University of Southern California Dissertations and Theses

Content analysis of articulated thoughts of chronic worriers

(USC Thesis Other)

Content analysis of articulated thoughts of chronic worriers

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UMI films
the text directly from the original or copy submitted. Thus, some thesis and
dissertation copies are in typewriter face, while others may be from any type of
computer printer.
The quality of this reproduction is dependent upon the quality of the
copy subm itted. Broken or indistinct print, colored or poor quality illustrations
and photographs, print bleedthrough, substandard margins, and improper
alignment can adversely affect reproduction.
In the unlikely event that the author did not send UM I a complete manuscript
and there are missing pages, these w ill be noted. Also, if unauthorized
copyright material had to be removed, a note w ill indicate the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, beginning at the upper left-hand comer and continuing
from left to right in equal sections with small overlaps.
ProQuest Information and Learning
300 North Zeeb Road, Ann Arbor, M l 48106-1346 USA
800-521-0600
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CONTENT ANALYSIS OF ARTICULATED THOUGHTS OF CHRONIC
WORRIERS
by
Andrei David Sachs
A Thesis Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment o f the
Requirements for the Degree
MASTER OF ARTS
(PSYCHOLOGY)
May 2002
Copyright 2002 Andrei David Sachs
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UNIVERSITY OF SOUTHERN CALIFORNIA
T he G raduate School
U niversity Park
LOS ANGELES, CALIFORNIA 90089-1695
This thesis, w ritten b y
A noz& i S a C H $
U nder th e direction o f hJ5... Thesis
Com m ittee, an d approved b y a ll its m em bers,
has been p resen ted to a n d a ccep ted b y The
Graduate School, in p a rtia l fulfillm en t o f
requirem ents fo r th e degree o f
Dean o f Graduate Studies
D ate May 1 0 , 2002
THESIS COMMITTEE
Chairperson
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table of Contents
List of Tables..........................................................................................................iii
Abstract................................................................................................................. iii
Introduction.............................................................................................................. 1
Method..................................................................................................................... 8
Results.................................................................................................................... 23
Discussion..............................................................................................................35
Reference List....................................................................................................... 43
Appendices............................................................................................................47
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
List of Tables
Table I: Demographics and self-report ratings of
participants............................................................................................................ 12
Table 2: Pooled within-cell correlations for baseline
self-report data.......................................................................................................24
Table 3: Repeated self-report ratings of mood (PANAS)................................ 27
Table 4: ATSS Verbalization Ratings................................................................29
Table 5: Within-subjects effect sizes for repeated
self-report ratings o f mood (PANAS).................................................................33
Table 6: Pooled within-cell correlations for repeated
self-report ratings o f mood (PANAS).................................................................33
Table 7: Pooled within-cell correlations between ATSS
verbalization ratings of negative appraisal (NEG1/2) and
self-report ratings o f Negative Mood Activation (NA1-4).............................. 34
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
I
Introduction
The objective of the proposed research project was to examine interpretative
bias in people who report clinical-level symptoms of Generalized Anxiety Disorder
(analogue GAD).
Specifically, this project examined differences between analogue-GAD and
nonanxious controls (people who deny having GAD symptoms) in terms o f certain
cognitive and emotional functioning aspects. Appraisal, temporal, affective, and
attentional/cognitive features of think-aloud verbalizations were evaluated in
response to two content-relevant ambiguous stimuli. Participants were asked to
articulate their cognitions and emotions (i.e., their mental state) when confronted
with the stimuli via a laboratory procedure known as Articulated Thoughts in
Simulated Situations paradigm (ATSS) (Davison, Robins & Johnson, 1983). ATSS
allows for the assessment o f participants’ cognitions and emotions while they are
being experienced in experimenter-controlled, simulated situations.
The hypotheses o f the proposed study were as follows:
1. People with analogue-GAD status will construct more threatening interpretations
of the ambiguous narratives presented via ATSS.
2. The appraisal o f threat will be associated with the following cognitive and
affective phenomena:
a. GAD participants will exhibit more negative affect and less positive
affect.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2
b. People with GAD status will make fewer references to the present time
and more references to the future than nonanxious controls.
c. GAD participants will show less shifting from one topic to another topic
(a measure o f cognitive rigidity).
d. Cognitive distortions (e.g., should, must, etc.) will be greater for GAD
participants than for nonanxious controls.
