiring mental rotation, target-directed motor skills (e. g., guiding
and intercepting a dart), disembedding simple shapes from complex pictures, and right-left
discriminations. Females, on the other hand, typically perform at higher levels than males
on tests of perceptual speed, memory for location of common objects, verbal ﬂuency, and
ﬁne-motor skills (e.g., Benton, 1959; Harris, 1979, 1981; Harris & Gitterman, 1978;
Maccoby & Jacklin, 1974; Kimura, 1992).

Evidence suggests that at least some of these differences are related to upbringing
and social training. For example, boys are oﬁen encouraged to engage in activities that
incidentally can enhance visuo-spatial skills (Sherman, 1976; for review see Harris, 1979,
1981). Neuropsychological evidence, however, raises the possibility that the more
fundamental force may lie in sex-related differences in cerebral organization. This
evidence, from both the clinical and non-clinical literature, suggests that the general
conclusions cited earlier about laterality and cognitive ﬁmctions apply more strongly to
males than to females (e. g., McGlone, 1980; McGlone & Kertesz, 1973; Kimura, 1992).
Put another way, language and spatial ﬁinctions tend to be under a greater degree of
unilateral control, or a lesser degree of bilateral control, in males than in females. What
still remains to be shown, however, is precisely how such diﬂ‘erences in cerebral
organization might be translated into performance differences in verbal and non-verbal
tests of the kind cited above. If, for example, greater unilateral organization gives males
an advantage on spatial tasks because the right hemisphere is more purely committed to
non-verbal processing, it is not clear why the same should not apply in the case of left

hemisphere skills, that is, why males should not have the same advantage for verbal skills

3

as well. There appears to be slippage between the neuropsychological evidence and the

cognitive performance data.

 

Given the evidence for sex-related diﬂ‘erences in cerebral organization for language
and visuo-spatial ﬁrnctions, the question arises whether there may also be sex-related
differences in cerebral organization for emotion. Another aim of this study is to address
this possibility.

One reason to raise this possibility is that, just as the search for sex differences in
neurological organization for language and visuo-spatial processing has been spurred, in
part, by demonstrations of sex diﬁ‘erences in performance on language and visuo-spatial
tasks, so is there evidence for sex differences on tests of “emotional” functions.

Some of the evidence pertains to the perception of emotion. For example, it is
often said that women perceive emotional stimuli more readily than men, that they are
more likely than men to "see when something is wrong with a ﬁiend" or to "hear a
problem in a friend’s voice. " Such accounts are largely anecdotal, but the general theme
ﬁnds support in scientiﬁcally rigorous studies. For example, Soloman and Ali (1972)
found that women were better than men in identifying the aﬂ‘ective implications of tone of
voice. Soloman and Ali interpreted this to mean that females have a better understanding
of emotional situations. Buck, Savin, Miller, and Can] (1972) found that females were
better than males in identifying emotional expressions in faces, and Natale, Gur, and Gur
(1983) found that females were better than males on tasks that required discriminating
between faces depicting diﬁ‘erent expressions.

Other evidence suggests that females and males also express and W

4

emotions diﬂ‘erently. For instance, researchers using data from psychotherapy clinics have
found that females report more symptoms of depression than males (Klerman, Weissman,
Rounsaville, & Chevron, 1984). Also, surveys have found that females report higher
average levels of depression and anxiety compared to males (Aneshensel, 1992; Mirowsky
& Ross, 1986) and exhibit more warmth and compassion than males (Deaux & Lewis,
1983; cited in Deaux, 1984; Intons-Peterson, 1988). It is unclear, however, whether these
data reﬂect a sex-related difference in the expression or in the experience of emotion, or
perhaps both.

Most attempts to account for the kinds of sex diﬁ‘erences cited earlier see them as
products of social training and experience. Three such attempts are the remme-bim
theory, the gendered-response theory, and the structured-strain theory. In support of the
idea that the main difference between the sexes is in greater female expression of
emotions, the response-bias theory proposes that women and men differ in their likelihood
of expressing emotions or feelings to an interviewer (Phillips & Segal, 1969; Ritchey,
LaGory, & Mullis, 1993; Seiler, 1975). According to Mirowsky and Ross (1995), the
reason is that, for women, emotional expression does not have a negative connotation.
Instead, it symbolizes the transmission of signals between connected individuals. To men,
however, the expression of emotion does have a negative connotation because such
behavior may feel to them that they are revealing weaknesses. According to this
hypothesis, womens’ greater apparent distress results primarily from differences in their
willingness to talk openly about distress and not from differences in the amount or quality
of emotional experience. One type of data oﬁ‘ered in support of this hypothesis shows

that, in experimental and survey situations, when men and women are asked to report how

5

happy they are, women do not report less happiness than males (Wood, Rhodes, &
Whelan, 1989; Stoppard & Gruchy, 1993). It would be difﬁcult to reconcile the
contradiction that women are more distressed than men but equally happy.

Where diﬁ‘erences in the W of emotion are concerned, the gendered-
response theory suggests that for negative emotions, the type of emotional response may
be different when depression is inﬂuenced by sex-role socialization (Mirowsky & Ross,
1995). Speciﬁcally, depression is thought to be the by-product of ﬁustration of desires
and aspirations and that this ﬁ'ustration generates rage and hostility (Rosenﬁeld, 1980).
Depression is seen when an individual focuses the anger inward, punishing the self for the
perceived failure and deﬁciency. Furthermore, men are generally socialized for aggressive
and combative roles that permit the outward expression of anger and hostility whereas
women are generally socialized for nurturing and supportive roles that discourage such
expression. Thus, the gendered-response theory pr0poses that frustration makes men
angry and women depressed (Dohrenwend & Dohrenwend 1976, 1977). Also, it has been
proposed that standard indices for measuring distress ask more questions about depression
and anxiety than about anger and hostility. Such question biases would make females
appear more distressed than males even though they may experience the same amount of
overall distress (Rosenﬁeld, 1980).

Although the gendered-response theory may seem intuitively reasonable, some
data have not supported it. In a clinical study, Frank, Carpenter, and Kupfer (1988)
followed women and men through treatment for recurrent depression and found that when
entering treatment, women reported signiﬁcantly more anger and hostility than men. In

another study Conger, Lorenz, Elder, Simons, and Ge (1993) reported results ﬁ'om a

6

survey of 451 rural Midwestern married couples that showed that wives reported
signiﬁcantly greater marital hostility than their husbands. Although the survey results
were interpreted as reﬂecting sex-related diﬁ‘erences in experience, the possibility exists
that these survey ﬁndings are the product of sex-related reporting bias, making it possible
that both an experience and reporting bias aﬁ‘ected these results.

The third explanation of sex-related differences in emotion is oﬁ‘ered by the
structured-strain theory, which incorporates both the expression and experience
diﬁ‘erences described above. The structured-strain theory proposes that gender inequality,
gender roles, and gender-based frequency of contact with social stressors account for
diﬁ’erences in the amount of distress reported by men and women. Gove and Tudor
(1977), Pearlin (1989), and Ross and Huber (1985) claim that women’s positions in the
labor force and in the family disadvantage them compared to men because of the women’s
greater burden of demands and limitations. This pressure creates stress and ﬁ'ustration
and becomes evident in higher levels of felt- and reported-distress.

While not denying the important contributions and inﬂuence of social training and
response-biases to sex differences, as represented by the response-bias, gendered-
response, and snucnired-strain theories, the evidence for sex differences in cerebral
organization for language and visuo-spatial ﬁrnctions at least raises the possibility that
these differences in emotion likewise reﬂect some degree of cerebral “hard-wiring.” Even
so, environmental inﬂuences should not be assumed to play an insigniﬁcant role.

Where sex-related diﬁ’erences in cerebral “hard-wiring” are concerned, in some
mammals and birds the diﬂ'erences appear to have their origins in at least two periods of

development: 1) prenatal and 2) around the onset of puberty. Prenatally, hormones set in

7

motion the process of organizing the sexually undifferentiated brain into distinct “male”
and “female” type brains (Harris, 1981).. At the onset of puberty, when many distinctly
male and female characteristics present more clearly, hormones again can be found in the
central nervous system. in increased quantities. Although these ﬁndings come from studies
of animals and birds, they may also hold true for human beings (for a more detailed
explanation see Harris, 1979, 1981). Examples of the environmental inﬂuences that are
also assumed to impact these processes were explained when the response-bias,
gendered-response, and structured-strain theories were discussed and, therefore, will not
be covered here.

Epi-genetic models like the ones discussed above have been the subjects of
research studies in a variety of ﬁelds. For instance, some researchers use epi-genetic
models partially to explain sex-related diﬁ‘erences for approaching tasks such as divergent
thinking, object synthesis, and ideational ﬂexibility (Coren, 1995). Clinical
psychopharrnacologists have studied possible sex-related diﬁ'erences in the behavioral and
physiological mechanisms of responses to threat and stress, hypothesizing that those
diﬂ’erences contribute to the diagnostic over-representation of women in anxiety disorders
(Blanchard, Greibel, & Blanchard, 1995). In both examples, it is believed that the
physiological and behavioral/environmental systems interact by reinforcing the other.

The rest of this manuscript reviews the evidence for possible cerebral diﬂ’erences
for cerebral organization of emotion. To begin, a review of two hypotheses about cerebral
organization for emotion will be presented. After that, a review of the evidence for sex
diﬁ‘erences in this organization and discussion of some possible explanations for

inconsistent results is presented. Following this, is a discussion of the rationale for a new

8

study including possible clinical implications for studies of the neuropsychological
organization of emotion. Finally, a description is given of a new experimental test of the
sex-differences hypothesis.

H II°III'EE'IE'

Historically, the brain region named as the primary area for the initiation and
control of emotional behavior has been the limbic system (Rinn, 1984; Joseph, 1992).
More recent research shows that nee-cortical regions also play key roles. Some of the
research has focused on the peneepﬁpn of emotion (e.g., Bever, Hurtig, & Handel, 1976;
Safer & Leventhal, 1977; Landis, Assal, & Perret, 1979; Coﬁ‘ey, 1987), some on the
Men of emotion (e.g., Campbell, 1978; Sackeirn, Gur, & Saucy, 1978; Borod &
Caron, 1980; Rubin & Rubin, 1980), and some on the W of emotion (e.g.,
Davidson & Schwartz, 1976; Davidson, Schwartz, Saron, Bennett, & Goleman, 1979).
Such studies have given rise to two main hypotheses about hemispheric specialization for
emotion: the so-called Right Hemisphere Hypothesis and the so-called Valence
Hypothesis. These hypotheses are summarized in Tables la and lb.

According to the Right Hemisphere Hypothesis (Table 1a), the right hemisphere is
specialized for the perception, expression, and experience of emotion and takes the leading
role regardless of the Meme of the emotions, that is, whether it has a more positive or
negative quality (Ley & Bryden, 1982; Borod, Koﬁ', & Caron, 1983; Buck, 1984;
Heilman, Bowers, & Valenstein, 1985).

The Valence Hypothesis (Table 1b), follows the Right Hemisphere Hypothesis in
proposing that the right hemisphere has primary control over the perception of emotion,

regardless of valence. It then proposes diﬂ’erent hemispheric roles for expression and

9

experience, depending on valence, with the left hemisphere having primary control over
positive emotion, the right hemisphere for negative emotion (Borod, 1992).
E'I EIE'IH'IHI'

Evidence for the Right Hemisphere Hypothesis comes ﬁ'om a variety of sources.
Evidence that the right hemisphere plays the leading role in the perception of emotion
includes dichotic listening studies showing a left ear, or right hemisphere, advantage for
perceiving non-speech, positive and negative sounds such as laughing and crying (Kimura,
1964; King & Kimura, 1972), as well as positive and negative aﬂ’ectively-toned speech
(Haggard & Parkinson, 1971). Similar hemispheric diﬁ‘erences have been reported in
visual-recognition studies. For example, using a tachistoscopic method‘, Strauss and
Moscovitch (1981) found a left-visual ﬁeld advantage (right hemisphere) in subjects'
pereeptipns of facial expressions, both smiles and frowns, depicted in photographs,
cartoons, and line drawings.

A proposed limitation of the tachistoscopic method is that it necessarily restricts
the subject’s head-movement and greatly restricts viewing time of the stimulus. Zaidel’s
(1975) Z-lens is less restrictive. As described by Springer and Deutsch (1993, p. 43), the
Z-lens is a contact lens that permits ﬁ'ee movement of the eyes, without a time limit, while
ensuring that only one hemisphere receives the visual information. Regardless of the
proposed limitation mentioned above, where the visual recognition studies are concerned,

studies using the Z-lens have often produced results consistent with those using the

 

'A tachistoscope projects visual information onto a screen in either the subject’s right- or
left visual ﬁeld, allowing only one hemisphere to receive the stimuli.

10
tachistoscopic method (e.g., Dirnond & Farrington, 1977).

Some of the evidence for the Right Hemisphere‘Hypothesis for emotional
expressipnpomes from clinical studies. For example, Borod, Koﬁ‘, Lorch, and Nicholas
(1986) compared patients with right hemisphere damage, left hemisphere damage, and
normal controls on tests of facial expression in posed and spontaneous conditions. For
positive and negative expressions alike, patients with right hemisphere damage were
signiﬁcantly impaired compared to patients with left hemisphere damage and normal
controls.

Other evidence for the Right Hemisphere Hypothesis comes from non-clinical
studies. For instance, Moscovitch and Olds (1982) compared side of facial activation in
adults while they were engaged in spontaneous conversations. Any change ﬁom a neutral
facial expression was considered a valid data point. Therefore, both positive and negative
expressions were recorded (e.g., smiles, snickers, sneers, and ﬁowns). The results showed
that, overall, subjects had greater left than right side expressivity, indicating that facial
expression was mediated by the right hemisphere. Similar ﬁndings were reported by
Borod, Koﬂ‘, and White (1983). In their study, adults generally showed greater
expressivity on the left side of the face.

E . I II I I! I H I .

Inasmuch as the Valence Hypothesis is identical to the Right Hemisphere
Hypothesis for the perception of emotion, all the evidence on perception already cited
would apply here too. The only diﬁ‘erence between the models, therefore, would be for
studies of expression and experience, where ulenee is now hypothesized to play a role.

Support for the Valence Hypothesis comes ﬁ'om a variety of sources. One consists

ll

of studies of emotional changes in patients with lateralized brain injuries. Gainotti (1972)
found that patients with left hemisphere lesions were three times more likely than right
hemisphere-lesion patients to show catastrophic reactions, whereas right hemisphere-
lesion patients were more likely to be indifferent to their emotions or even to show a
“euphoric” reaction. Goldstein (193 9) originally characterized the catastrophic reaction as
"crying, pessimistic statements, guilt, feelings of nothingness, indignity, despair, and
complaints and worries about the ﬁrture," whereas the euphoric reaction constituted "lack
of comprehension, smiling, joking, laughing, mimicry, relaxation, optimism, and a sense of
well being" (cited in Silberrnan & Weingartner, 1986, p. 336; Goldstein page reference not
given). Additional support for these ﬁndings is reported by Robinson and Price (1982).
They found that patients with left hemisphere damage were more likely to be clinically
depressed than patients with right hemisphere damage. Robinson (1983) also found that
patients with right- but not left-sided strokes became inappropriately cheerﬁrl. Similar
differences were reported by Sackeirn, Greenberg, Weirnen, Gur, Hungerbuhler, and
Geschwind (1982) in a review of studies of pathological laughing and crying following
destructive lesions. Left hemisphere lesions were more likely to produce catastrophic
reactions, whereas right hemisphere lesions were more likely to produce euphoric
reactions.

Gainotti (1972) interpreted ﬁndings such as these to mean that the left hemisphere
is dominant for the experienee and expmsjen of positive emotions, whereas the right
hemisphere is dominant for the experienee and expressien of negative emotions. The
reasoning behind this interpretation draws on Hughlings-Jackson's (1874) inhibition

theory. That theory supposes that both hemispheres are normally in a state of balance,

12

with each exercising a measure of control, or inhibition over the other, as the site of
greater arousal passes ﬁ'om one side to the other. The catastrophic reaction associated
with left hemisphere damage thus is seen as a product of the left hemisphere’s release of
inhibition over the right hemisphere. The inverse is also true; the euphoric reaction
associated with right hemisphere damage reﬂects the release of right hemisphere inhibition
over the left hemisphere.

Support for the characterization that the left hemisphere usually inhibits the right
hemisphere, and that the right hemisphere usually inhibits the left hemisphere, has also
come from patients who have undergone the Wada procedure, a pharmacological method
that selectively anesthetizes one cerebral hemisphere at a time. Several early studies
reported that when the left hemisphere was anesthetized, the patient often exhibited a
”catastrophic reaction" whereas when the right hemisphere was anesthetized, the patient
more often showed a “euphoric reaction” (Alema, Rosadini, & Rossi, 1961; Perria,
Rosadini, & Rossi, 1961; Rosadini & Rossi, 1967; Terzian, 1964). The assumption is that
whatever behaviors are lost following this procedure lie within the province of the
anesthetized hemisphere, whereas whatever behaviors are retained or newly emerge are
under the control of the other hemisphere.

The problem with these data is that both the catastrophic and the euphoric
reactions occurred when the anesthesia was lessening, which means that the emotional
change could represent the actions of either side. In other words, it is possible that when
either hemisphere was "waking-up," it was able to produce the emotion usually under its
own control.

This problem of interpretation is clariﬁed by data reviewed by Sackeirn et al.

13

(1982) on epileptic patients during ictal episodes. The data revealed that left-side ictus
was more often identiﬁed with laughter, and right-side ictus with crying. Because, during
ictus, there is increased electrical activity, blood ﬂow, and metabolic activity associated
with the side of epileptic discharge, the laughter and crying are assumed to reﬂect the
direct activity of the left hemisphere and right hemisphere, respectively, and thus to be
consistent with the left hemisphere-positive, right hemisphere-negative Valence Model.

In summary, given all of the aforementioned research, there is ample evidence to
support the Valence Hypothesis.