BACKGROUND
Although in recent years researchers have started to learn about the nature
and content of worry (Roemer, Molina & Borkovec, 1997; Borkovec, 1994; Craske,
Rapee, Jackel & Barlow, 1989), much remains to be discovered about inner
experiences o f worry. Excessive and uncontrollable worry is the central feature of
Generalized Anxiety Disorder (Diagnostic and Statistical Manual o f Mental
Disorders-Fourth Edition, DSM-IV, 1994). Borkovec, Robinson, Pruzinsky &
Depree (1983) proposed a tentative definition of worry: “worry involves a chain of
thoughts and images, negatively affect-laden and relatively uncontrollable; it
represents an attempt to engage in mental-problem solving on an issue whose
outcome is uncertain but contains the possibility o f one or more negative outcomes;
consequently worry relates closely to the fear process.”
There is emerging evidence that cognitive and emotional interpretations of
anxious individuals seem to be biased in a relatively negative direction, and that they
show selective attention to threat information and construct a more threatening
meaning when appraising ambiguous information. The findings o f attentional and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3
interpretative biases play an important role in theories of etiology and maintenance
of anxiety disorders. For example, Mathews (1990) suggested that attentionai biases
to threat cues in the environment result in increased perception of danger and
therefore more frequent or intense experiences of anxiety. In a study where
participants were asked to read descriptions of an ambiguously-threatening event,
being interviewed for a job, Hirsch & Mathews (1997) found evidence for the notion
that nonanxious individuals infer positive outcomes to an ambiguous event, while
highly anxious individuals do not. The authors suggested that the lack o f protective
interpretative bias might be linked to a vulnerability to anxious mood. In a study of
the attentionai biases for emotional facial expressions, Bradley, Mogg, Falla &
Hamilton (1998) found evidence for an attentionai bias on the part o f high trait
anxious individuals to threatening facial expressions. Threat, happy, and neutral face
stimuli were presented, and an attentionai bias favoring threatening facial
expressions, but not emotional faces in general was found, in high trait anxiety.
These findings were subsequently replicated with a GAD clinical population
(Bradley, Mogg, White, Groom, & De Bono, 1999). According to the analysis of
Mogg & Bradley (1998), vulnerability to anxiety stems mainly from a lower
threshold for appraising threat, rather than a bias in the direction o f attention
deployment. Thus, relatively innocuous stimuli are evaluated as having higher
subjective threat value by high rather than low trait anxious individuals. In a recent
review o f cognitive theories in GAD, Aikins & Craske (in press) evaluated the
evidence for threat-related interpretation o f ambiguity and concluded, like others
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4
(MacLeod & Cohen, 1993), that the results so far are more likely due to
experimenter demand characteristics and response selection bias effects. Aikins &
Craske (in press) concluded, “conclusions regarding interpretations of ambiguity
remain very equivocal.”
The ATSS think-aloud paradigm allows for an assessment of differences in
the interpretation of ambiguity in a way that may not be compromised by the two
methodological problems just mentioned. First, the dependent measures of
interpretative bias and its correlates of affect and related cognitive features, are taken
entirely without the subjects’ knowledge. Second, the dependent variables that result
from ratings o f ATSS verbalizations are not based on subjects’ relative willingness
to endorse different response options, but they are derived from coding the streams
of consciousness produced by them. Given these methodological advantages, ATSS
is being applied in this experiment to address the hypothesis that chronic worriers
will construct more threatening meanings when exposed to ambiguous narratives.
While most findings of cognitive bias are based on trait anxious samples (MacLeod
& Cohen, 1993; Calvo & Castillo, 1997; Hirsh & Mathews, 1997; Byme &
MacLeod, 1997), the present study aims to extend the research on interpretative bias
in anxious people to a chronic worrier sample. To my knowledge, this was the first
attempt to use complex and ambiguous stimulus scenarios about events in real life,
with high external validity, in order to examine the presence o f interpretative bias in
an analogue-GAD population.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5
Many clinical researchers have proposed that cognitive biases play a central
role in the development and maintenance of emotional disorders and therefore are
associated with emotional vulnerabilities and disturbances, such as anxiety (Beck &
Clark, 1997) and worry (Andrews & Borkovec, 1988). Others have suggested that
when an ambiguous situation is interpreted in a threatening manner, heightened state
anxiety is likely to be the emotional consequence (MacLeod & Cohen, 1993). In
accord, some emotion theorists (e.g., Lazarus, 1984; Smith & Ellsworth, 1985;
Frijda, Kuipers & ter Schure, 1989) view the appraisal process as an antecedent of
the emotional response. Frijda et al. (1989) regard emotion as an “experience of
forms of appraisal”. However, other emotion theorists (Zajonc, 1984) continue to
support the primacy and independence of affect. In this study, it was expected that
high trait worry participants would experience and therefore verbalize more negative
affect and less positive affect than nonanxious control participants. The negative and
positive affect were measured according to the Watson & Tellegen (1985) two-factor
model o f affective structure.
Consistent with Ellis’s (1969) and Beck’s (1976) theories o f anxiety and
other emotional disturbances, more recent empirical findings have emphasized the
role o f cognitive distortions in worry and anxiety (Ingram & Kendall, 1987; Ingram,
Kendall, Smith & Donnell, 1987). In an analysis o f streams o f consciousness of
GAD participants, Molina, Borkovec, Peasley & Person (1998) found support for the
prevalence o f cognitive distortions in the content o f worrisome streams compared to
neutral streams. In accord, it was hypothesized in this study that absolute words
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6
{always, never), words suggesting catastrophic interpretations {awful, terrible,
horrible, what if..), and the use of rigid rules and commands in daily living {should,
must, have to) will be more prevalent in the high trait worry group. Prior research
has shown that the majority of worrisome content (76.4%) refers to the future or the
past (Borkovec et al., 1983). High trait worry individuals were expected to make
fewer present-oriented and more future-oriented statements than control participants.
Most of what we know today about the content and nature o f worry has been
derived either by clinical interviews or post hoc subjective self-report measures.
Self-report measures are also referred to as endorsement methods o f cognitive
assessment or self-statement inventories by some researchers (Haaga, 1997). Such
self-statement inventories require respondents to endorse pre-selected thoughts
presented by the experimenter. Even though questionnaires have shown good
psychometric properties of criterion and discriminant validity (Glass & Amkoff,
1997), concerns remain regarding content and convergent validity. For example, an
item on an inventory may be endorsed, not because the respondent had the thought
but because the participant agrees with it. Additionally, most questionnaires refer to
how thoughts have occurred in the past and they are dependent on recall and subject
to any biases imposed on that recall (Haaga, 1997).
With the limitations of endorsement methods in mind, researchers have
developed complimentary means of assessing cognitions. One such alternative
method o f assessing cognitive products and processes is the think aloud approach,
which does not offer pre-selected thought examples (Chamberlain & Haaga, 1999).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7
In this method, cognitions are assessed concurrently with their occurrence, as
participants are asked to verbalize their thoughts as they occur (Davison, Vogel &
Coffman, 1997). In one example of the use of the think aloud method developed by
Pennebaker, Czajka, Cropanzano & Richards (1990) for the study of worry,
participants were first instructed to simply describe their ‘stream of consciousness’,
or ‘to worry in your usual way’ and then ‘do the streaming’ (Molina et al., 1998).
Unlike the Pennebaker et al. (1990) approach, Articulated Thoughts in Simulated
Situations (ATSS) (Davison et al., 1983) is a think aloud approach to cognitive
assessment that affords experimental control over the stimulus situation. So in other
words, with ATSS there is the opportunity to expose all participants to the same
stimulus, instead of asking the participants to choose a topic to think about in the
moment according to some criteria (e.g., worry). With ATSS, the stimulus is
presented on audiotape with instructions for the participant to imagine himself or
herself in the situation, and then think aloud after each brief interval of the situation.
The ATSS paradigm is a think-aloud approach that enables researchers to
assess participants’ cognitions “on-line” and incorporates an unstructured response
format. It has been successfully employed in the study of a variety of psychological
issues, including marital and family conflict (Eckhardt, Barbour & Davison, 1998),
aggression (Williams, Davison, Nezami &DeQuattro, 1992), anxiety and depression,
(White, Davison, Haaga & Kerrin, 1992) and smoking cessation (Haaga, Davison,
McDermut & Hillis, 1993). In the ATSS procedure participants are presented with
audio taped scenarios, which serve as the stimulus. Each scenario is typically split
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8
up into several segments, each about ten to twenty seconds long and each followed
by a block of time (30 seconds) for participants to verbalize their thoughts.