I I H I . S I'

So far, we have reviewed a variety of methods for studying hemispheric
specialization for emotion and for testing the Right Hemisphere Hypothesis and Valence
Hypothesis. In recent years, still another method has come into common use, namely the
experimental induction of mood. This method will be discussed separately ﬁ'om the rest,
and, because it oﬁ’ers certain advantages over prior methods, it is the method used in the
current study.

Dozens of techniques have been used to induce mood-states, including the playing
of music (Thaut & de I’Etoile, 1993; Bouhuys, Bloern, & Groothuis, 1995; Lewis,
Dember, Scheﬁ’t, & Radenhausen, 1995), self-generated imagery (Harman & Ray, 1977),
the tachistoscope and Z-lens presentations of visual stimuli (Graves, Landis, & Goodglass,
1981; Landis, Assal, & Perret, 1979; Rizzolatti, Umilta, & Berlucchi, 1971; Geﬁ‘en,
Bradshaw, & Wallace, 1971; Dimond, Farrington, & Johnson, 1976), and the presentation
of words and other verbal stimuli (Baruch, Stolar, Heller, & Goldman, 1992). In general,

the results of these studies support the Valence Model.

14

For example, Dimond and Farrington (1977) used the Z-lens procedure to study
subjects’ reactions to ﬁlms presented selectively to each hemisphere. The ﬁlms were
either humorous, pleasant, unpleasant, or horriﬁc. The subjects’ heart-rates were
measured, with increases used as an index of an emotional response. The result was that
heart-rate changes were greater for ﬁlms presented to the right hemisphere than to the left.
The subjects independently also rated the ﬁlms as more pleasant when presented to the left
hemisphere and as more unpleasant and horriﬁc when presented to the right hemisphere.
The results thus are consistent with the Valence Hypothesis.

Further support comes from a study by Davidson, Schwartz, Pugash, and
Broomﬁeld (1976). They showed infants ﬁlms of a woman smiling or frowning. At the
same time, they recorded electroencephalographic activity ﬁom the infants’ ﬁ'ontal and
parietal lobes. The results showed greater right hemisphere activity to the ﬁ'owning face
and greater left hemisphere activity to the smiling face. Even though the infants’ overt
behaviors were not described, it seems reasonable to suppose that the smiling woman
elicited a positive state, and the frowning woman a negative state, so that the cortical
activation patterns may be assumed to reﬂect the infants’ experiential state while viewing
the ﬁlms. The results thus would be consistent with the Valence Hypothesis.

Harman and Ray (1977) used events from a subject’s past to elicit both positive
and negative emotions. Cortical EEG measures were used to study the eﬂ’ect of the
valence on each hemisphere. The results showed that the left hemisphere had a signiﬁcant
increase in power for positive emotions and a signiﬁcant decrease in power for negative
emotions. The right hemisphere changes were in the reverse direction but to a much lesser

extent. Again, these results are consistent with the Valence Hypothesis.

15

As a last example, Banich, Stolar, Heller, and Goldman (1992) induced depressed
or neutral mood-states in college students, as measured by self-rated sadness and
decreased happiness, using a combination of both music and imagery. On a task requiring
right hemisphere processing, subjects induced into a depressed mood showed longer
reaction times (i.e., poorer performance) than did subjects induced into a neutral mood.
The depressed subjects, however, were not slower on tasks requiring left hemisphere
processing. The authors interpreted this pattern of results as evidence that depression was
being regulated by the right hemisphere.

The current study used a new type of mood induction procedure. This procedure,
developed by Schiﬁ’ and Larnon (1989), uses facial manipulations to selectively stimulate
each hemisphere and, thereby, to induce diﬁ’erent mood-states. To explain how a facial
manipulation may be able to change mood, we ﬁrst must note certain features of the
anatomy of the facial sensory and motor pathways. After that, we can examine the Schiﬂ’
and Larnon (1989) study in detail.

The facial sensory and motor pathways in the lower two-thirds of the right side of
the face primarily enervate the left hemisphere of the brain. Similarly, the lower two-thirds
of the left side of the face primarily enervate the right hemisphere. The effects of
contracting the facial muscles on the lower two-thirds of me side of the face thus would
be expected to be on the contralateral hemisphere. Still, it is not clear where exactly the
stimulation is taking place (beyond the motor strip and somatosensory strip), or what
other cortical areas are being aﬂ’ected. Even so, following the Valence Hypothesis, one-
sided facial contractions would be expected to produce either a positive or a negative

mood, depending on the side of contraction.

16

The ﬁrst such experiment was performed by Schiﬁ‘ and Lamon (1989, exp. 2). In
this experiment, 30 college undergraduates were instructed to "pull back and lift each
corner of their mouth" (p. 925), manipulating one side of their face at a time. Side of
manipulation was alternated, with each manipulation lasting 45 seconds. The subjects
were then instructed to "attend to their feelings rather than to thoughts or facial sensory
experiences" (p. 929). Between contractions, there was a 30-second relaxation period
during which the subjects reported any feelings that they experienced. Each session was
audio taped and transcribed. The transcriptions were randomized and two judges read the
transcripts and then classiﬁed each mood report as positive or negative. Of a total of 120
mood reports from 30 subjects, two judges were in agreement about 106 reports. Fifty-
six of these agreements were that the report followed either a left- or right-facial
contraction. The other 50 reports on which the judges agreed were that the reports could
not be classiﬁable (e.g., “I felt nothing” or “I felt the same as before”). Since these reports
were not classiﬁable, Schiﬁ’ and Lamon did not include these data in their analyses. The
remaining 14 responses on which the two judges disagreed were ones for which one judge
believed the response was unclassiﬁable (BB. Schiﬂ', personal communication, September
12, 1996)’. Ultimately, only ambiguous responses were included in the analysis.
Results showed that the left facial contraction produced more positive remarks and the
right facial contraction produced more negative remarks, consistent with the Valence

Hypothesis.

 

2This information was available in Schiﬂ’ and Lamon’s (1989) original publication. The
way in which they handled the data was clariﬁed, however, through a personal
communication with Dr. Schiﬁ‘.

17

These data were based on unstructured self-report, which can be problematic for
statistical validity and reliability. Schiﬁ’ and Lamon (1989) therefore had another group of
36 college undergraduates perform the same facial contractions and then tell stories about
selected pictures ﬁom the Thematic Apperception Test (TAT). Three cards were used.
Card #11, depicting a mountain side with a prehistoric-looking creature sticking its head
and long neck out of the side of a cave and preparing to attack an animal resembling a
buﬁ‘alo, was chosen as the target card because it is known to oﬂ’er clues about a subject’s
mood (Bellak, 1986 cited in Schiﬁ’& Lamon, 1989). The other two cards, #13B,
depicting a young boy, barefoot and dressed in overalls, with an ambiguous look on his
face (e. g., perplexed, concentrating, thinking), sitting in the doorway of a wooden building
(perhaps a log cabin), and #14, depicting the silhouette of a man in a dark enclosure
looking out an open, and back-lighted window, “were chosen as practice pictures because
they were described as not being sensitive to emotional states” (Schiﬂ’& Lamon, 1989, p.
931). Again, the subjects’ responses were taped and transcribed. The responses were
classiﬁed into ”propositions” deﬁned as “an idea unit loosely consisting of a topic and a
comment about that topic” (p. 932). Two raters evaluated a total of 924 propositions as
either positive, negative, or neutral for emotional tone. The raters agreed on over 90% of
the propositions.

There were no diﬂ'erences in the tone of the subjects’ stories on the practice cards.
On the target card, however, responses diﬁ‘ered according to the side of facial
manipulation. The left face contraction produced more negative propositions than the
right face contraction and also produced stories that were distinctly negative in tone. The

right face contraction produced more positive and neutral propositions than the left face

18

contraction; however, the stories were not readily characterized as to emotional tone
(Schiﬁ’& Lamon, 1989, p. 933).

Based on these results, Schiﬁ‘ and Lamon (1994) suggested that "Unilateral
contraction of any muscles with comparable sensory motor organization [comparable to
the face] should be effective in inducing emotion" (p. 247). To test this idea, Schiﬁ‘ and
Lamon (1994) used hand contractions instead of facial contractions. Because the hand-
brain connections are ﬁrndamentally the same as the lower-face-brain connections (i.e.,
contralateral connections are stronger than unilateral connections), they reasoned that
right-and left-hand contractions should have effects in the same direction as right- and left-
face contractions.

In this study, 57 undergraduates were asked to squeeze a nrbber ball as hard as
they could for 45 seconds and then to relax for 10 to 15 seconds for a total of four trials.
Again, the three TAT cards were used, along with the same instructions for responding,
and responses were classiﬁed into positive, negative, and neutral propositions. There also
was a control group that gave stories to the TAT cards without performing the hand
contractions.

The results corroborated the facial contraction studies with one exception. As in
the facial contraction studies, subjects told signiﬁcantly more negative propositions about
the target picture after leﬁ-hand contractions than after right-hand contractions, and told
more positive propositions after right-hand contractions. The exception, for the
experimental groups, was in response to picture #133. Because #13B is generally
considered a non-emotionally sensitive picture, left-and right-hand diﬁ‘erences were not

expected. There were, however, more negative responses to this picture following left-

19

hand contractions than right-hand contractions. Schiﬂ' and Lamon (1994) were unable to
account for this ﬁnding but did note that the result at least was consistent with the results
for the target picture. Finally, the control group produced signiﬁcantly fewer negative
propositions about the target picture than the left-hand contraction group. Taken
together, these results are consistent with the Valence Hypothesis and serve to validate
Schiﬁ’ and Lamon’s’ mood induction technique.

Other studies support Schiﬁ’ and Lamon’s’ (1989, 1994) ﬁndings. For example,
Brockrneier and Ulrich ( 1993) used an assortment of standard mood induction techniques
proposed by Velton (1968) and Brewer and Doughtie (1980), including the life event
recollection technique. Brockrneier and Ulrich (1993) predicted that, in a semi-
standardized face-to-face interview, subjects would show greater right-sided lifting of the
corners of the mouth during positive mood states and greater left-sided lowering during
negative mood states. A video camera and microphone recorded the subjects’ responses
to the mood induction. Brockrneier and Ulrich (1993) examined those video segments
comprising a single frame of motion that was unrelated to speech or other gestures. Each
selected instance was at least one minute apart ﬁ'om the others so that it was clearly
distinct. The results were consistent with both predictions.

In summary, the results of these mood induction studies are consistent with the
Valence Hypothesis. That is, expressipn and experience of positive emotions are
regulated primarily by the left hemisphere, whereas expressien and W of negative
emotions are regulated primarily by the right hemisphere.

20

S-BI ID'ﬁI .1 1.5!. EE'
With this review of the literature on the Right Hemisphere Hypothesis and Valence

Model of emotion, we now can re-examine the literature to see how the results compare
when the sex of the subject is taken into account. This review must be somewhat limited
because many of the published studies either did not report analyses for possible sex
diﬁ’erences or included only male subjects (e.g., Schiﬁ’ & Lamon, 1989, 1994; Brockrneier
& Ulrich, 1993). As for the studies that did make this comparison, no clear, consistent
picture about the nature of or direction of the differences emerges. For instance, in some
studies laterality eﬁ‘ects appear only in males, in others only in females, and in still others,
in both males and females. Furthermore, where laterality eﬁ’ects do appear in both sexes,
sometimes they are stronger in men, sometimes in women. A review of these ﬁndings is
presented here. The review will be divided into two broad categories — ﬁrst, studies of
perception, then studies of expression and experience.

Beginning with studies of the peneeptipn of emotion, the ﬁrst question is, do men
and women alike show the right hemisphere advantage for the perception of emotional
stimuli, irrespective of valence?

Bryden, Free, Gagne, and Groﬁ’(l991), using a dichotic listening technique, had
subjects listen to four words presented in happy, sad, angry, or neutral tones’ of voice.
These four words were paired with nine other stimulus words that differed both in
emotional tone and phonetic content. For the emotional recognition task, subjects were
asked to listen for a speciﬁc emotional tone and to report whether it had or had not been

used in a dichotic pair presentation. The results indicated that, for subjects who showed

21

an ear-advantage, 80% of the men exhibited a leﬁ-ear advantage compared to only 62% of
the women, showing a sex diﬁ’erence on the emotion task.

McKeever and Dixon (1981) tested the abilities of men and women to recognize
tachistoscopically-projected pictures of previously memorized faces. The results showed
that women, but not men, had a left visual ﬁeld (right hemisphere) advantage when the
faces were paired with negative aﬂ’ective imagery (induced by having the subject imagine a
negative scene). However, when the faces were paired with neutral imagery, neither men
nor women showed signiﬁcant visual ﬁeld asymmetries.

Ladavas, Umilta, and Ricci-Bitti (1980) tested men’s and women’s ability to
discriminate emotional stimuli on three distinct levels: discriminating emotional faces ﬁ'om
neutral faces, regardless of the emotion being expressed, deciding whether the emotional
expressions were the same or diﬁ’erent, independent of the emotion being expressed, and
recognizing discrete, speciﬁc emotions. Six different human faces were shown in six
expressions: happiness, surprise, fear, sadness, anger, and disgust. The faces were
presented either to the subject's left or right visual ﬁeld, and the task was to identify the
face that displayed the same emotion as the target emotion. The subjects responded by
pressing a button when the target appeared while the researcher measured the subjects’
reaction time only to correct responses. Women had faster reaction times with left visual
ﬁeld presentation, indicating right hemisphere superiority for this discrimination task. Men
showed no consistent lateralization eﬁ’ects.

In a similar study Rizzolatti and Buchtel (1977) showed pictures of either positive
or negative emotional faces to men and women. Each picture was presented either to the

subjects’ left or right visual ﬁeld. Subjects were instructed to press a key if the emotion

22

was positive but not to press if it was negative. Again, only the reaction time for
accurately recognizing the positive emotion was recorded. This time, only the men
showed a signiﬁcant laterality eﬁ‘ect. They responded the fastest for faces displaying
positive emotions presented in the left visual ﬁeld (right hemisphere). Females showed no
signiﬁcant visual-ﬁeld differences for either emotion. The authors do not report the
responses or response times for the negative emotions presented to the right visual ﬁeld
(left hemisphere). They do state, however, that although there were very few errors,
between 10-15%, most males made fewer errors than females, and the male performance
for faces presented to the right hemisphere was the most accurate of all.

In summary, Bryden et a1. (1991) reported a non-signiﬁcant trend towards a male
leﬁ-ear (right hemisphere) advantage on an emotional perception dichotic listening task.
In Rizzolatti and Buchtel’s study ( 1977), only men showed a right hemisphere advantage
for the perception of emotional faces, whereas in McKeever and Dixon’s (1981) and
Ladavas et al’s. (1980) studies, only women did.

5 I' [E . I E . EE .

We turn now to the laterality studies that compare the sexes in the expression and
expenen' ee of emotion. In studies using electroencephalography (EEG), Davidson and
Schwartz (1976) and Davidson, Schwartz, Pugash, and Broomﬁeld (1976) measured EEG
activity while subjects generated both positively and negatively charged visual images.
For both positive and negative images, EEG activity was greater in the right hemisphere
than the left hemisphere for both males and females. Yet, female subjects’ right
hemisphere activity was greater than male subjects’ right hemisphere activity. These

results thus support the Right Hemisphere Hypothesis for both sexes but suggest that the

23

degree of the lateralized control may actually be stronger in women than in men.
Previously, the study by Moscovitch and Olds (1982) was discussed with regard to
their general ﬁndings. In addition to those general ﬁndings, they also looked at sex
differences, comparing side of facial activation in men and women while they were
engaged in spontaneous conversations. Again, any change from a neutral facial expression
was considered a valid data point, and both positive and negative expressions were
recorded (e.g., smiles, snickers, sneers, and ﬂows). As previously mentioned, the results
showed that, overall, both men and women had greater leﬁ-than right-side expressivity.
However, women had a signiﬁcantly greater left-side expressivity than did men. These
ﬁndings indicate that for both sexes, facial expression was mediated by the right
hemisphere but perhaps more selectively for the women, thus showing support for the
right-hemisphere hypothesis. Similar sex-related ﬁndings were reported by Borod, Koﬁ’,
and White (1983). In their study, both men and women showed greater expressivity on
the left side of the face, with the left side expression being more exaggerated for females
than for males. This was true, however, only for negative expressions. For positive
emotions, males showed signiﬁcantly greater expressivity on the left side, and females
showed a non-signiﬁcant asymmetry on the right side. Other studies have reported similar
sex diﬂ‘erences for the intensity and quality of asymmetries in facial expressions. These
studies thus suggest that both sexes show greater left-side expressivity but that the
diﬂ’erences are larger in women than in men (Campbell, 1978; Strauss & Kaplan, 1980).
Finally, in their study of mood changes associated with ictus, Sackeirn et al. (1982)
found that pathological crying and dysphoric mood responses occurred more commonly in

women than in men in association with right hemisphere site. When these changes

24

occurred in men, however, the response was more intense.

In summary, based on the mood induction and facial expressivity studies, the
research for the Valence Hypothesis seems to be as compelling as the dichotic listening
and facial expressivity studies, which supported the Right Hemisphere Hypothesis.
Furthermore, the overall results from expression and experience studies that 'meludedeex
djﬂerenees reveal many inconsistencies. Again, Davidson and Schwartz (1976) and
Davidson, Schwartz, Pugash, and Broomﬁeld (1976) found support for the Right
Hemisphere Hypothesis and for the notion that, where Meme is concerned, the degree
of the lateralized control may be stronger in women than in men. Moscovitch and Olds
(1982) and Borod, Koﬂ’, and White (1983) showed that for the expressien of positive
emotions, men showed signiﬁcantly greater left-side facial expressivity (right hemisphere)
and women showed a non-signiﬁcant asymmetry on the right side. Yet, in other studies of
expreseipn (Campbell, 1978; Strauss & Kaplan, 1980), the results suggest that both sexes
show greater left-side expressivity but that the differences are larger in women than in
men.