Participants are instructed to imagine themselves in the situation presented on tape
and to articulate their thoughts during the segment breaks. Their articulated thoughts
are recorded and later transcribed for content analysis.
SIGNIFICANCE OF THE PROPOSED STUDY
The current study proposes to examine the presence o f interpretative bias in
the articulated thoughts of analogue general anxiety disorder participants. Cognitive
processing of high worriers is important to study because worry processes have been
hypothesized to underlie many disorders of emotion (Brown, Antony, & Barlow,
1992). Such processes are present not only in GAD, but in other anxiety disorders
and depression as well. The current study extends prior research in interpretative bias
in anxiety disorders in a number of ways. First, the use o f a complex stimulus
scenario as a standardized stimulus, which was designed based on findings from
content of worry (Roemer at al., 1997) and the ultimate object o f worry research
(Hazlett & Craske, in preparation) is new. Second, the use o f a think aloud protocol
(ATSS) to measure “mental state” is new in GAD research as well. Only one prior
study (Molina et al., 1998) employed such a protocol, however in more limited
context. Third, the significance of valid assessment o f cognition for the ongoing
development o f CBT models o f psychopathology has been endorsed by many
researchers (Haaga, 1997; Clark, 1997). A think aloud production method, such as
ATSS (Davison et al., 1983), with its experimental control over stimulus situations
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9
and an open ended response format, may be helpful in revealing more about the inner
experiences of chronic worriers.
METHOD
Experimental Design
This study employs a 2 X 2 mixed-subject design. Participants were grouped
into chronic worriers and nonanxious controls on the basis o f self-report data
collected with the Generalized Anxiety Disorder Questionnaire (GAD-Q-IV;
Newman, Zuellig, Kachin, Constantino & Cashman, in press) in the pre-testing
process. The within-subjects experimental variable was type o f ATSS audio
narrative, with two conditions. In the first condition, all participants were exposed to
an ambiguous social situation scenario, in which both threatening and nonthreatening
interpretations were possible. In the second condition, participants were exposed to
an ambiguous scenario that tapped into thoughts and beliefs about academic life and
achievement. This second scenario was designed to be non-threatening to the
research participants. Only one measure, the self-report measure of mood, was
administered as a repeated measure four different times. All other self-report
measures were administered once.
In this study, all participants (i.e., GADs and controls) were exposed to the
same two stimulus scenarios. This choice was appropriate based on prior findings
about the content of worry in GAD, which suggest that chronic and normative worry
are different in uncontrollability and excessiveness, but not in content. Roemer et al.
(1997) found the same worry themes, family/interpersonal and achievement, to be
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10
predominant for both GAD and for nonanxious controls. Their worry processes are
different, though, in intrusiveness, uncontrollability and duration (i.e., time spent
worrying). Both the Roemer et al. (1997) and this study relied on a college sample.
The first stimulus scenario (the social situation scenario) contains a complex
and ambiguous situation about the possibility o f making friends when attending a
social event on campus. The topic for this scenario, making friends, was selected
based on an analysis of the data set from the Hazlett & Craske (in preparation) study.
Hazlett & Craske (in preparation) examined feared outcomes underlying the worries
of GAD and nonanxious college students The results of the analysis of their data
suggested that worrying about making friends was the most frequent worry topic
within the social functioning domain. In other words, when asked, “what do you
worry about when you worry about social relations”, more participants selected
“making friends” than any other topic. In this ambiguous scenario, both threatening
and nonthreatening interpretations are possible. Specifically, when presented with
the imaginal opportunity to make friends, some may welcome it while others may
fear it. The second stimulus scenario (the unknown professor scenario) presents a
topic related to taking a class and the quality o f instruction received, and it may have
tapped into cognitions related to academic achievement. Achievement was found to
be the second most frequent topic of worry (Roemer et al., 1997).
Grey & Mathews (2000) found that training in unselected volunteers can
induce interpretative biases, and that prior exposure to threatening information
resulted in relative faster access to congruent meanings. Although the two
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
11
experimental narratives used in this study were carefully designed to be ambiguous
and therefore they do not contain overtly threatening information, the possibility
exists that a training effect may influence the responses to the narrative which is
presented second. Given that the study hypotheses are focused on the social
situation scenario, the two narratives were not counterbalanced in this design and the
social scenario was presented first to study participants. This design eliminated the
possibility that the responses to the social scenario may be confounded by responses
to the other scenario. No specific hypotheses were made regarding the responses to
the second narrative and they are being treated strictly treated as exploratory
research.
Participants
A total o f 89 participants were recruited to participate in this study, 44
analogue GADs (35 females), and 45 nonanxious controls (27 females). Fifteen
hundred copies of the GAD version o f the recruitment flyer and an equal number of
control flyers were distributed across campus. Flyer respondents were administered
the GAD-Q-IV as a phone screen by either the principal investigator or the project
coordinator. A total of 35 paid participants were recruited via a USC campus-wide
flyer and were paid $ 12 each for their participation. O f these 35 participants, 19
qualified for the GAD group and 16 for the control group. A total o f 54 subjects
participated in exchange for class credit, 25 in the GAD group and 29 in the control
group. These 54 participants were males and females who were administered the
Generalized Anxiety Disorder Questionnaire (GAD-Q-IV; Newman, Zuellig,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
12
Kachin, Constantino & Cashman, in press) survey measure during in class pre
testing. All participants (i.e., both paid and for class credit) were included if they
self-reported symptoms o f generalized anxiety disorder (i.e., current uncontrollable
and excessive worry and three symptoms such as restlessness, difficulty
concentrating, sleep disturbance, irritability, fatigue, chronic muscle tension) or
denied meeting any diagnostic criteria for generalized anxiety disorder (nonanxious
group) according to their responses on the GAD-Q-IV. There were no inclusion or
exclusion criteria that are based on gender or racial/ethnic origin. The ethnic mix was
as follows: Caucasian: 43.8%, Asian-American: 20.2%, Latino-American: 19.1%,
African-American: 6.7% and other: 6.7%. The paid GAD participants had a mean
age of 34.94 (SD=12.16), while the GAD subjects who participated for class credit
had a mean age of 20.4 (SD=2.04). The paid control participants had a mean age of
24.56 (SD=7.8), while the control subjects who participated for class credit had a
mean age o f 20.00 (SD=1.16). The demographics and baseline characteristics are
presented in Table 1.
Table 1. Demographics and self-report ratings of participants
GAD
(n = 41)
M/(SD)
Control
(n = 44)
M/(SD)
Age 26.02/(10.33) 21.68/(5.2)
% Female 80.5% 59.1%
% Caucasian 41.5% 45.5%
PSWQ 62.12/(9.31) 41.29/(11.98)
BIS 21.25/(2.72) 17.09/(2.94)
GAD-Q-IV sum score 25.02/(2.95) 2.15/(2.31)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
13
Table 1. Continued
BDI 16.8/(9.75) 5.97/(4.16)
FNE 34.61/(6.66) 29.47/(5.97)
SUDS 35.36/(27.02) 13.79/(17.5)
[US 75.81/(19.40) 51.58/(14.02)
Self-report measures
I) Generalized Anxiety Disorder Questionnaire (GAD-Q-IV; Newman, Zuellig,
Kachin, Constantino & Cashman, in press). The GAD-Q-IV was designed for
use as an initial screen to diagnose Generalized Anxiety Disorder (GAD)
based on DSM-IV criteria and as a measure to compare symptomatology
across groups. Diagnoses made by GAD-Q-IV were found to have a 17%
false positive rate and a 7.6% false negative rate (Newman et al, in press).