The data presented so far demonstrate that our current level of knowledge is at
least incomplete in the area of hemispheric lateralization for emotion, especially where
possible sex-related diﬁ’erences are concerned. The data ﬁ'om both hemispheric studies of
emotion and sex-related diﬂ'erences in emotion are diﬁcult to understand because the
results are inconsistent. An attempt is made here to explore a few of the possible
explanations for the inconsistencies. Following these explanations, a rationale is presented
for the current study. The rationale includes important clinical implications that could

result from such a study.

25

1M 5 I I . E' I' 9

The conﬂicting results in the literature may be partly attributable to the use of
diﬁ’erent experimental techniques and measurement methods by diﬁ‘erent researchers. The
abundance of methods used is partly the result of the difﬁculty researchers have faced in
deﬁning the feelings that presumably underlie different emotions. For instance, feeling
cocky, or smug, is often described diﬁ‘erently by diﬁ’erent people, so that where one
experimenter may consider an emotional response to be valid, another experimenter may
disagree. The conﬂicting results thus may stem, at least in part, from the use of diﬁ‘erent
labeling criteria in each experiment.

\Vrthout a consensus on the value of one experimental design over another,
researchers are encouraged to use diﬂ’erent methods; diﬁ’erent methods, however, may
produce results that diﬂ’er slightly but noticeably ﬁ'om one another. For example, recall
the previously mentioned list of studies that used a variety of techniques to induce mood-
states: playing music (Thaut & de l’Etoile, 1993; Bouhuys, Bloem, & Groothuis; 1995;
Lewis, Dember, Scheﬂt, & Radenhausen, 1995), self-generated imagery (Harman & Ray,
1977), the tachistoscope and Z-lens presentations of visual stimuli (Graves, Landis, &
Goodglass, 1981; Landis, Assal, & Perret, 1979; Rizzolatti, Umilta, & Berlucchi, 1971;
Geﬂ’en, Bradshaw, & Wallace, 1971; Dimond, Farrington, & Johnson, 1976), and the
presentation of words and other verbal stimuli (Banich, Stolar, Heller, & Goldman, 1992).
Although many of these techniques produced research ﬁndings that were signiﬁcant, and
some produced ﬁndings that supported other research, such methodological

inconsistencies still make it diﬁcult to draw reliable comparisons between studies. That

26

is, it is diﬁcult to know how the methodological differences presented above affect the
studies that do not generate signiﬁcant results or results that are consistent with previous
studies.

Since no consensus has been reached in support of either the Valence Hypothesis
or the Right Hemisphere Hypothesis, or for the strength or direction of laterality
differences between sexes, it is presumed that by replicating past studies one may either
validate or invalidate past ﬁndings. Schiﬂ’ and Lamon’s (1989) facial manipulation study is
an example of a technique that was well documented in the original report and that
produced signiﬁcant results but that has never been successﬁrlly replicated by researchers
other than Schiﬁ’ and Lamon (1994). The current study is the ﬁrst attempt at a replication.
As in Schiﬁ‘ and Lamon’s study (1989), the subjects were college undergraduates,
enrolled, like Schiﬂ’ and Lamon’s subjects, in an Introductory Psychology Course. Aside
from the difference in citizenship (American in the current study, Canadian in Schiﬂ’ and
Lamon’s study), the only sample diﬂ’erence was inclusion, in the current study, of an
adequate number of men and women to permit analysis of possible sex-related diﬁ’erences.

Probably the most important diﬁ’erence between Schiﬁ‘ and Lamon’s and the
current study was in the dependent measure. Instead of the TAT used by Schiﬁ‘ and
Lamon (1989), the current study used the Depression Adjective Check List. The DACL is
considered a well-validated and reliable way to measure emotional responses. This is
demonstrated by the number of studies that use the DACL as a comparison test or a “gold
standar ” for validating new instruments (Lubin, 1993a, 1994a; Lubin, Fielder, & Van
Whitlocle 1996; Lubin, Van Whitlock, Dale, & Kelly, 1996; Lubin & Van Whitlock,

1995). The DACL's well-established history and strong statistical support as a valid and

27

reliable measure should give conﬁdence in validity and reliability of the results.

This study is also important because it tackles a ﬁrndamental question pertaining to
sex differences. That is, almost no one would argue that differences between men and
women do not exist. Many, however, would argue that differences are primarily a product
of the environment, as represented by the three diﬁ‘erent theories reviewed earlier, while
others would argue that the physiological/biological inﬂuences are more important.
Unfortunately, current methods of studying hemispheric lateralization of emotion, such as
showing movies through a Z-lens or listening to music, do not exclude the unwanted
inﬂuence of the environment. After all, society is not immune to promotions pushing the
notion that “men like action movies and women like love-dramas.” This type of inﬂuence
could have an unwanted eﬂ‘ect on the subject watching a movie that is intended to induce
an emotional response. The facial manipulation technique, however, would not seem to be
contaminated in this way. That is, it would be hard to make the case that society has
inﬂuenced peoples’ willingness or desire to manipulate one side of one’s face. Moreover,
no equipment such as Z-lenses or tachistoscope are needed, making facial manipulations a
readily available technique for the ﬁxture.

CI' . I I I' .

In addition to the important scientiﬁc questions about possible sex-related
diﬁ‘erences in cerebral lateralization for emotion, there are equally important clinical
implications to this research. For example, if the results provide a clearer understanding of
the human “emotional system,” they could improve the clinician’s ability to serve clients.
This is particularly important in the therapeutic climate that exists today; therapists are

being pressured to develop techniques that are more time-eﬁcient. Research using facial

28

manipulations that stimulate emotional responses could oﬂ‘er one more way for therapists
to understand their clients’ emotions and to implement time efﬁcient techniques. For
instance, if males were found to be more clearly right hemisphere dominant than females
for the expression and experience of negative emotions, and therapists become aware that
imagery is a process that maybe largely controlled by the right hemisphere (currently the
evidence explaining hemispheric control of mental imagery is mixed; Farah, 1984; Farah,
Gonzzaniga, Holtzrnan, & Kosslyn, 1985; Kosslyn, 1987; Sergent, 1990), a therapist who
currently does not use imagery may be encouraged to, speciﬁcally for men. Such
techniques might help facilitate access to negative emotions.

It is at least possible that one reason why more women than men seek therapy is
that therapists have often used a procedure (i.e., talk therapy) better suited to the female
than male processing style. Although this supposition may not have documented support,
there is precedence for similar ﬁndings in the literature. That is, recent research suggests
that other groups who have not been large users of therapeutic services (i.e., African-
Americans, Asians, Hispanics) have felt that the therapy process conﬂicted with their
styles, beliefs, or backgrounds (Sue, 1987, 1994; Takeuchi, Sue, Yeh, 1995). As more
information is gathered showing how these groups can be served more effectively, it is
predicted that they will seek therapeutic services more readily.

From another perspective, it has been argued that, although psychopharmacology
has made advances in developing antianxiety, antidepressant, and antipsychotic
medications since the 19508, and the availability of these medications to individuals
suﬁ‘ering ﬁom distress has greatly increased in recent years, the sites of action for these

medications are only recently becoming understood. For instance, Haldol, a commonly

29

used antipsychotic agent, is known to antagonize dopamine receptors. It is also known,
however, that dopamine receptors are found in many areas of the brain including the
limbic system, ﬁ'ontal cortical areas, the hypothalamic-pituitary axis, as well as certain
centers of the brainstem (Julien, 1988). Although there are theories that speciﬁc receptor
areas have primary control over certain behaviors, if males and females diﬂ’er in cerebral
organization for emotions, medications that affect all of those areas could actually act at
sites diﬁ’erent for men and women. So, when a non-medical therapist consults a
psychiatrist about using medication with a client, the client’s sex could be an important
factor to consider. Furthermore, if the American Psychological Association’s currently
unresolved decision about prescription privileges for therapists were to be resolved by
granting such privileges, the possibility of individual diﬂ’erences in cerebral organization of

emotion could become an important and helpﬁrl knowledge base for therapists.

CHAPTER 2
CURRENT EXPERIMENT
The current study used Schiﬂ‘ and Lamon’s (1989) mood induction technique in an
attempt to elucidate potential interactions between sex and hemispheric lateralization for

emotion.

E . E I' .
Since both the Right Hemisphere Hypothesis and the Valence Hypothesis are

supported by a substantial amount of research, and since there are discrepant results where
sex and hemispheric lateralization for emotion are concerned, making predictions with
conviction is diﬁcult. Nevertheless, three primary predictions are made along with three
secondary predictions. The rationales for these predictions stem from research in diﬂ’erent
but related areas.

WWW: First, in choosing between
the Right Hemisphere and the Valence Hypotheses, it was predicted that the Valence
Hypothesis would be supported. The main reason was that the current study was
designed to be a partial replication of Schiﬂ’ and Lamon’s study, the results of which did
support the Valence Hypothesis. In the current study, therefore, it was predicted that
subjects will make more negative responses (e. g., report feeling low, somber) following

manipulations of the leﬁ side of their face (right hemisphere) and more positive-like (e.g.,

3O

31

report feeling smug-like, cocky) responses following manipulations of the right side (left
hemisphere) of their face.

W. Given the modest but statistically signiﬁcant evidence for
sex-related differences for lateralization of cognitive tasks, a second prediction was that
men would be more clearly lateralized than women on the mood induction task. Thus,
following left-face manipulations (right hemisphere), it was predicted that men would
produce greater change toward more negative proﬁles resulting in higher scores on the
DACL’s (i.e., higher depression scores) than women, and following right-face
manipulations (left hemisphere) men would produce greater change toward more positive
proﬁles resulting in lower depression scores than women. Also, when the last
manipulation was a left-face contraction, it was predicted that the POMS-BI, the
additional measure used to rate emotional responses, would show men producing more
intense negative responses resulting in higher scores on negative items than women.
When the last manipulation was a right-face contraction, this measure would show men
producing more intense positive responses resulting in higher scores on positive items than
women.

Sexﬂelatedﬁesppnseﬂanems. Third, as stated above, the majority of research
studies comparing questionnaire response patterns between men and women suggest that
women report more emotion than men. Based on the results ﬁ'om these studies it was
predicted that women would report more overall emotion than men. In other words,
regardless of having performed either a leﬁ- or right-face manipulation, it was predicted

that women would select more DACL adjectives, both positive and negative, than men.

32

S I E I' .

IimingefManipuleﬁenmm. Following Schiﬁ’ and Larnon’s supposition
(1989; see also Schiﬂ”& Lamon, 1993) that a subject notices a diﬂ’erence between the
effects of two manipulations more than the eﬂ‘ects of manipulating one side alone, it was
predicted that the eﬂ’ect of the manipulations would be more apparent alter a subject
manipulated both sides of the face than after only the ﬁrst manipulation. On the DACL,
this means that scores following the second facial manipulation would be more
exaggerated, suggesting increasing aﬂ‘ective experience, than scores following the ﬁrst
manipulation.

WWW. Likewise, the eﬂ’ects of the third and fourth
manipulations would produce an increase in aﬁ’ective experience beyond that of the ﬁrst
and second manipulation resulting in higher scores following the last two manipulations
compared to the ﬁrst two manipulations.

WWW. Finally, on the hypothesis that men are
more clearly lateralized than women, it was predicted that DACL scores following the last

two manipulations ought to be more exaggerated for men than for women.

CHAPTER 3
METHOD

Subjests

Subjects were 40 right-handed male and 40 right-handed female undergraduate
students ﬁ'om the Psychology Department Human Subjects Pool at Michigan State
University. Their age ranged between 18 and 27 years old. This population was selected
because of its availability and because of the desire to select a healthy population.
Subjects received credit in their undergraduate psychology course for their participation.
Anyone with a history of depression, other psychiatric disorders, and/or facial paralysis
was excluded for the following reasons: 1) it would be unethical to induce potentially
dysphoric emotions in someone with a history of depression or other psychiatric disorders,
and 2) an individual with facial paralysis may be unable to perform the facial manipulation
as required by the experimental protocol. To eliminate these individuals from the study,
criteria for exclusion were clearly stated on the subject sign-up sheets, on the informed
consent form, and on the medical history questionnaire.
Design

This study used a 2 x 2 x 2 within-subjects design with side of the facial
contraction (leﬁ vs. right) and sex of subject (male vs. female) and condition (RLRL vs.
LRLR) as the independent variables. The side that was contracted ﬁrst was
counterbalanced across subjects. Halfof the subjects, therefore, began with a right-sided
contraction (RLRL) and the other halfbegan with a left-sided contraction (LRLR). The

33

34

dependent variables were the subject’s emotional responses after facial contractions as
measured by Lubin’s (1994b) Depression Adjective Check List (DACL) and by Lorr and
McNair’s (1988) Proﬁle of Mood States-Bipolar (POMS-BI). Each subject performed
two left-sided and two right-sided face contractions, alternating the side of the contraction
each time.
Materials

The materials used included an informed consent form, a subject identiﬁcation
form, a medical history questionnaire, a participant background and demographic survey,
an 8-item handedness questionnaire, the DACL, and the POMS-BI.

InfennedﬂpnsenLEan. The Informed Consent Form (Appendix A) includes the
following information: 1) that this is a study of the eﬁ’ects of facial muscle contractions on
mood in a normal, healthy population, 2) that the subject will be asked to complete a
medical history questionnaire, a background and demographic survey, an 8-item
handedness questionnaire, and two adjective checklists, 3) that the subject will be asked to
pull back and lift one corner of his or her mouth and to hold that position for 45 seconds,
4) that the subject will perform four such contractions, alternating sides for each
contraction, and 5) that the subject will be asked to complete a checklist, as quickly as

possible, following each contraction.

 

3Schiﬁ’ and Lamon (1993) found that having subjects alternate side of facial contraction
after each trial (as opposed to performing several sequential same-side contractions)
produced more robust and reliable mood states.

35
WW. On the Subject Identiﬁcation Form (Appendix B), the

subject’s name and identiﬁcation number was recorded for the purpose of maintaining
conﬁdentiality. This was the only form that identiﬁes and links subject name to
identiﬁcation number. Only the investigator and his designates had access to this
information. All other documents had only the subject’s identiﬁcation number on them.
W. The Medical History Questionnaire
(Appendix C) asked whether the subject currently or at any time in the past had been
clinically diagnosed with any of the following disorders: 1) Mq'or Depression, 2) Bipolar,
3) Generalized Anxiety, 4) Panic, 5) Post Traumatic Stress, 6) Substance Abuse, 7)
Schizophrenia, 8) Hypertension, 9) any other emotionally related disorders, or 10) a
history of facial paralysis. This measure was created to screen out individuals who would
be inappropriate for psychological or physical reasons.

' . The Personal Information

 

and Background Questionnaire (Appendix D) asked for the subject’s age, sex, years of
education, academic major, and ethnicity/race. The reported age and sex of each subject
was used to score responses on both the POMS-BI and the DACLs; both measures use
age and sex of subject as scoring variables. Years of education, academic major, and
ethnicity/race were obtained in order to provide the experimenter with a more complete
understanding of each subject’s background. These variables were not assumed to be
important to the hypotheses and are considered only if they appear to relate to diﬁ’erences

between the groups.

36

LateraligLQuestiennaiLe. Only right-handed subjects were used in this sample.

Limiting the sample to right-handers is methodologically important since research has
suggested that left-handers have diﬂ’erent cerebral organization from right handers
(Bulman-Fleming & Bryden, 1994; Levy & Reid, 1978; Luh, Redl, & Levy, 1994). In
order to verify subject handedness, a Laterality Questionnaire (Appendix E) was used.

The questionnaire asked the subject to choose which of the following descriptions best
describes the subject’s general handedness: 1) strongly left-handed, 2) moderately left-
handed, 3) ambidextrous (either-handed), 4) moderately right-handed, or 5) strongly right-
handed. It then asked the subject to indicate his/her hand preference for eight common
tasks: 1) write a letter, 2) hammer a nail, 3) throw a ball at a target, 4) unscrew lid of a jar,
5) use a knife to cut bread, 6) use a tooth brush, 7) hold a match while striking it, and 8)
hold a tennis racket. Strength of hand preference was rated on a ﬁve-point Likert Scale,
ﬁom 1 (always left) to 5 (always right). Subjects also had to indicate whether they write
with an inverted or non-inverted hand posture by circling one of two pictures depicting the
two writing postures‘. Lastly, subjects were asked whether anyone (including themselves)
has ever tried to change what hand they use for certain tasks. This question was included
to check for the possibility that a person who was once left-handed may have changed
hand-use for certain tasks, either because of instruction from a teacher, parent, or
grandparent, or because of certain life experiences, such as diﬁculties in using tools

designed for right-hand use. Many leﬁ-handers report such experiences and, therefore,

 

‘The reason for using this measure is that research ﬁndings have shown that there may
be a diﬁ‘erence in the cerebral organization between leﬁ-handers who write with an
inverted posture and leﬁ-handers who do not (Levy, 1982; Levy & Reid, 1976; Levy &
Reid, 197 8).

37

come to use their right hands for certain tasks (Harris, 1990).

W. The Depression Adjective Check List (DACL)
(Appendix F) is a multiple—item paper- and pencil-checklist measuring either state or trait
transient moods, feelings, and emotions (in this case only state moods, feelings, and
emotions were considered) (Lubin, 1993a). Seven parallel forms exist, making the DACL
useﬁrl in studies where these factors are measured repeatedly. In the current study,
subjects’ emotional experience was measured ﬁve times using ﬁve of the seven highly
correlated (Table 2) state-DACL forms (i.e., Forms A, B, C, D, G). Each form includes
adjectives that represent state negative mood sub-scale (S-Neg) and state positive mood
sub-scale (S-Pos). As described in the manual, the S-Neg sub-scale depicts the degree of
state negative mood and is calculated by adding the negative valence adjectives endorsed.
The S-Pos sub-scale depicts the degree of state positive or elated mood and is calculated
by adding the positive valence adjectives endorsed. Using these sub-scales, a Total score
is calculated. The Total scores is the total number of adjectives, both positive and
negative valence, endorsed by a subject. Although elevated S-Pos scores suggest higher
levels of state positive mood, S-Pos is used diﬁ’erently to calculate the Total raw score.
That is, “Because one component of depression is conceptualized as the absences of
positive mood, the Total raw score is calculated as the sum of the negative adjectives
endorsed and the number of positive adjectives that were not endorsed (i.e., [10 minus S-
Pos])” (Lubin, 1993b, p. 7).

Finally, Total scores were converted to standardized scores (both t-scores and
percentile scores) according to normative data separating subjects’ by age and sex (Lubin,

1993b). In the following analyses, unless otherwise speciﬁed, all statistics are generated

38
from percentage scores.

W. The POMS-BI (Appendix G) is a 72-item paper
and pencil test that reportedly measures the following mood dimensions in terms of six
bipolar affective states: 1) composed-anxious, 2) elated—depressed, 3) agreeable-hostile,
4) energetic-tired, 5) clear-headed-confused, and 6) conﬁdent-unsure. Table 3 shows the
reported correlations between mood states reportedly measured by the POMS-BI (Lorr &
McNair, 1988). Each subject rates the degree to which he/she subscribes to 72 adjectives
on a four-point Likert Scale ranging from zero (“much unlike this”) to three (“much like
this”). The form is designed to rate the subject’s feelings during the past few minutes.
Adjectives include untroubled, relaxed, tense, uneasy, kindly, sympathetic, grouchy,
annoyed, and vigorous.

The formula that is recommended for scoring the POMS-BI is as follows: the total
score (8,) is the sum of the positive item scores (8,) minus the sum of the negative item
scores (8,) plus 18 (S, = Sp - SN + 18)’. The constant 18 is used to make all scores
positive, making each scale’s range from 0 to 36. Using this scoring procedure, a low
total score represents a less negative proﬁle and a high total score represents a more
negative proﬁle.

This POMS-BI scoring system can be used to gauge the intensity of the subjects’

emotional responses. For this reason, the POMS-BI was administered as a pretest,

 

’Initially, the procedure reported here was the intended scoring procedure for this
study. Since the six-factor model put forth by Lorr and McNair (1988) was not found in
this study, however, the scoring procedure was necessarily changed. The new procedure
included summing the Likert-scale responses for selected adjectives on the positive and
negative factors and then comparing the means of those responses.

39

measuring the subjects’ emotional state at baseline (i.e., prior to performing an
experimental manipulation) and as a post-test (i.e., following the last experimental
manipulation). The post-test measure was used to assess the intensity of the last facial
manipulation; comparing leﬁ- and right-face contractions. In contrast, the DACLs were
administered between each experimental manipulation, as well as before the ﬁrst
manipulation and alter the last manipulation. The DACLs, therefore, measure not only the
emotional state at baseline but also the eﬁ’ects of individual facial contractions.

To estimate the degree of POMS-BI test-retest reliability, Lorr and McNair (1988)
had a sample of 66 subjects complete the form a second time (see Table 4). As Table 4
shows, no signiﬁcant diﬁ’erences in mean scores were found.
Premium

Prior to the start of the experiment, 40 subject packets were numbered 001
through 040. These served as subject numbers and were used for female subjects only.
All even-numbered packets were given to subjects in the LRLR condition, all odd-
numbered packets to subjects in the RLRL condition. Additionally 40 packets, numbered
041 through 080, were used for male subjects. As before, all even-numbered packets
were given to subjects in the LRLR condition, all odd-numbered packets to subjects in the
RLRL condition. Using this system, the ﬁrst female subject to tested was assigned subject
number 001 and received the RLRL condition, the second female subjected was assigned
number 002 and received the LRLR condition, and so on.

When the subjects entered the testing room, they were seated at a small table
across from the experimenter and were given an Informed Consent Form to read and sign.

This form was also read aloud by the experimenter. Following this, the subject who

40

agreed with the conditions and signed the consent form completed a Medical History
Questionnaire. Anyone who answered “yes” to any of the disorders listed on the
questionnaire, was told that, for the purposes of this study, no one could be included who
has suffered from any of the listed disorders (as previously mentioned, these criteria for
exclusion were stated on the sign-up sheet). To understand the characteristics of persons
who were excluded, information ﬁ'om the remaining pre-test measures was collected. If
the disorders reported included histories of emotional distress (for example, depression,
bipolar, generalized anxiety, or substance abuse), the subject was told that counseling
services are available at the University Counseling Center. After completing the pre-test
measures, these persons were excused and reminded that they still received ﬁrll credit for
their participation. Following this, all subjects completed the remaining three pre-test
measures: 1) the Laterality Questionnaire, 2) the POMS-BI, and 3) the DACL.

Next, the experimenter, while facing the subject, explained how the subject was to
perform the facial manipulations. The following instructions were read:

I want you to pay close attention to the following instmctions.

Don’t do anything yet; just listen. I want you to pull back and

raise the left/right side of your mouth as much as you can, just like this

[experimenter demonstrates by making the mirror-image movement].

Now you try it. [After the subject successﬁrlly demonstrates the

movement]. That’s ﬁne. When the experiment starts, I want you to

hold it in that position for 45 seconds until I say relax.

41

If necessary, ﬁrrther instructions were given until the subject understood how to perform
the facial manipulation. Subjects were asked whether they understood the procedure.
Once they understood what they were to do, the following instructions were read:

Now when I say “start,” I want you to pull back and raise

the left/right side of your mouth in the same way you just did

and to hold that position until I say “relax.” Remember to

pull back and raise your mouth as much as you can. While

you’re pulling back and raising your mouth, I want you to

breathe normally and to look directly at me. I also want

you to pay close attention to your mood, to how you are feeling.

The side of initial contraction depended on the predetermined order to ensure
equal cells. After maintaining the contracted position for 45 seconds, the subjects were
told to relax and to complete one version of the DACL. Following that, subjects were
told to perform the same contraction on the opposite side of the face after which they
were asked to complete another version of the DACL. This procedure continued until
each subject had performed a total of four contractions and had completed a total of four
versions of the DACL. After completing the fourth facial contraction and the ﬁfth DACL,
the subject completed another POMS-BI checklist. Subjects were then told that their
participation in the study was complete. A debrieﬁng period followed.

For the debrieﬁng period the subjects were asked what they believed to be the
nature of the study. The experimenter then explained its actual purpose. The subjects

were asked whether they were feeling any emotional residue (either positive or negative)

42

from their participation. When subjects reported feeling distressed or if the researcher
detected stress (for example, if the subject was tearﬁrl or upset), the researcher (an
advanced clinical psychology graduate student) continued to talk with the subject until
both agreed that the subject’s emotional state had stabilized. Through pilot testing, it had
been demonstrated that the entire procedure took between 40 and 50 minutes to complete.
Statistics

All statistical procedures were performed using Statistical Package for Social
Science (SPSS) and only ﬁndings resulting in p >.05 were considered signiﬁcant.
Frequencies and descriptive statistics were calculated for the age, year in school, and
ethnicity/race for all the subjects. Since no assumptions were made regarding the subjects’
age, year in school, or ethnicity/race, statistical analyses were not performed on these data.

The data simply provide information about the subjects who participated.

CHAPTER 4
RESULTS

Two women and two men were excused from participating in the experiment due
to positive responses on the Medical History Questionnaire. One woman reported having
suﬂ‘ered a concussion following a sporting accident and said that she still felt the eﬂ’ects,
and the second women has a history of Obsessive-Compulsive Disorder. The two men
reported histories of Obsessive-Compulsive Disorder and Depression, respectively. In all
four instances, subjects received ﬁrll credit and were excused from ﬁrrther participation.
These data suggest that the Medical History questionnaire effectively screened out
subjects who were inappropriate for this study. The four excused subjects were replaced,
permitting the originally intended sample size to be reached.

WW

Factor analysis of the Proﬁle of Mood States-Bi-Polar was performed using all
time-one responses from all the subjects along with left handed male subjects ﬁom the
related study, for a total N = 120. Since Lorr and McNair (1988), the developers of the
POMS-BI, claim that the measure contains six bi-polar scales, a Principal Components
Factor Analysis with orthogonal rotation was performed to test this claim“. The data did

not support a six bi-polar scale factor structure. For example, results showed that 27% of

 

‘Aﬁer the Principal Component Factor Analysis, all statistics were calculated on only
right-handed male and female subjects.

43

44

the variance was accounted for by the ﬁrst variable, which had an eigenvalue of 19.43,
whereas the second variable accounted for only 8.8 % of the variance, with an eigenvalue
of 6.31. As shown in Table 5, the remaining variables with eigenvalues greater than 1.0
accounted for only a small percentage of the variance. With so many factors, and with
substantial item overlap, the six-factor solution would be diﬁcult to interpret with any
conﬁdence.

Since the six-factor structure proposed by Len and McNair (1988) was not
supported, Principal Component Analyses were performed forcing all 72 items into a ﬁve-,
four-, three, two-, and one-factor model. With each model, all items with eigenvalues
over 1.0 were found to have identical loadings, to account for the same amount of
variance as the original six-factor model (see Table 5), and to have substantial item
overlap. These factor solutions, therefore, were also un—interpretable.

To generate a reliable bi-polar scale, items with eigenvalues over 1.0 were selected
from the two-factor model. Labels for variables were attached to item numbers so that an
adjective list could be generated based on content. Items that did not cluster into two
factors based on content (e.g., one positive and one negative) were deleted ﬁ'om the
potential list of variables. A Principal Component Factor Analysis forcing a two-factor
structure, as well as a reliability analysis, were computed based on the remaining variables.
The resulting two factors include a “positive” scale with 12 items and a “negative” scale
with 15 items. Table 6 shows the adjectives selected and the inter-item correlations.
Table 7 shows factor variables, factor loadings, the reliability coeﬁcient for both factors,
and the alpha variance if any of the items were to be deleted. The “alpha variance if item

delet ” scores support the inclusion of each item in the factor. The content of the two

45

factors suggest a bi-polar scale and the inter-item correlations support the distinction
between the factors.
I1 I . I E .

These data pertain to demographic information about subjects in this study and do

not pertain to the hypotheses or predictions.

Signiﬁcant differences exist between male and female subjects’ age, “mm, =
20.02, SD = 1.95; 11.95,“, = 19.02, SD = 1.64; 1(78) = -2.48, p = .015, and year in

school, winded”: 2.08, SD = 1.05; umhwpm = 1.53, SD = .72; {(78) = -2.74,
p = .008. Signiﬁcant differences did not exist, however, between male and female
subjects’ ethnicity/race, 108) = .73 7, p = .833. Figure 1 shows frequency data for 40 male
and 40 female subjects’ age, year in school, and ethnicity/race.
W

The results of 80 subjects' responses to the 8-item handedness questionnaire
showed that all subjects were strongly right-handed (see Figure 2). Subjects self-reported
their hand preference for completing eight speciﬁed tasks. A total of 640 responses were
collected. Of those responses, 569 fell into the “right hand only and usually right hand”
category, 65 in the “both hands” category, and 6 in the “usually left and left only”

categories.

E l [E . E I' .
Valeneeﬂypmhesis. Following the Valence Hypothesis, the ﬁrst primary

prediction supposed that left-face manipulations would produce a more negative affective

experience and right-face manipulations would produce a more positive affective

46

experience as measured by DACL scores, where the higher the DACL score, the higher
the negative aﬁ‘ect. Two steps were required for testing the Valence Hypothesis. The
ﬁrst step was to show that there is a signiﬁcant interaction between the between-subject
variable Condition (RLRL, LRLR) X the within-subject variable Time (DACL B, C, D,
G). Ifthe interaction was signiﬁcant, one could say only that some effect took place
between the Condition and Time variables. This information, however, would not explain
whether these ﬁndings jibe with the Valence Hypothesis. So, given a signiﬁcant
interaction, the second step was to compare the average of all DACL scores following
left-face manipulations with the average of all DACL scores following right-face
manipulations. Ifthe average following left-face manipulations proved to be signiﬁcantly
higher than the average following right-face manipulations, then the Valence Hypothesis
would be supported.

To perform the ﬁrst step in the analysis, the test for the interaction, a repeated
measureeANCQXA was computed for all 80 subjects, with the pre-measure DACL-A as
a covariate 1(1, 77) = 4.24, p S 0.00.

The results ﬁ'om the AN COVA analysis indicated that the Condition X Time
interaction was signiﬁcant E(3, 234) = 2.96, p = .03, and that Time had a large eﬁ’ect (see
Table 8 and Figure 3). Therefore, the next step was to compare the average of all DACL

scores following left-face manipulations (um = 105.23, SD = 13.78; see Figure 4) with

the average of all DACL scores following right-face manipulations (“W = 105.15, SD

= 13.17; see Figure 4). For this comparison, a We: was used. The

results showed that the diﬂ’erence was not signiﬁcant t(79) = 0.06, p = 0.95; therefore, the

47

Valence Hypothesis was not supported.

W. The second prediction supposed that men would be
more clearly lateralized than women on the mood induction task. Thus, following left-face
manipulations it was predicted that men would produce greater change toward more
negative proﬁles resulting in higher depression scores on the DACLs than women, and
following right-face manipulations men would produce change toward more positive
proﬁles resulting in lower depression scores than women.

To examine whether sex of subject had an effect on DACL scores (DACLs B, C,
D, G), the same WA presented above was used, this time
including Sex (male, female) as an additional between-subject variable. To repeat, a 2 X 2
X 4 (RLRL, LRLR and Male, Female and DACL B, C, D, G) Lepearedmeamm
ANCQXA was computed for 40 male and 40 female subjects, with the pre-measure
DACL-A as a covariate 1(1, 77) = 4.24, p S 0.00.

As shown in Figure 5, the Condition (RLRL, LRLR) X Sex (male, female)
interaction was not signiﬁcant E(3, 228) = 0.89, p = 0.45. Table 9 contains the means and
standard deviations for male and female subjects in both conditions.

W. This prediction supposed that when the last facial manipulation was a
leﬁ-face contraction, men would produce more intense negative responses on the second
administration of the POMS-BI, resulting in higher scores on negative items, than
women’. Likewise, when the last manipulation was a right-face contraction men would

produce more intense positive responses, resulting in higher scores on positive items, than

 

’Where the POMS-BI data is concerned, all data was analyzed using the two factor
model described in the POMS-BI factor analysis portion of this report.

48

women. To examine these predictions, two simplefaetenﬂANQVAs, each co-varying

pre-measure POMS-BI at time 1, were computed for the between-subject variables
Condition (RLRL, LRLR) X Sex (male, female) and for the within-subject variables
Negative or Positive Adjectives from the POMS-BI Factor Analysis‘. Neither the negative
adjective analysis E(1, 78) = 0.12, p = 0.73 nor the positive adjective analysis E(l, 78) =
0.09, p = 0.77 produced signiﬁcant results (see Table 10).

W. For this hypothesis it was predicted that,
regardless of having performed either a left— or right-face manipulation, women would
select more DACL adjectives, both positive and negative, than men. To examine this
prediction a 2 X 4 (male, female and DACL B, C, D, G) W with
Sex as a between-subject variable and DACLs as a within-subject variable, the mean
number of adjectives endorsed by women on each DACL (DACL A, B, C, D, G) was
compared to the mean number of adjectives endorsed by men on each DACL (DACL A,
B, C, D, G).

As shown in Table 11, there were no signiﬁcant diﬁ’erences between women and
men at any time for any of the mean DACL scores.
B I E S I E I .

IIfumn' ' 3 pi Manipulatipn (pm 1). Next, it was predicted that the eﬁ‘ect of the
manipulations would be more apparent after subjects manipulated both sides of their face

than after only the ﬁrst manipulation. On the DACL, this means that scores following the

 

'One subject had missing data for the second administration of the POMS-BI. That
missing data was replaced with means from subjects of the same sex and same condition
allowing for an N=80.

49

second facial manipulation would show an increased affective experience compared to
scores following the ﬁrst manipulation. To test this hypothesis, the absolute amount of
change from the ﬁrst manipulation to the second manipulation was computed for 20 male
and 20 female subjects in the RLRL condition and 20 male and 20 female subjects in the
LRLR condition (see Table 12). Said another way, DACL-B scores were subtracted from
DACL-A scores producing a “DACL-AB difference score”. Similarly, DACL-D scores
were subtracted ﬁ'om DACL-C scores, producing a “DACL-CD difference score”. These
results were computed using a W and co-varying pre-measure
DACL-A. Once again, Sex and Condition (male, female and RLRL, LRLR) are between-
subject variables and Time (DACL B, C, D, G) is a within-subject variable. Since the
direction of change is not considered here, the absolute values of all diﬁ‘erence scores were
used in the W

The results were not signiﬁcant E(1, 78) 1.72, p = 0.19; the eﬂ’ect of the second
facial manipulation was not signiﬁcantly diﬂ’erent ﬁ'om the eﬁ’ect of the ﬁrst facial
manipulation.

IiminngManipulatierLLparLZ). In an extended version of the same hypothesis, it
was predicted that the effects of the third and fourth manipulations would produce an
increase in affective experience beyond that of the ﬁrst and second manipulation. One

would expect, therefore, greater absolute change following the last two manipulations

 

’Two necessary conditions existed in this data that limit the risks associated with the
use of “diﬂ’erences scores”. First, there were no data points in danger of creating either a
ceiling or a ﬂoor eﬁ’ect, and second, a co-variate (DACL-A) was available to support the
reliability of the results (for ﬁrrther explanation Collins, 1996; Wilkinson, Blank, Gruber,
1996).

50

compared to the ﬁrst two manipulations. To test this prediction, the sum of absolute
change scores for DACLs-AB and DACLs-BC was computed for 40 subjects in the RLRL
condition. Next, the sum of the absolute change scores for DACLs-CD and DACLs-DG
was computed for 40 subjects in the LRLR condition. Finally, a Lepeatedmeesum
ANCOXA, co-varying the effect of pre-measure DACL-A, was computed to examine
possible differences between the ﬁrst- and second-set of manipulations. Here, Time
(DACL absolute change scores) were within-subject variables and Condition (RLRL,
LRLR) were between-subject variables.

The results were not signiﬁcant E(1, 78) 0.73, p = 0.40; compared to the ﬁrst two
manipulations, the last two manipulations did not produce an increase in emotional
experience.

WWW. Finally, on the hypothesis that men
are more clearly lateralized than women, it was predicted that DACL scores following the
last two manipulations ought to be more exaggerated for men than for women. In light of
the results presented above, it is clear that no diﬂ’erences would exist between men and

women for this same hypothesis.

CHAPTER 5
DISCUSSION

Three primary predictions and three secondary predictions were investigated in this
study. None of the predictions were supported. Although not a primary prediction, one
interaction was statistically signiﬁcant. The discussion that follows explores both the
signiﬁcant interaction as well as the non-signiﬁcant ﬁndings. Also covered are discussions
about possible reasons for discrepancies between the current experiment and past
experiments, discussions about possible clinical implications of this research, and ideas for
future research in this area.

The interaction that was signiﬁcant was Condition (RLRL, LRLR) X Time (DACL
B, C, D, G) whereby a dramatic increase in negative aﬁ’ect across time was seen. As
shown in Table 8 and Figure 3, subjects’ responses on the DACLs steadily increased
creating a signiﬁcant change over time (DACL B, C, D, G). Still, it is unclear why
increases were found only in negative aﬂ’ective experiences. Although this is purely
speculative, one possibility is that subjects were uncomfortable with the facial
manipulation procedure. If they felt awkward about manipulating their faces, they may
have been responding to that awkwardness and not to the effect of the manipulation itself.

XaleneeHypethesis. The ﬁrst prediction supposed that, where the experience of
emotion is concerned, subjects’ responses to facial manipulations would support the
Valence Hyporhesis. The results presented above, however, did not support the Valence

51

52

Hypothesis. That is, contrary to Schiﬂ’ and Lamon’s (1989, 1993, 1994) results, subjects
did not report more negative emotional experience following left-face manipulations (i.e.,
right hemisphere) or more positive emotional experiences following right-face
manipulations (i.e., left hemisphere). Although there may be no deﬁnitive ways to explain
these diﬂ’erences, several possible explanations can be suggested.

One possible explanation for different results follows from the use of different
criteria to include and exclude data ﬁ'om ﬁnal analyses. The current study analyzed all
data that were collected and found results that are less clear but perhaps more accurate
than Schiﬁ‘ and Lamon’s (1989). In contrast, as previously noted (see page 16), Schiﬁ’
and Lamon (1989 exp. 2) included only unambiguous responses in their analyses. Among
the data that they excluded were subjects’ responses that could not be agreed upon by
raters or responses not clearly identiﬁed by raters as having followed a right- or left-facial
contraction (i.e., “I felt nothing” and “I feel the same as before”). This lead Schiﬁ' and
Lamon to exclude 64 of 106 data points. Although this procedure may have been
suﬂicient for their purposes, excluding those data may have removed the ambiguity that
would have led to less clear results. Ifthe 64 responses left out of their analyses had been
included, one wonders whether their data would look more like the data in the current
study.

Another possible explanation for the disparate ﬁndings could stem from differences
in the dependant measures used. In the current study, ﬁve of seven Depression Adjective
Checklists, all designed to measure State depression, were used to measure subjects’
aﬂ’ective experiences. As stated above, the DACL was selected as the primary dependant

measure because its parallel forms could be used for repeated measurement, because it

53

does not require a lot of time to pass between administrations, because it has been
standardized on populations similar to the population ﬁom which subjects were drawn in
the current study. and because the DACL has a reputation for being statistically sound
(Lubin, 1993a, 1993b, 1994a, 1994b; Lubin, Fielder & Van Whitlock, 1996; Lubin, & Van
Whitlock, 1995). For all these reasons the DACL was a good choice for a dependant
measure.

Nevertheless, there may have been a limitation that was not foreseen; the DACL,
while it appears to be adequate for detecting negative emotional experiences, may not
have been designed to have the ﬂexibility to detect the ambiguous positive-like responses
supposedly produced by right-face manipulations. As Schiﬂ‘ and Lamon (1989, pp. 926)
noted, “right-face contractions produced reports of positive aﬂ’ect (‘up’ ‘ good’) and what
can be characterized best as a mixture of positive aﬁ'ect and aggression (‘cocky’, ‘srnug’).
Given this, the contrast between the negative and positive—like responses may have been
diluted because the DACL is intended to detect clearly positive, not positive-like,
responses.

Furthermore, because the DACL does place each response clearly into the positive
or negative realm, using the DACL as the main dependent measure may have inadvertently
forced the subjects to respond as if they were pressing the poles of the mood continuum
and, thereby, limiting their willingness and ability to respond honestly. In hindsight,
perhaps it was unreasonable to expect normal, healthy subjects to experience mood swings
so dramatic that they would report feeling very negative emotions at one minute and then
very positive emotions 45 seconds later. It may be more realistic to assume that a shift in

mood would take place but that it would not press the poles of the mood continuum. If

54

this is so, the DACL may be less useﬁrl in studies using normal healthy subjects and where
both ends of the mood continuum are being evaluated.

To their credit, Schiﬁ‘ and Lamon (1989, exp. 2 and 3) may have circumvented
these issues by using a more ﬂexible dependant measure (i.e., TAT cards) able to capture
subtle and ambiguous responses produced by facial contractions. Although users of the
TAT do not usually use a statistically validated scoring system, its properties may have
made it useﬁil for evaluating the subtle changes in affect (Wagner, 1995; Obrzut, 1983).

Wedﬂiﬂerenees. The second prediction pertained to the possibility of
diﬂ‘erences in cerebral organization for emotion. It was predicted that men would produce
more negative proﬁles than women following left-face manipulations, and more positive
proﬁles following right-face manipulations. These predictions were not supported.

There may be several ways to explain this non-signiﬁcant ﬁnding. The ﬁrst
possibility is that there are in fact no signiﬁcant diﬂ’erences in the cerebral organization for
emotion between men and women. In other words, the null hypothesis was correctly not
rejected. Although this explanation is plausible, it does not explain those experiments that,
whether focusing on the expression or experience of emotion or on either the Right
Hemisphere Hypothesis or the Valence Hypothesis, found signiﬁcant sex-related
diﬁ’erences in hemispheric control of emotion (Davidson, Schwartz, Pugash, &
Broomﬁeld, 1976; Moscovitch & Olds, 1982; Borod, Koﬁ’, & White, 1983; Campbell,
1978; Strauss & Kaplan, 1980). Other possible explanations for the absence of sex-related
diﬁ’erences, therefore, must be considered.

One other possibility is that the lack of sex-related differences is due, at least in

part, to male examiners testing all subjects. There is a large literature replete with

55

examples of experimental confounds produced by an interaction of sex of examiner and
sex of subject (Franco, 1985; Kupke, 1983; LeVine, 1983), so it is possible that the sex of
the examiner had an inﬂuence here as well. Given this possible explanation of the null
ﬁndings, coupled with the unusual ﬁnding of men reporting more overall emotion than
women (even though there was no signiﬁcant diﬁ’erence; see Table 11), one could
hypothesize that women had a harder time than men reporting emotions to the male
examiner. Although the potential inﬂuence of this confound was recognized before the
study began, the constraints of time and resources prohibited an appropriate remedy.
Namely, it would have meant doubling the sample size and the number of examiners (two
female examiners) to control for the inﬂuence of sex of examiner on subjects. Although
the price for not taking these precautions can be substantial, realistic constraints must also
be recognized.

The current results —- the absence of sex-related diﬂ’erences and the unusual
ﬁnding that men responded to more adjectives than women, albeit not signiﬁcantly more,
— therefore, can be added to the growing number of studies already mentioned that have
reported inconsistent results for sex-related differences for the exmrienee and expmsipn
of emotion (e.g., Davidson & Schwartz, 1976; Davidson, Schwartz, Pugash, &
Broomﬁeld, 1976; Moscovitch & Olds, 1982; Borod, Koﬁ’, & White, 1983).

Here, attempts will be made to expand the possible explanations for inconsistent
ﬁndings in studies of sex-related diﬁ‘erences in emotion. The inconsistencies, in part,
might turn on the distinction between the terms “sex” and “gender.” Although these terms
are often used interchangeably, there is a difference between them that is useﬁrl to keep in

mind when doing this type of research. The term “sex” refers to physiological

56

characteristics, including those hemispheric diﬁ‘erences reported in the cognitive literature.
The term “gender,” however, refers to those socio-psychological characteristics used to
deﬁne traits stereotypically considered either male or female. Although the technical
diﬁ’erence is reasonably clear, the actual product of these differences is diﬁicult to tease
apart.

The primary focus of the current study was to test the possibility of sex-related
differences in the cerebral organization of emotion. A smaller part of this study, the one
that predicted support for the response-bias theory, focused on gender. In either case, it
would be difﬁcult to state that the tested emotional experiences were solely the product of
either sex or gender. It may be more accurate to assume that all the hypotheses examined
a combined inﬂuence of both sex and gender.

Not separating these two important variables, however, has not hindered the ability
to ﬁnd consistent results in studies comparing men’s and women’s diﬁ’erences in cognitive
ﬁrnctioning. There appears to be slippage between the cognitive literature and the
emotion literature where sex-related differences are concerned. As previously noted, the
literature on cognitive differences consistently reports small but statistically signiﬁcant sex-
related diﬁ’erences. That is, males typically perform at a higher level than females on tasks
with a strong visuo-spatial component (right hemisphere dominant tasks) and females
typically perform at higher levels than males on tests of perceptual speed, memory for
location of common objects, verbal ﬂuency, and ﬁne motor skills (left hemisphere
dominated tasks). Perhaps where the emotion literature is concerned, the tasks used to
examine these issues are not sensitive enough to detect differences. Or, there may not be

sex-related differences for emotion even though there are diﬁ’erences in cognitive abilities.

57

As a ﬁnal possibility, if one could clearly separate which responses were derived from
biological mechanisms and which were derived from socio-psychological mechanisms, and
examine them individually, perhaps studies comparing men’s and women’s affective
differences would produce more consistent results. Whatever the reason for the different
ﬁndings, the gap that exists between these areas of study is clearly deﬁned and hopeﬁrlly,
one day, will be narrowed.

WI. The third prediction was that the interaction between the side of the
face manipulated last and the sex of the subject would guide the intensity of responses on
the POMS-BI. This prediction was not supported. That is, regardless of the side of the
face manipulated last, the POMS-BI did not discriminate the subjects' emotion as being
more intensely positive or negative, and it also failed to disclose diﬁ‘erences between men
and women.

As with any measure, the POMS-BI has both strengths and weaknesses. In
choosing this measure as a dependent variable, it was decided that its strengths
outweighed its weaknesses. The next section will describe these strengths and
weaknesses.

On the positive side, the POMS-BI took between three and ﬁve minutes to
complete, making it easy to administer, and it is easily scored with scoring keys. Also,
since the responses are coded on a Likert Scale, information about the intensity of the
subjects’ feelings is available. These features allowed much information to be gathered in
a short time.

There are also important limits to the POMS-BI that must be examined. For

instance, the manual for test administration and scoring does not provide clear validity and

58

reliability scores (aside ﬁom test-retest scores) from standardized samples. Also, although
the six bi-polar scales are given in the manual, no data verifying these six bi-polar scales
are presented. It was necessary, therefore, to verify the assumptions discussed above by
completing a Principles Component Factor Analysis before any strong conclusions could
be stated. The six-factor prediction presented by its authors was not supported, so a more
reliable factor structure had to be determined in order to test hypotheses generated ﬁ'om
use of the POMS-BI.

Unfortunately, data gathered ﬁom the POMS-BI are of questionable value because
the POMS-BI was not found to have the structural stability reported by its developers,
Lorr and McNair (1988). Although attempts were made to modify this test, trying to
create a valid tool to measure the intensity of emotional responses in the current study has
dramatic limitations. These limitations hinge on the problem of having only a small sample
for testing the validity and reliability of the modiﬁed measure. In the end, any results
generated from this measure must be viewed with great caution.

W. This prediction supposed that women would
report more overall emotion than men. In other words, even after a speciﬁc manipulation,
women would produce more emotional responses, both positive and negative, than men.
This prediction was not supported. Instead, just the opposite was found; the average
DACL scores following the last manipulation showed that, regardless of condition, men
reported more overall emotion than did women. The phenomena explained by the
response-bias theory discussed earlier were, therefore, not supported in these ﬁndings. To
repeat, that theory suggests that women view the expression of emotion as a positive

symbol of connectedness between individuals, whereas men view it as a symbol of

59

weakness. Thus, the theory supports the notion that women express emotions or feelings
to others, including interviewers, more readily than men (Mirowsky & Ross, 1986), which.
may also suggest that men actively inhibit the expression of emotion. There are data
supporting sex-related diﬂ’erences whereby men were thought to actively inhibit emotional
processes (Burton & Levy, 1989). The issue regarding men actively inhibiting emotion

will be considered in more detail below.

11' . ES I E l' .
W In the next Predictions, it was thought

that the effects of the manipulations would become more apparent after the subject
manipulated both sides of the face than after only the ﬁrst manipulation. These predictions
were not supported.

In an attempt to explain the lack of agreement between the current results and
those reported by Schiff and Lamon (1989), Kop, Merckelbach, and Muris (1991), we
return to the original rationale for these predictions. First, a restatement is made regarding
both Schiﬁ‘ and Lamon’s (1989, 1993) reasoning for this procedure and these predictions,
as well as the criticism brought against them by Kop et al., (1991) for using this approach.
After that, the current study’s rationale for following Schiﬁ‘ and Lamons’ procedure is
discussed and possible reasons for incongruent ﬁndings are given.

Recall that two opposing theories have been suggested regarding these predictions.
First, Schiﬂ’ and Lamon (1993) proposed that the effect of the manipulations would

become more apparent alter the subject manipulated both sides of the face than after only

 

10’The underlying assumptions are the same for both “timing of manipulation”
predictions and, therefore, they are discussed together.

60

the ﬁrst manipulation. They drew on this assumption as support for using an alternating
manipulation sequence (e.g., RLRL and LRLR) as their method of mood induction.

In contrast, Kop et al., (1991) argued that the eﬂ‘ect of the manipulation should
not change depending on whether alternating contractions or consecutive contractions
(i.e., RRRLLL, LLLRRR) are used. In their reply to Kop et al. (1991), Schiﬁ’and Lamon
(1993) stated: “We believe that alternation of contractions within subjects makes the
induced states more salient through contrast, and thereby increases the probability that
they [the feelings] are reported” (p. 550). What is not clear is why Schiﬁ’ and Lamon
(1989) believed that the contrast was important for producing reports of the emotional
experience. This is especially unclear given their initial experience when three subjects
wept following left-face manipulations (Schiﬁ‘& Lamon, 1989, exp. 1). The act of
weeping is a clear and strong emotional response and would not seem to require a contrast
to bring it about or to make the underlying emotion salient.

As one possible explanation for these results, Schiﬂ’ and Lamon (1989) pointed out
that the subjects who wept were atypical in that they were “. . . acquaintances of the
experimenter. They were selected because the experimenter considered that they would
be sufﬁciently comfortable feeling unexplained emotions, they would not interfere with
such experiences and would also be willing to report them” (p. 925, exp. 1). These
characteristics presumably inﬂuenced the intensity of these subjects’ responses. Had they
not been acquainted with the examiners, however, they may have responded in a less
dramatic way. Furthermore, had this not been the ﬁrst time Schiﬁ’ and Lamon used the
manipulation, one could argue that other, less dramatic experiences would have led them

to believe that the contrast produced by alternating the side of facial contractions was

61

necessary to produce clear results. As it stands, however, the subjects’ relationship to the
experimenters, and the fact that this was the experimenters’ ﬁrst experience with the
procedure, might lead one to believe that Schiﬁ’ and Lamon’s (1993) explanations were
generated as a post hoc response to Kop et al. (1991).

Be that as it may, in the current study, it was believed that the emotional
experience muldjesuhﬂe and in some instances might resist clear labeling, especially
those experiences that followed right-face manipulations. It was predicted, therefore, that
these subjects would require the contrasting emotions before they could clearly realize
that their emotional experience had changed. These assumptions follow, in part, ﬁ'om the
explanations just mentioned. Nevertheless, choosing to alternate facial-contractions was
apparently insufﬁcient in bringing about a signiﬁcant emotional experience. As previously
stated, another reason for not ﬁnding signiﬁcant differences may be related the DACL’s
inability to detect diﬂ‘erences between negative and positive-like emotional experiences.

Iimingpﬁmanipulanpnmmubjeet. Finally, on the hypothesis that men are
more clearly lateralized than women, it was predicted that men would produce more
exaggerated DACL-D and DACL-G scores than women. The results did not conﬁrm
these predictions; the last two manipulations for men did not produce more exaggerated
DACL responses than those for women.

Although these results do not support the predictions set forth above, they are
consistent with the other results of the experiment. That is, none of the ﬁndings presented
in this study showed signiﬁcant sex-related diﬁ’erences. This may suggest that the
methods used in this study to measure mood changes were reliable but perhaps had poor

validity.

62

Cl" II 1"

This research is thought to have implications that apply not only to
neuropsychology but also to clinical psychology. It was mentioned earlier that clinicians
could beneﬁt greatly ﬁ'om a more thorough understanding of emotional systems and
ﬁrnctioning. This is certainly true where sex-related differences exist. To restate,
clinicians tend to see more women in psychotherapy than men. It was previously
suggested that men may not be readily attracted to traditional talk-therapy (a left-
hernisphere dominant approach) and may be more inclined toward using a visualization-
type technique (arguably a more right-hemisphere dominant approach). The ﬁndings in
this study, however, would not support implementing such techniques based on the sex of
a client.

The lack of signiﬁcant sex-related differences presented in the current study does
not mean that no differences exist between men and women. After all, there have been a
number of studies already discussed that found differences between men and women for
other emotion-related processes (Bryden, et al., 1991; Ladavas et al., 1980; McKeever &
Dixon, 1981; Rizzolatti & Buchtel, 1977). Additionally, differences between men and
women in the prevalence of, and intensity of, various clinical problems like depression
(Aneshensel, 1992; Klerman, W eissman, Rounsaville, & Chevron, 1984; Mirowsky &
Ross, 1986) and anxiety (Aneshensel, 1992; Mirowsky & Ross, 1986) have also been
covered in some detail. Since other studies have found important differences, one cannot
discount the possibility of useﬁrl therapeutic techniques that would beneﬁt men and
women diﬁ‘erently. In ﬁrture research, therefore, therapeutic techniques speciﬁcally

designed for men and women may be an important avenue to explore.

63

In other ways, some of the issues presented in the current experiment may still be
helpful when conceptualizing clinical issues. For example, diﬁ’erent clinical theories have
their foundations in diﬁ‘erent approaches to understanding emotional processes. As
previously noted, many of the techniques used to induce and examine emotional
experiences are conta. .mated by unwanted cognitive inﬂuences on research manipulations
(e.g., societal inﬂuences on watching movies). It is still assumed that the technique used
here is less inﬂuenced by cognitive interference. Given this, two diﬁ‘ering approaches to
therapy will be discussed; 1) Cognitive-Behavioral Therapies (CBT) and 2) Affective
Therapies (AT). Then, the ways in which these therapeutic approaches may beneﬁt from
this type of research will be discussed.

CpgmnyezﬂehanpmlAppmm. As explained by Bloom (1992, section 3, pp.
158), a primary premise at the foundation of cognitive-behavioral therapy is that small
changes in faulty cognitive processes and negative behaviors will produce signiﬁcant
changes that will reverberate throughout the personality and can bring about changes in
many areas of ﬁrnctioning. The implicit message is that, where change in therapy is
concerned, cognition and behaviors are the primary mechanisms and emotions are
secondary mechanisms. Research that focuses primarily on emotion and secondarily on
cognition, therefore, would likely hold opposing views to those espoused by theorists
supporting a CBT approach.

W. Contrary to the Cognitive-Behavioral theorists, Aﬁ’ect
theory, as presented by Tomkins (1987), assumes that the primary mechanism for
successful therapy and therapeutic change is affect, not cognition. That is, affect is the

blueprint for cognition, decision making, and actions. A therapist who is partial to using

64

affect theory to guide the therapeutic process, therefore, would focus on the affect a client
is experiencing and assume that the cognitive and behavioral changes will follow.

Clearly, the foundation of these theories rests on the question, which comes ﬁrst,
the cognition or the aﬁ’ect? Currently, we have little concrete knowledge to answer this
question. As Schiﬁ‘ and Lamon (1989) suggest, there is no clear answer to the dispute
over whether emotion always has a secondary inﬂuence on the cognitive appraisal of
events (Schacter & Singer, 1962; Lazarus, 1984) or whether emotions can come before
cognition (James, 1884; cited in Schiﬁ’ and Lamon, 1989; Zajonc, 1980).

The theory behind the current study supports the supposition that one can
experience aﬂ’ect primarily and have cognition secondarily because, prior to the induced
emotional change, subjects had no cognitive awareness of impending emotional changes.
Although no concrete answers to these complex questions are generated by the current
study, it oﬂ'ers another way to approach these questions.

Where clinical neuro-psychopharmacology is concerned, it was previously stated
that the type of research presented in the current study may help explain how various
mood-altering medications may have diﬁ’erent sites of action for men and women. The
current study, however, did not oﬁ‘er results that would support sex-related diﬂ’erences
but still may have oﬁ’ered helpful information in this research area. These results may
guide one away from certain investigations, since, for instance, no sex-related diﬁ’erences
in lateralization for the experience of emotion were found or it may guide changes in one’s
approach to ﬁrture investigations. For instance, instead of using only male rats in
pharmacology laboratory studies, a common practice because of less predictable female-

rat hormone cycles, use both male and female rats in laboratory studies of drugs known to

65

alter emotional experiences. This may help to elucidate the current discrepancies found in
the literature on sex-related differences in emotion.

Although this study has raised questions that, if studied in the ﬁiture, could move
us closer to understanding the lateralization of emotion in general, as well as possible sex-
related differences in lateralization of emotion, there are many other ways to approach
these questions as well. Some of these other ways are discussed below.

E B I 11° .

As stated above, the dependent measure used in the current study has received
support in the literature as a statistically valid and reliable measure. For this study,
however, it may have lacked the sensitivity needed to detect subtle emotional changes. It
is possible that a projective measure such as the TAT would be better at detecting these
subtleties. A study comparing results ﬁ'om the DACL with those ﬁ'om the TAT may
resolve this question and would begin to establish a consistent replication procedure,
oﬁ’ering a guide for future researchers to know which tools provide the richest
information.

Also, the question was raised whether the subjects would report more changes in
emotion if the experimenter and subjects became acquainted prior to the experiment.
Meeting with subjects ahead of time and becoming somewhat acquainted perhaps would
enhance the subjects’ willingness to reveal their true range of feelings. Of course,
experimenters should take care to control for experimental biases, since they will not be
blind to their subjects prior to the experiment.

Another way to enhance the probability of revealing emotional responses would be

to include a pro-test scenario wherein an emotional experience is shared by potential

66

subjects and the examiner. By sharing an emotional experience, subjects may feel more
connected to the examiner, which would establish an atmosphere that would permit
subjects to report their feelings more ﬁ'eely.

Next, using both male and female examiners in ﬁrture research will be vitally
important, especially where possible sex-related differences are concerned. In order to
establish appropriate control over all possible conditions, both male and female subjects
should be tested by both male and female examiners. This design would let experimenters
determine whether the sex of the examiner plays a role in the subjects’ performance.

Furthermore, it would be interesting to see whether adding another set of
contractions may intensify or enhance the emotional responses, especially if the
environment were comfortable enough to permit greater expression. A RLRLRL
condition may reveal more intense emotion than a RLRL condition. Similarly, a hand
contraction, used as a priming mechanism prior to a facial contraction, may produce a
clearer, more robust response than a facial contraction alone. There is already evidence
suggesting that, by using various mood induction techniques and by priming either positive
or negative emotions, subjects' social judgments change (Constans, 1993 ), visual
recognition task responses change (Hermans, 1994), and ratings of emotional intensity
become more congruent with the subjects’ own emotional state (Hansen, 1995).

Based on the description of the facial sensory and motor pathways to the Central
Nervous System (see p. 16), it is clear that the mood induction method used in this study
stimulates primarily the contralateral hemisphere of the cerebral cortex. Yet, it is not clear
where exactly the stimulation is taking place (beyond the motor strip), or what peripheral

areas are being aﬂ’ected. Functional Magnetic Resonance Imaging (fMRI) or Positron

67

Emission Tomography (PET) could be used to better understand the neuronal mechanisms
correlated with the manipulation and the subsequent emotional responses.

There are still other ways to use imaging techniques to examine sex-related
diﬂ’erences. It would be an interesting but difﬁcult task to test the notion, as presented as
an extension of the response-bias theory, that men actively inhibit the expression of
emotion. One possibility for testing this type of a question would be to use imaging
techniques such as ﬂVIRI and to compare activation of regional areas of the brain. Given
the roles played by the frontal and parietal regions for expression and perception,
respectively, these regions also may serve as a likely starting point for the study of the
experiencing of emotion. The diﬁculty with such a study would be trying to determine
whether “activation” in a speciﬁc region represented excitation (increasing the likelihood
of a response), or “inhibitory” (reducing the likelihood of a response). Presumably, a clear
understanding of which neurotransmitters and receptors (excitatory or inhibitory) are
involved in those regions and which are actually activated would be necessary.

Finally, even if these questions were answered, additional conceptual issues
remain. For instance, if men were found to “actively inhibit” the expression of emotion,
the active inhibition might reﬂect a soda-psychological inﬂuence whereby men are taught
to inhibit the expression of emotion. Still, the questions remains whether men may be
biologically predisposed (i. e. hard-wired) to actively inhibit the expression of emotion.
These questions cannot be addressed in detail here, but they can be considered as ﬁxture
research questions.

In addition to the methodological issues discussed above, there are conceptual

issues that have been raised by the current study. For instance, the importance of

68

understanding the implications of the differences between the terms “sex” and “gender”
cannot be over-emphasized. Although many people use these words interchangeably, the
actual differences are signiﬁcant and arguably should be considered pivotal in deﬁning the
goals of future research in the area of emotional experience. One example of ﬁxture
research that emphasizes this diﬁ‘erence is presented here.

Finally, in an attempt to continue bridging the gap between research and clinical
work, ﬁxture studies should consider potential clinical implications. For instance, as
previously stated, one question raised here was whether it was reasonable to expect a
dramatic mood shift in a normal and healthy population. In individuals whose moods are
more labile (e.g., individuals with bi-polar disorder or hysterical tendencies), more robust
ﬁndings may develop that have implications for both the neuroscience and the clinical

communities.

APPENDICES

APPENDIX A

APPENDIX A

INFORMED CONSENT FORM

This is a study of the effects of facial muscle contractions on mood in normal,
healthy people. It is a “normative study”, which means that we don’t know how normal,
healthy people will respond, and we want you to help us ﬁnd out. There are no right or
wrong ways to respond. For example, you may feel a change in your mood or you may
feel no change at all. Either response is possible, and that is why there is no right or
wrong way to respond. The study will take about one hour to complete. You will receive
one hour of research credit for participating.

Your record will be kept completely conﬁdential. To ensure this, you will be
assigned a subject number, which will be the only way to identify you in any reports about
the study. Only the experimenter will have access to ﬁles and lists that can be used to link
a name with a subject number. Any publications resulting ﬁ'om this work will not identify
you by name or in any way that would allow your identity to be discovered.

For this study, you will be asked to complete several tasks.

1. On the Subject Identiﬁcation F arm, you will be asked to write your name.
This is the only form that can identify you by name and link you to the subject
identiﬁcation number that has been assigned to you. Only the experimenter and his
designates will have access to your ﬁle. To ensure conﬁdentiality, all other documents you
ﬁll out will have only your subject identiﬁcation number.

2. On the Medical History Questionnaire, you will be asked whether you have a
history (either past or present) of clinical depression, bipolar episodes, substance abuse,
anxiety disorders, hypertension, and facial paralysis. This form is a screening measure
which will be used to identify and exclude individuals with any of the above-mentioned
histories. For this study, such individuals cannot be included.

3. On the Personal Information and Background Questionnaire, you will be asked
to provide your age, sex, ethnicity, religion, years of education, and academic major.

4. On the Laterality Questionnaire, you be asked to identify which hand you

prefer to use for a variety of common tasks. You will be asked whether anyone (including
yourself) has tried to change your hand preference.

69

APPENDIX A (cont’d).

5. On the Proﬁle of Mood States-Bipolar form, you will be asked to check off
adjectives on a list that describe how you feel at a given time.

6. After completing these questionnaires, you will be asked to perform four facial
manipulations, speciﬁcally, to move the lower part of your face into a certain position and
then to hold that position for 45 seconds.

7. Aﬁer each facial manipulation, you will be asked to complete another mood
check list. For this measure you again will be asked to check oﬂ’ adjectives describing
how you feel.

8. Lastly, you will be asked to complete a diﬁ’erent Proﬁle of Mood State-Bipolar
check list.

A debrieﬁng period will follow your participation in the experiment. You will receive
one hour of research credit for your participation. You have the right to discontinue your
participation at any time and for any reason, and to do so without explanation or penalty.
There is also an alternative method of earning extra course credit if you do not wish to
participate in this or other experiments. Ifyou wish to seek this alternative, speak with
your Introductory Psychology professor for speciﬁcs. The assignment may vary from
professor to professor, but it is usually to write a short paper.

If you would like ﬁxrther information regarding your rights as a research subject,
you may contact the Oﬁce of the UCRIHS at Michigan State University by telephoning
(517) 355-2180.

After the entire study is completed, if you have any questions or if you want a
written summary of the general results, you may contact the investigators at their
university oﬁices.

 

I have read this consent form, I understand the conditions, and I voluntarily agree to
participate in this study.

 

 

Subject’s Signature Date

 

 

Investigator’s Signature Date

70

APPENDD( B

APPENDIX B

SUBJECT IDENTIFICATION FORM
This is the only form that can identify you by name and link you to the subject
identiﬁcation number that has been assigned to you. Only the investigators and their

designates will have access to your ﬁle. All other documents you ﬁll out will have only
your subject identiﬁcation number in order to ensure conﬁdentiality.

Please print your ﬁxll name.

Subject Name: Identiﬁcation Number:

 

 

71

APPENDIX C

APPENDIX C

MEDICAL HISTORY QUESTIONNAIRE

Please check whether you have, or at some time in the past have had, any of the following
conditions:

CURRENT PAST

Yes N 0 Yes No
Major Depression _ _ __ _
Bipolar Disorder _ __ _ _
Generalized Anxiety _ _ _ _
Panic Disorder

Post Traumatic Stress Disorder
Substance Abuse

Hypertension

Schizophrenia

Other:

 

Have you ever suffered ﬁ'om any
type of facial paralysis?

Ifyes, please describe:

 

 

 

 

 

 

72

APPENDIX D

APPENDD( D

PERSONAL INFORMATION AND BACKGROUND QUESTIONNAIRE
. What is your age?
. Are you? Male: _ Female: __

. What year in school are you in now?

 

. What is your academic major?

 

. Which ethnic/race category best ﬁts you?

Caucasian: _ Asian: _
African-American: _ Hispanic: _
Native American:. _ Other: _ Deﬁne:

 

73

APPENDIX E

APPENDIX E

LATERALITY QUESTIONNAIRE

A. General Handedness

Circle the description that best applies to you: '
a. Strongly leﬁ-handed

b. Moderately left-handed

c. Ambidextrous (either-handed)

d. Moderately right-handed

e. Strongly right-handed

Circle the picture which best depicts how. you hold a pencil when writing.

Len-Handed Writers Right-Handed Writers

 

 

74

APPENDIX E (cont’d).

B. Hand Use for Specific Tasks

For each of the tasks listed below, check the column that corresponds to the hand you
would use to perform that task.

Always Usually Usually Always
Left Left Both Right Right
1 2 3 4 5

Hand Use

1. Write a letter

2. Hammer a nail

3. Throw a ball at a target
4. Unscrew lid of a jar

5. Use knife to cut bread
6. Use tooth brush __ _ __ _ _
7. Hold a match while striking it_
8. Hold a tennis racket

Has anyone (including yourself) tried to change what hand you use to perform any of the
tasks listed above or any other tasks? No Yes

If your answer was yes, please identify the task(s) and describe the circumstances.

 

 

 

 

 

75

APPENDD( F

APPENDIX F

POMS-BI

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 
    

  

  

— j EOENDFICADON
—"“_M“ “'5 l I ' I I I I

- Below are words mar describe Outrage mm: mm ”can «an EVERY m cart-tum. Thorn Mt I I I I .

— saga-mummammunvmcnaummnaoaoumungounmcmEPmWEEK IOOOOOOOOO
- Marksman “alumna theme-rm amen-s ebmttohowmmmlm W'WG Igggggggga
_ THE PAST WEEK metucmc room. I mama clascncus —I .®®®®®®®®Q
- The numbers Mar to I m—I IGGQQQGQQGI
- the" phrases . use a so. 2 '91ch can. I®®®®®®®®®
"" 0 - Macaw-him ‘- muons.“ malts .®®®®®®®©@I
- r - swim ma. rm '. ruse cumtv. L®®®®®®®®®I
"' 2 - 940mm- rm ! counter was Inca-stir «an IQ©®®®®®®®I
- J carbonation». 3 '1 0‘.“ 3?:3’; .Q’ QQ‘ QEDMJ’ G
: «2f. I it § 3‘11 I 1““? :
- it". sf: s21. ’ ‘3 i «a ‘ 3;”: 2'2 ‘
-: ems. ram W were. I
- \ ”2‘13” ’3 ms ’5 75 «up 3 \3‘ r; a"; g
-' \ \o’ o \ Q‘s \ ‘3’}; \\.$$ \:
_‘ \ J"'\\ \ ’\ . \ ' \ \ I
-; r. em... ....... ©O®®m. v-ooro... ............. @QQGIsr. son... ................ @®® Isa. lady-woo ......... ©®®GI

—; 2.11m ............. ©O®®Izo M... ............. ©®®©Iaa um... ....... @QQQI-rr. one“... ......... «more:

—I :l. Cheerful .......... @OCDQIZL may ................ @GQGIa. MM ................. ©®®®Isr. Good-natured ....... @OQCDI

—j a. w... ............. ©®®®I22. hug-ac .............. @GGGHD. was...” ........ ©0®®Isa w... ................ (we I

-' 5. road. ............. ©0®®Iu lord ................... @CDQQDIar. shorty ................ @0Q 59. Conﬁdent ............ @GQCﬁ

-E a. sour...“ ........ @(DCDCDlzc. uric...“ .............. ®®®®ua Paolo-dd ............ @OQGIeo. W ......... ©®®®

—: 1. may ............. GOG 23. hoe-cu .............. @®® do. Adm ................. @®®®Isr. Mead. .............. ©0®®I
-I a. 3.x ............... (cos) :0. M... ............... (media.

—I s. M .......... @GQ) 21. 59mm ......... @®®®I«.

-:ro. rind .............. ©®® as. M ............... ©®®®Iu.

-:rr. Strong ............ @063 as. Jim”; ............ ®®®®Ida

-:u. w ..... @®® Ian. W .......... @OQQICI Mucmu©®®®ilt .
—'l:t um ....... @GGQIJI. W ............. ®®®®lds Calm .................. @®®®m. Moo... ........... ®®®Q
-_rd. om... .......... @oexan. a»... ............ :...®®®®gso. a... ................... ®®®®Isa on..." .............. ®O®Gi
-;rs. mm ............ @ocacam. Svmm ......... @oeqsr. 1.... ................... coaxed... Arm ......... @oom
— . ' l
—j... rm ............. @oeegu an...” ........... @oeqsz. on...“ ............ ©®®®Iro Mm... ......... (goes):
—§n. Nam .......... ©®®®Iu w... ............. @o@@§ra. Arum: .............. @QQQIn. 3...“... ......... @oeq
nus. mm ......... I . ............. .54. Mice ............. 000012. . alert ....... oooo
_ mm m. . mo arm as suns YOU HAVE ANSWERED EVERY ITEM

76

APPENDD( G

APPENDD( G

ST-DACL
Form A

Directions: Below you will ﬁnd some words that describe different kinds of moods and
feelings. Check the words that describe how you feel right now - at this moment. Some
of the words may sound alike, but we want you to check all of the words that describe your
feelings. Work rapidly and check all of the words that describe how you feel right now. If
you need to change an answer, clearly erase the incorrect the answer.

List A
01. Wllted 017. Strong
0 2. Safe 018. Tortured
03. Miserable 019. Listless
O 4. Gloomy O 20. Sunny
05. Dull 021. Destroyed
0 6. Lively O 22. Wretched
O 7. Low-spirited O 23. Broken
0 8. Sad 0 24. Lighthearted
0 9. Unwanted 0 25. Criticized
010. Fine 026. Grieved
011. Brokenhearted 0 27. Dreamy
012. Downcast 0 28. Hopeless
O 13. Enthusiastic O 29. Oppressed
014. Failure 0 3o. Joyous
015. Aﬁlicted 031. Weary
O 16. Active 0 32. Droopy

77

ST-DACL
Form D

Directions: Below you will ﬁnd some words that describe diﬂ‘erent kinds of moods and
feelings. Check the words that describe how you feel right now - at this moment. Some
of the words may sound alike, but we want you to check all of the words that describe your
feelings. Work rapidly and check all of the words that describe how you feel right now. If
you need to change an answer, clearly erase the incorrect the answer.

List D
O l. Depressed 017. Fit
0 2. Elated 018. Lonesome
0 3. Awful 019. Unloved
O 4. Lifeless O 20. Glad
O 5. Grief stricken O 21. Grave
0 o. Inspired 0 22. Sunk
O 7. Woeful O 23. Shot
0 8. Lonely O 24. Merry
0 9. suffering 025. Wasted
O 10. Mellow O 26. Washed out
0 11. Drooping O 27. Clear
0 l2. Rejected 028. Gruesome
O 13. Fortunate O 29. Tired
0 14.Dreary 030. High
0 15. Lousy 031. Worse
0 16. Good 032. Drained

80

ST—DACL
Form C

Directions: Below you will ﬁnd some words that describe diﬁ‘erent kinds of moods and
feelings. Check the words that describe how you feel right now - at this moment. Some
of the words may sound alike, but we want you to check all of the words that describe your
feelings. Work rapidly and check all of the words that describe how you feel right now. If
you need to change an answer, clearly erase the incorrect the answer.

List G
O l. Heartsick O 18. Enthusiastic
0 2. Healthy 0 19. Bleak
O 3. Sad O 20. Grief stricken
0 4. Aﬁlicted 0 21. Eager
0 5. Lonesome 0 22. Drained
O 6. Fine 0 23. Desolate
O 7. Alone 0 24. Miserable
O 8. Gloomy O 25. Merry
O 9. Depressed O 26. Dull
0 1o. Alive 0 27. Melancholy
O 11. Heavyhearted O 28. Interested
0 12. Failure 0 29. Unwanted
0 13. Glad 0 3o. Gruesome
O 14. Despondent O 31. Whole
0 15. Sunk O 32. Oppressed
O 16. Optimistic O 33. Lifeless
0 17. Jovial 0 34. Elated

81

82

Table 1a: Right Hemisphere Hypothesis

Left Hemisphere Right Hemisphere
Positive Negative Positive Negative
Perception + +
Expression + +
Experience + +

Table lb: Valence Hypothesis

Left Hemisphere Right Hemisphere
Positive Negative Positive Negative
Perception + +
Expression + +

Experience + +

83

Table 2. Correlations among ﬁve state lists for male and female
normals (N=86).

DACL-A DACL-B DACL-C DACL-D DACL-G

DACL-A — 0.88 0.86 0.88 0.88
DACL-B 0.91 _ 0.88 0.90 0.90
DACL-C 0.85 0.89 — 0.91 0.90
DACL-D 0.91 0.92 0.89 — 0.93
DACL-G 0.90 0.91 0.86 0.92 —

Note: Correlations for males (n=51) are below the diagonal; those
for females (n=35) are above the diagonal.

 

Table 3. Correlations among mood scales for Proﬁle of Mood States-Bipolar
(Lorr & McNair, 1988).

l 2 3 4 5 6
Composed 1 —— —— —— —— _ __
Agreeable 2 0.59 — — — __ __
Elated 3 0.62 0,73 — —— __ __

Conﬁdent 4 0.59 0.46 0.67 — — —

Energetic 5 0.40 0.49 0.68 0.64 —— ——

Clearheaded 6 0.68 0.58 0.64 0.74 0.57 ——

84

Table 4. POMS-BI retest correlations and mean scores
(Lorr & McNair, 1988) (N=66).

r lst Mean 2nd Mean
Composed 0.55 24.26 23.91
Agreeable 0.33 28.70 29.55
Elated 0.59 23.76 24.77
Conﬁdent 0.57 23.80 23.73
Energetic 0.60 21.59 20.91
Clearheaded 0.72 27.53 26.64

Table 5. Results of Conﬁrmatory Factor Analysis for a six-factor model
of the POMS-BI (N=120).

Variable Eigenvalue Percent of Variance

1 19.43 27.00
6.31 8.80

3 3.85 5.30
4 2.88 4.00
5 2.67 3.70
6 2.39 3.30
7 2.00 2.80
8 1.63 2.30
9 1.59 2.20
10 1.52 2.10
11 1.45 2.00
12 1.39 1.90
13 1.35 1.90
14 1.25 1.70
15 1.16 1.60
16 1.12 1.60
17 1.02 1.40

18 1.01 1.40

85

co.— nnd .Nd nvd 3.6 ovd 9N6 9.6 and ovd 3.6 and «Nd nNd ovd >325.“—
8._ 3.6 and Om... 3.6 wed had and 3.9 5.6 36 and and NV... tome-635.3
8; med :6 mad vmd 36 on... cud 31¢ and nNd and mad Bonn—chain...”—

8.. .4... 9.... R... an... 3.... an... 8... .e... 2... .3. :3 88.0.2
8.. S... «v... a... an... an... «e... 9.... an... en... 8... e3%.:

8.. m2. 3... 2.... e2. 2... R... S... «a... 8... 8.6.3....

8.. 2... x... 5. 3.... 3.... 8... 9... 8... £26.96

8.. .2. 8.... :3 3... a... a... on... egg...

8.. a... 2... ea... 8... .2. 8... gm...

8.. an... a... 2... a... .2. 88......

8.. S... 3... 2... we... 3....“

8.. an... .2. 3.... 8..-niece...

8.. .e... 3... Been

8.. 2.... 632nm

8.. 382....
2:22:2eaeeaenn.5.92.33
8.. an... an... 2... 8... 8... 8... a... 9.... an... 8... 8... 85.6mm.
8.. 8... en... e2. 8... 2.... 8... x... on... a... .2. 8698......

8.. a... S... a... R... .e... 3... we... 2.... we... 5.6.2...

8.. 3... «N... on... 5... .2. 3.... an... on... 8.3.2885

8.. on... 2.... 8... an... 9.... e2. 9... 22......

8.. on... an... 2... en... .2. 3... .33.:

8.. 8... on... on... an... em... 2889‘...

8.. em... on... 8... an... 2.6:

8.. 3.... a... .2. .8360...

8.. 3... 2.... 58865

8.. 3.... bad

8.. 9....o .3”...

N.._e.o....ememu.e>£8..§8£a

.8215 Ems—Om o... :e easing. 38-h o... 8.. @8023 «3:00.93 25am»... ‘5 25.3.. 3.. .252: cow-.230 6 03:.

86

Table 7. Factor loadings, reliability analyses, and alpha variance if items deleted
for positive and negative factors on the POMS-BI (N=120).

Factor 1 (positive)
Full of Pep
Kindly
Peaceﬁrl
Conﬁdent
Joﬂy
Agreeable
Playful
Lively
Good-Natured
Joyful
Energetic
Relaxed

Reliability Coefﬁcient

Factor 2 (negative)
Sluggish
Drowsy
Tired
Downhearted
Tense
Exhausted
Annoyed
Grouchy
Fatigued
Wwy
Gloomy
Bad Tempered
Discouraged
Uneasy

Reliability Coefﬁcient

Loadings Factor 1
-0.76

-0.62
-0.56
-0.64
-0.71
-0.60
-0.60
-0.75
-0.53
-0.73
0.74
-0.50

Alpha 0.91

Loadings Factor 1
0.75

0.71
0.62
0.52
0.47
0.71
0.39
0.56
0.73
0.70
0.57
0.52
0.61
0.50

Alpha 0.91

Loadings Factor 2
0.21

0.10
-0.21
-0.26
0.27
0.05
0.34
0.27
0.17
0.19
0.31
0.78

LoadingsFactorZ
-0.11

-0.24
—0.27
0.48
0.22
-0.30
0.50
0.33
-0.28
-0. 15
0.40
0.42
0.47
0.42

Alpha ifItem Deleted
0.90
0.90
0.90
0.91
0.90
0.90
0.90
0.90
0.90
0.90
0.90
0.91

Alpha ifItem Deleted
0.90
0.90
0.91
0.91
0.91
0.91
0.91
0.91
0.91
0.91
0.91
0.91
0.91
0.91

Na.» 5.3
a.” 2.3
we.» Nvdn
g6 cwdn
Gm 53—2
Orr—0(a—

87

ame Ne. _ m
2 .w 3 .3
No.5 mmdv
N56 oven
Gm =82
quu<a

mad Swan
ems ceNn
c. .h 3.2
R6 86m
Gm :32
Urs—U<Q

mad 3.:
mm.” :09
use 3.3
end omdm
Gm 582
mrqu<D

med 3.:
and code
3% 3 .me
new 2.;
Gm Ewe—2
<rs~U<Q

£153-:qu 3
digs—-52
garcoﬁog
rad—-52

.8»qu 3052.3 1‘5: v.8 .215— 05 E 5863 v.8 =2: 8.. Gm Ea 8.8m AU<Q :82 .a 03:.

El.
ﬁnd
mod
86

ood
cod

:6

a...

09..

id

cod

mg .m
«NN .m
ms .—
VmN .m

2 .mn
oﬁme
NYNN
3...:

mad

3.:

26
m

S. .—

vmm .m

K. .—
a.

m. .mnm

363

3...:
mm

5m x 08:. x 55980
new 0. 08:.

.8m x 529.00

25,—. x 5:680

xom
25,—.

5:550

.812. .295. 6.2.18 8e .6 d .o .m .63. as... .25... .45...
53.950 .8 8§ta>reu .6 £9.92 8.382 3.33% ..o $3.8..— .a «3:.

Hue—o 00. wows—8 0.. 0.320 0.08010. >ZO<>m 0: 0.0 80000 33.0.3830: 0*. 0.0
0030.2. 0000 000003 .50 00053 00.00.76... 000 0020300 :5 03. 0000302000: EH03.

2000.78 2:29.00 mm 0.. 0.. 0

00.. 000.00 _. 00 0.00 0.00
000050.. 00.00 .. 00 .00 0. .0
me... x 000050: 0.3 r .00 0. .0 0.00
00:05 >& 8:48 mm 0.. m. 0

max 00.00 .. 00 E00 0.00
000030: 0.00 _. 00 0.00 0.00

mm... x 0000300 0.00 _. 00 0.00 0.3

88

89

_N.0

V0.0

0.0

V0.0

2.0

00.0

v0.0 :00
00.0. .06
mmé N0...
3.0 2.0

Om 0.8.2
MAM—1—

;.m 00.0

.04. _0.v

.04~ 00.0

00.0. 00.0~

Gm :32
12:

Q 0:55 O AU<D

U 35—: n AU<Q

m 2:08 U AU<D

< 0:005 m AU<Q

.80qu 800:8 «1:: s .25. as.» s 38.33 ___W 5 8.8... 88200:. 228% 0o 00 2a :82 .2 n=0:

N0.—

00.0

woN

00.0

mmN

00.0

a:
M: ._
w:
a:
M: ._

00. .—
=0

0.000

_0.0

2.00

00.0

2.00

.00
mm

32

0:. 2.:

00.0 00.0

00.0 00.0

00.0 9.0

00.0 0N0

am :32
0050?»

$2

0_.v mu...

0N.v 00.0—

000 00.0

02‘ 00.»

N00 00.0

Gm :82
:02

.545 1.0.0.0
0.1545
9-4045
U-..—U<Q
94043

<JU<D

.8qu 8:53 05 Be

.00 5 .D .U .m 2: AU<G :08 :0 038 8.000%»: 0530500 <>OZ< 3.88.0 2950 a 50.0 3380
0:0 5 .D .U .m .3 AU<Q :08 =0 08.005 magi? 00 508:: :82 .2 03:.

9O

 

 

3390:2005 ..00000 E 30> .om<

‘WOUJV
myseong
alowoqdos

uquscu :1

“L2

9Z'VZ

EZ‘ I Z
0Z'8 I

19‘110
new
unouauzv

E .
I l
E
l
H .10an
I

Joguss

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

222 n. _

 

 

 

 

29:00 I _

 

 

 

.Aownzv b00800
0083:2050 000 40008 E 300 50088 000 :03 5 300.33 0.0800 05 03E 00 0:00.00 ._ 0.50:—

warned

91

9i? 085

80.08 323 :35 0:0 .8“ a 00 8008
£008 :80:— 508 8 00:0— 0: 05 a 3 :3 =00 a
a 20: a 20: a 95 a 82 3200:: 0 30.5.0 3.55:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

222 0
one”... I

 

 

 

. will! In .‘lIulllnlll.

 

.30qu 8.80 306 :0 8008008065: .00 02:03 05 :2: E00 32008.. :82 .0 0.503

300.
a 255

 

 

 

Kluo u=I=o

$01003 mm

=9

Kluo um:

92

OJU<Q

000005 00 8:032

QJU<D UJQ<Q

p p

mJU<D <JU<Q

 

q d

n .00

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MAKAIIII
Avid ID. I

 

 

 

 

 

.smuzv <-.:0<0 23.3.:

-000 00050700 .93”: 0:130 0030000 0:10 .D .0 .m Ao<e 085. 0003000 0000825 .m 0.5»:—

SERIODS 'IOVCI "39W

93

0.00 0000-00-— . 3:0 0000-00000

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.Aowuzv 00503 000 00:. .000 .00 000000000 000.00 00.5.50 000003008... 0000-30: .00 00.000
AU<Q :00... 05 000000000 000.00 000.050 000000.300... 0000-02 .00 00.000 AU<Q 0002 .0 00:0..—

02

00—

00—

00—

891005 mm mm

94

Gomuoem MO 05mg:

0-4046 D.._O<n_ O.I_O<D

m.._0<o

(.4 9.3

 

 

 

 

 

 

 

 

 

 

 

 

00121000080..— IIOI

 

 

053.202 '0'
40.00-200.00 . . 9 . . .\

 

  

550-202 I 9 I

 

 

 

 

.80uzv 080.008

0000 05 0000 20 5 sec; 05 85 a. G .0 .0 .0 .< 09.8 0280 03.0 :82 .0 200.0

891038 ”lava "29W

LIST OF REFERENCES

REFERENCES

Alema, G., Rosadini, G., & Rossi, G. F. (1961). Psychic reactions associated with
intra carotid amytal injection and relation to brain damage. ExgemMedica, 31, 154-155.

Aneshensel, C. S. (1992). Social stress: theory and research. W
W: 1.8, 15-38

Banich, M. T., Stolar, N., Heller, W., & Goldman, R B. (1992). A deﬁcit in right
hemisphere performance after induction of a depressed mood. Wm

MW 5, 20-27.
Bellak, L. (1986). W (4th ed.). Larchmont: Grune

& Stratton.

Benson, F., & Zaidel, 13. (Edit)- (1935). WW9!)
inﬂumans. New York: Guilford.

Benton, A. L. (1959). W. New
York: Hoeber-Harper.

Bever, T. G., Hurtig, R R, & Handel, A B. (1976). Analytic processing elicits right
ear superiority in monaural presented speech. mm 14, 175-181.

Blanchard, C. D., Greibel, G., & Blanchard, R. (1995). Gender bias in the pre-
clinical psychophannacology of anxiety: Male models for (predominantly) female disorders.

WM, 29 79-82

Bloom,B. L. (1992). ' z - '
(pp. 157-159). Boston, Massachusetts: Allyn and Bacon.

 

Borod, J. C. (1992). Interhemispheric and intra hemispheric control of emotion: A
focus on unilateral brain damage mmmmmsmmxmmm QQ, 339-

348.
Borod, J. C., & Caron, H. S. (1980). Facedness and emotion related to lateral
dominance, sex, and expression type. W 18, 237-241.

95

96

Borod, J. C. ,Koﬂ‘, E. ,& Caron, H. S. (1983). Right hemispheric specialization for
the expression and appreciation of emotion: A focus on the face. In E. Perecman (Ed )

Wammmﬂemisnhm (pp 83- 110). New York: Academic Press.

Borod, J., C., Koﬁ‘, E., Lorch, M. P., & Nicholas, M. (1986). The expression and
perception of facial emotion in brain-damaged patients. Wada, 251,, 169-180.

Borod, J. C., Koﬁ‘, E., & White, B. (1983). Facial asymmetry in spontaneous and
posed expressions of emotion. W 2, 165-175.

Bouhuys, A L. ,Bloem, G. M, & Groothuis, T. G. (1995). Induction of depressed
and elated mood by music inﬂuences the perception of facial emotional expressions in healthy

subjects Wanders, 3.3., 215-226

Brewer, P. ,& Doughtie, E. B. (1980). Induction of mood and mood shiﬁ. humal
W, 16, 215-226

Brockrneier, B. ,& Ulrich, G. (1993). Asymmetries of expressive facial movements
during experimentally induced positive vs. negative mood states: A video-analytical study.

Wandﬁmoﬁm, 1, 393-405

Bryden, M. P., Free, T., Gagne, S., & Groﬂ‘, P. (1991). Handedness eﬂ‘ects in the
detection of dichotically-presented words and emotions. Cortex, 21, 229-235.

Buck, R. (1984). W. New York: Guilford Press.

Buck, R. N. ,Savin, V. J., Miller, R E., &Caul, W. F. (1972). Communication of

aﬂ‘ect through facial expressions in humans Watchman
21, 362-371.

Bulman-Fleming, M B., & Bryden, M P. (1994). Simultaneous verbal and affective
laterality effects. Nmmpmholggia, 12, 787-797.

Burton, L. A., & Levy, J. (1989). Sex diﬂ‘erences in the lateralized processing of
facial emotion. W11, 11, 210-228.

Campbell, R. (1978). Asymmetries in interpreting and expressing a posed facial
expression. (Long, 111,, 327-342.

Coﬂ‘ey, C., E. (1987). Cerebral laterality and emotion: The neurology of depression.
WM 1, 28, 97-219.

. Conger, R D ,Lorenz, F. 0, Elder, G. H., Simons, R. L., and Ge, X. (1993).
Husband and wife diﬁ‘erences 1n response to undesirable life events. Wham

Wat, 3.4, 71-88

97

Constans, J. I. (1993). Mood and the subjective risk of ﬁrture events. Cognitignmd
Emotion, 1. 545-560-

Coren, S. (1995). Differences 1n divergent thinking as a function of handedness and
sex Amenmloumalnfﬁmholm 108, 311-325

Collins, L. M. (1996). Measurement of change in research on aging: old and new
issues from an individual grth perspective. In J. E. Birren and K.W. Schaie (Eds), (pp.
38-56). San Diego, Ca.: Academic Press.

Dau'dson, R, Schwartz, G. (1976). Patterns of cerebral lateralization during cardiac
feedback versus self-regulation of emotion: Sex differences. Psychgphysiglggy, 11, 62-74.

Davidson, R, Schwartz, 6., Pugash, E., & Broomﬁeld, E. (1976). Sex differences
in patterns of EEG asymmetry- Watchmen :1, 119-138.

Davidson, R J, Schwartz, 6., E. ,Saron, C. ,Bennett, J., & Goleman, D. J. (1979).
Frontal vs. Parietal EEG asymmetry during positive and negative aﬂ‘ect. Bmhgphysjglogy,
16, 202-203. (Abstract).

Deaux, K. (1984). From individual diﬁ‘erences to social categories: Analysis of a
decade’s research on gender. W31, 32, 105-116.

Deaux, K, & Lewis, L. L. ( 1983) Assessment of gender stereotypes: Methodology
and components. W11. 25 (Ms #2583)

Dimond, S. J ., & Farrington, L. (1977). Emotional response to ﬁlms shown to the
right or leﬁ hemisphere of the brain measured by heart rate. W 11, 255-260.

Dimond, S. J., Farrington, L., & Johnson, P. (1976). Differing emotional response
from right and left hemisphere. Mann, 261, 690-692.

Dohrenwend, B., & Dohrenwend, B. S. (1976). Sex diﬂ‘erences in psychiatric
disorder. Amednanlcumalnfﬁociolm 82, 1447-1459.

Dohrenwend, B., & Dohrenwend, B. S. (1977). A Reply to Gove and Tudor.
Amerimlcunnlnﬁmziolon 82,1336-1345

Farah, M J., Gonzzaniga, M. S., Holtzrnan, J. D., & Kosslyn, S. M. (1985). A leﬁ
hemisphere basis for visual imagery? W 21, 115-118.

Farah, M J. (1984). The neurological basis of mental imagery. Cognition, 18, 245-
272.

98

Franco, J. N. (1985). Eﬁ‘ects of examiner variables on reported self-disclosure:

Implications fro groan Personality testing WNW 1,187-
197.

Frank, E., Carpenter, L. L., & Kupfer, D. (1988). Sex diﬁ‘erences in recurrent
depression: are there any that are signiﬁcant? WM, 155, 41-45.

Gainotti, G. ( 1972). Emotional behavior and hemispheric side of lesion. gangs, 8,
4 1-5 5.

Geffen, G., Bradshaw, J. L. & Wallace, G. (1971). Interhemispheric eﬁ‘ects on
reaction time verbal and nonverbal stimuli. WM, 81, 415-422.

Goldstein K (1939) IhLQrganismaAnﬂolisiiLAppmadeiolmﬁenxed
fronLRatliologigalDatajnMan. New York: American Book

Gove, W. R & Tudor, J. (1977). Sex differences in mental illness: A comment on
Dohrenwend and Dohrenwend. Ammmloumalgﬂmglogyjz, 1327-1335.

Graves, R, Landis, T., & Goodglass, H. (1981). Laterality and sex diﬁ‘erences for
visual recognition of emotional and non-emotional words. Wham 12, 95-102.

Haggard, M. P., & Parkinson, A M. (1971). Stimulus and task factors as
detenninants ofearadvantases W 21. 168-177

Hansen, C. H. (1995). Emotion speciﬁc priming: Congruence eﬁ‘ects on affect and

recognition across negative emotions. W 21, 548-
557. '

Harman, D. W., & Ray, W. J. (1977). Hemispheric activity during aﬁ‘ective verbal
stimuli: An EEG study. WM 18, 457-460.

Harris, L. J. (1979). Sex-related diﬁ‘erences in spatial ability: A developmental

Psychological view In C B KOPP (Ed) WWW
(pp. 133-181). New York: Plenum.

Harris, L. J. (1981). Sex-related variations in spatial skill. In L. Liben, N. B & A
Patterson (Eds), : . * .. . . ~ m (pp. 83- 125).
New York: Academic Press.

 

Harris, L. J. (1990). Cultural inﬂuences on handedness: Historical and contemporary
evidence. InS. Coren(Ed.,) » . - - .

Wm, 61, (pp. 195-25s) 131.000 Science Publishers, B. v. North
Holland Book Series.

 

99

Harris, L. J ., & Gitterman, S. R. (1978). University professors’ self-descriptions of

right-left conﬁisability: Sex and handedness differences. W, 51,
819-823.

Heilman, K. M., Bowers, D., & Valenstein, E. ( 1985). Emotional disorders
associated with neurological diseases. In K. M. Heilman & E. Valenstein (Eds), Q'migal
Wham (pp. 337-402). New York: Oxford University Press.

Hermans, D. (1994). The aﬁ‘ective priming eﬁ‘ect: Automatic activation of evaluative
information in memory. WW, 8, 515- 533.

Hughlings-Jackson, J. (1874). On the Nature of the Duality of the Brain. In J. Taylor
(Ed) Wm (1932 V01 2 PP 129-145) London

Hodder & Stoughton.

Mons-Peterson, M., G. (1988). WWW
Hillsdale, NJ: Lawrence Erlbaum Associates.

James, W. (1884). What is an Emotion? Mfmd, 2, 185-205.

Joseph, R. (1992) The limbic system: emotion, laterality, and unconscious mind.

Wm 12, 405-456.

Julien, R M. (1988). Wm, (5th ed.). New York: W. H. Freeman
and Company.

Kimura, D. (1964). Leﬁ-right diﬁ‘erences in the perception of melodies. Quaﬂeﬂy
W 1.6, 355-358

Kimura, D. (1992). Sex differences in the brain. SmsmiﬁsLAmeiican, 118-122.

King, F. L., & Kimura, D. (1972). Left car superiority in dichotic perception of vocal
nonverbal sounds WM 26, 111- 116

Klerman, G. L., Weissman, M. M., Rounsaville, B. J., Chevron, E. S. (1984).

W. New York: Basic Books.

Kop, W, J., Merckelbach, H., & Muris, P. (1991). Unilateral contraction of facial
muscles and emotion: a failed replication. 9.121128, 21, 101-104.

Kosslyn, S. M. (1987) Seeing and imagining in the cerebral hemispheres.
W, 93, 148-175

100

Kupke, T. (1983). Effects of subject sex, examiner sex, and test apparatus on

Halstead Category and Tactile Perception Test. WWW
23112119190141, 624-626

Ladavas, E., Umilta, C., & Ricci-Bitti, P. E. (1980). Evidence for sex differences in
right-hemisphere dominance for emotions. Ngummhglggia, 18, 361-3 66.

Landis, T., Assal, G., & Perret, E. (1979). Opposite cerebral hemispheric
superiority’s for visual associative processing of emotional facial expressions and objects.

Nam 21%. 739-740-

Lazarus, R S. (1984). In W, R. Scherer and P. Ekrnan (Eds),
(pp. 247-258). Hillsdale, New JerseyzEarlbaum.

LeVine, E. S. (1983). Eﬁ‘ects of therapist’ s gender, ethnicity, and verbal style on
clients willingness to seek therapy. WM, 121, 51 5-7.

Levy, J. (1982). Handwriting posture and cerebral organization: How are they
related? KW 21 589-608

Levy, J. & Reid, M. (1976). Variations in writing posture and cerebral organization.
Win, 2, 409-441.

Levy, J. & Reid, M. (1978). Variations in verbal organization as a ﬁinction of

lundednesshandposture inwritins. and sex WWW
101,119-144.

Lewis, L. M, Dember, W. N., Scheﬁt, B. K. ,&Radenhausen,R A. (1995) Can
experimentally induced mood aﬁ‘ect optimism and pessimism scores? WW

WEI, L4, 2941

Ley, R. G., & Bryden, M. P, (1982). A discussion of right and left hemispheric
effects for recognizing emotional tone and verbal content. WM 1, 3-9.

LOU, M., & McNair, D. M. (1988). BLOW. San
Diego, Ca.: Educational and Industrial Testing Services.

Lubin, B. (1993 a). Self-assessment of state and trait depressive aﬂ'ect by college

students: Reliability and validity of an adjective checklist. W
Dﬂelnnment. 3.4.. 249-255-

Lubin, B. (1993b). : - ' ° ' '
W. Psychological Assessment Resources 1110. Odessa, Fl.

 

101

Lubin, B (1994a). Development of critical item scale for the Depression Adjective
Check List 199W 30.

Lubin. B- (1994b) WW Odessa. Florida:
Psychological Assessment Resources, Inc.

Lubin, B, Fielder, E. R, Van Whitlock, R (1996). Mood as a predictor of discharge
from AirForce basic training. Why, 82, 145-151.

Lubin, B., Van Whitlock, R. (1995). Levels of depression on the State-Trait
Depression Adjective Check List as measured by the Minnesota Multiphasic Personality
Inventory-D scale, the Beck Depression Inventory, and the Center for Epidemiologic Studies

Depression scale, Wm, 81, 826-831.

Lubin, B., Van Whitlock, R, Dale, T. M., Kelly, R. (1996). Further study of the
diagnostic and screening efﬁciency of the State-Trait Depression Adjective Check List.

Ioumalnﬂllinig .2519! 191:1 52, 26-31.

Luh, K. E., Redl, J., & Levy, J. (1994). Left- and right-handch see people
differently: Free-vision perceptual asymmetries for chimeric stimuli. Wan, 28,
141-160.

Maccoby, E., & Jacklin, C. (1974). W. Stanford.
Ca.: Stanford University Press.

McGlone, J. (1980). Sex diﬁ‘erences in human brain asymmetry. A critical survey.
Winners 3, 23 6-237

McGlone, J ., & Kertesz A (1973). Sex diﬁ‘erences in cerebral processing of visuo-
spatial tasks. 291102;, 2, 313-320.

McKeever, W. F., & Dixon, M. S. (1981). Right hemisphere superiority of
discriminating memorized ﬁ'om non-memorized faces: Aﬂ‘ect imagery, sex, and perceived
emotional eﬁ‘ects. W. 1.2, 246-260.

Mirowsky, J., & Ross, C. E. (1986). Social patterns of distress. W
5991919231, 12. 23-45-

Mirowsky, J., & Ross, C. E. (1995). Sex diﬁ‘erences in distress: Real or artifact.
WM 69, 449-468.

Moscovitch, M, & Olds, J. (1982). Asymmetry in spontaneous facial expression and
their possible relation to hemispheric specialization. mm 29, 71-81.

102

Natale, M., Our, R E., & Gur, R C. (1983). Hemispheric asymmetries in processing
emotional expressions. Ngumsychglggia, 21, 555-565.

Obrzut, J. E. (1983). The projective approach to personality assessment: An analysis
of the Thematic Picture technique. MW, 12, 414-420.

Pearlin, L. I. (1989). The sociological study of stress. WWW
Behaxiot. 3.0, 241-256-

Perria, P., Rossadini, G., & Rossi, G. F. (1961). Determination of side of cerebral
dominance with Amobarbital. Archimmfhleiimlm, A, 175-181.

Phillips, D. L., & Segal, B. (1969). Sexual status and psychiatric symptoms.
WM, 3.4, 58- 72

Rinn, W. E. (1984). The neuropsychology of facial expression: A review of
neurological and psychological mechanisms for producing facial expressions. 25291191931991

311119111125, 52- 77

Rizzolatti, G., & Buchtel, H. A. (1977). Hemispheric superiority in reaction time to
faces: A sex diﬁ‘erence. 291192;, 18, 300-305.

Rizzolatti, G., Umilta, C., & Beriucchi, G. (1971). Opposite superiority’s of the right
and left cerebral hemispheres in discriminative reaction time on physiognomical and
alphabetical material. Brain, 24., 431-442.

Robinson, R G. (1983). Mood changes in stroke patients: Relationship to lesion
location MW 211 555-566

Robinson, R G., & Price, T. R ( 1982). Post-stroke depressive disorders: A follow-
up study of 103 patients. Sanka, 13, 635-641.

Rosenﬁeld, S. (1980). Sex diﬁ‘erences in depression. Do women always have higher
rates? MW 21, 3342

Ross, C. E., & Huber, J. (1985). Hardship and depression. Wm
SmiaLBehaidor. 26, 312-327.

Rossadini, G. R, &Rossi, G. F. (1967). Experimentalanalysisofcerebral dominance

inman InD HMillikan&F L. Darley(Eds..) BrainMechanismsUnderlrdaaﬁamhand
14m. New York. Grune & Stratton.

Rubin, D. A, & Rubin, R T. (1980). Diﬂ‘erences in asymmetry of facial expression
between right- and left-handed children. WM 18, 373-3 77.

103

Sackeim, H. ,Greenberg, M. S. ,Weimen, A. L. Gur, R C., Hungerbuhler, J. P., &
Geschwind, N. (1982). Hemispheric asymmetry in the expression of positive and negative

emotions. Neurologic evidence. Aahixemﬂﬂgumlogy, 32, 210-218.

Sackeim, H. A, Gur, R. C., & Saucy, M. C. (1978). Emotions are expressed more
on the left side of the face, Sgienm, 202, 434-436.

Safer, M. A, & Leventhal, M. (1977). Bar diﬁ‘erences in evaluating tones of voices

and verbal content WM
29119111101993.7582.

Sclmc S., & Singer, 1 E. (1962). Cognitive, social, and physiological determinants
of emotional state. Wm, 62, 379-399.

Schiﬁ‘, B. B., & Lamon, M. (1989). Inducing emotion by unilateral contraction of
facial muscles: A new look at hemispheric specialization and the experience of emotion.

Neiimnmholosia. 22, 923-935.

Schiﬁ; B. B., & Lamon, M (1993). Unilateral contraction of facial muscles do effect
emotions: A "failed replication’s" failure to perform a replication. (291192;, 22, 549-551.

Schiﬁ‘, B. B., & Lamon, M. (1994). Inducing emotion by unilateral contraction of
hand muscles. Cums, 3_Q, 247-254.

Sergent, J. (1990). The neuropsychology of visual image generation: Data, method,
and theorY- WM 13., 98-129.

Sherman, J. A. (1976). Social values, femininity, and the development of female
competence loumalnfSnmallssuele. 181-195

Silberrnan, E. K., & Weingartner, H. (1986). Hemispheric lateralization of ﬁrnctions
related to emotions. Wu, 5, 322-353.

Soloman, D., & Ali, F. (1972). Age trends in the perception of verbal reinforcement.
Momentum 1, 238-243.

Springer, S. P., & Deutsch, G. (1993). Wain. New York: W. H.
Freeman & Co.

Stoppard, J. M, & Gruchy, C. D. (1993). Gender, context, and social expression of
positive emotion MW 12, 143-150

Strauss, E., & Kaplan, E. (1980). Lateralized asymmetries in self-perception.
Neumnmhologia. 16, 289-295.

104

Strauss, E., & Moscovitch, M. (1981). Perception of facial expression. Bainand
W. 13., 308-332.

Sue, S. (1987). The role of culture and cultural techniques in psychotherapy: A
critique and reformulation. Ameueanjsyehelggy, 32, 37-45.

Sue, S. (1994). Mental health research on Asian Americans. Special Issue: Asian-
American mental health. MW, 22, 61-67.

Takeuchui, D T., Sue, S., & Yeh, M. (1995). Return rates and outcomes from
ethnicity-speciﬁc mental health programs in Los Angeles. MW
86, 63 8-643.

Terzain, H. (1964). Behavioral and EEG eﬁ‘ects of intra carotid sodium amytal
injections. Winter... 12. 230-240.

Thaut, M. H., de l’Etoile, S. K. (1993). The effects of music on mood state-
dependent recall. W12! 29, 70-80

Tomkins, S. S. (1987). Shame. In D. L. Nathanson (Ed), IheMmEaeesgf
Shame. New York: Guilford Press.

Velton, E. (1968). A laboratory task for induction of mood states. Beheadeux
Ressamhandlhetapx, 6, 473-482.

Wagner, E. (1995). A logical analysis of projective technique based on independence
of items and latitude of responses. We, 81, (3 pt.-l), 868-870.

Wilkinson, L., Blank, G., & Gruber, C. (1996). Desk top data analysis with Systat:
regression on tirne-series data. Ch. 16. Upper River Valley: Prentis Hall

Wolpe, J. (1968). Psychophysiological correlates of systematic desensitization of
phobias. preliminary ﬁnds. Cunduimalﬁeﬂex, 8,139-.

Zaidel, E. (1975). A technique for presenting lateralized visual input with prolonged
exposure. 1609mm, 15, 283-289.

Zajonc, R B. (1980). Feeling and thinking: Preferences need no inferences.
Ameﬁganlsimholosist, 35, 151- 175

"1111111111111111