Participants were said to meet GAD criteria on the GAD-Q-IV if they
indicated: (1) having experienced excessive and uncontrollable worry (i.e.,
they answered yes to item I or item 4, or the combination of items 2 and 3, or
any other combination of the 1st four items that indicated that the individual
worried excessively and uncontrollably), (2) the worry occurred more days
than not for at least six months (i.e., they answered yes to item 6), (3) they
worried about a number o f events or activities (i.e., they indicated three or
more worry topics on item 5), (4) they reported experiencing three or more o f
six symptoms during the past six months (i.e., they checked off three or more
symptoms in response to item 7) and the symptoms have caused at least
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
14
moderate distress or moderate impairment (i.e., they scored 4 or more on item
8 or 9). The GAD-Q-IV was scored by creating a sum total score. To create
a total score for each participant, all yes answers were coded as 1 and all no
answers as 0 (e.g. items 1-4, & 6). In addition, individuals were given 1
point for each individual worry topic listed up to 6. Further, participants
were given I point for each physical symptom they experienced up to six.
Finally, the numbers circled for items 8 and 9 (i.e., degree o f distress and
interference) were added to the sum of the other items. Because the
questionnaire requests that individuals skip the remaining items when they do
not endorse initial criteria, such skipped items were scored as 0. In the
Newman et al. study, the mean score for the nonclinical sample was M = 5.27
(SD = 6.87) and the mean score for the GAD sample was M = 20.07 (SD =
7.73).
2) Penn State Worry Questionnaire (PSWQ; Meyer, Miller, Metzger &
Borkovec (1990). This is an instrument designed to provide a trait assessment
of pathological worry; its 16 items capture some of the important features of
clinically relevant worry, namely the (1) generality o f worry over time and
situations, the (2) intensity/excessiveness o f worry, and the (3)
uncontrollability o f worry. Results of psychometric analysis (Brown et al.,
1992) show that nonanxious controls have a mean o f 35, while the clinical
GAD mean was 68. A score of 56 has been used as a ‘cut-ofF for high
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
15
worriers. Other studies (Hazlett & Craske, in preparation) have shown that
analogue GADs (i.e., GAD by GAD-Q) have a mean score of 65 on PSWQ.
3) Behavioral Inhibition System Scale/Behavioral Approach System (BIS/BAS;
Carver & White, 1994). This measure has two scales. The BIS scale is
designed to assess dispositional behavioral inhibition, which regulates
aversive motivation, while the BAS scale is designed to assess individual
differences in the sensitivity o f regulation of appetitive motives, in which the
goal is to move toward something desired. A behavioral avoidance (or
inhibition) system (BIS) is said to regulate aversive motives, in which the
goal is to move away from something unpleasant. The BIS is sensitive to
signals of punishment, nonreward, and novelty. Greater BIS sensitivity
should be reflected in greater proneness to anxiety. Items other than 2 and 22
are reverse-scored. BAS Drive: 3,9, 12, 21; BAS Fun Seeking: 5, 10, 15,
20; BAS Reward Responsiveness: 4, 7, 14, 18, 23. BIS: 2, 8,13,16,19,22,
24. Items 1, 6, 11, 17, are fillers.
4) The Brief Fear o f Negative Evaluation Scale (FNES; Watson & Friend, 1969)
(Appendix D). The FNE is a 30-item scale measuring apprehension about
other's evaluations. This measure has demonstrated high test-retest reliability
and high known groups and criterion validity.
5) Demographics Questionnaire. Participants were asked to provide certain
personal information variables including gender, age, race/ethnicity, year in
college.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6) The Positive and Negative Affect Schedule (PANAS; Watson & Tellegen,
1998). This is a 20-item scale that measures the degree of positive and
negative affect (high activation) within an experimentally defined time
period. In this study the PANAS was administered as a state measure. The
internal consistencies for the positive and negative subscales range from .85-
.90 and test-retest reliabilities range from .42-.68 depending on the timeframe
reference. Ten items form the positive activation (PA) scale and the
remaining 10 form the negative activation (NA) scale. This measure was
administered four times during the course o f the experiment.
7) Beck Depression Inventory (BDI; Beck, Ward, Mendelson, Mock &
Erbaugh, 1961). The BDI contains 21 items rated from 0 to 3 in terms o f the
intensity o f depressive symptoms. The measure has high internal consistency
in psychiatric and nonpsychiatric samples and adequate concurrent and
discriminant validity (Beck, Steer & Garbin, 1988). Prior psychometric
research (based on an out-patient sample) has shown that while the GAD
mean was 14.46 (SD = 6.10), the depression mean score was 26.37 (SD =
6.94).
8) The Intolerance o f Uncertainty Scale (IU; Freeston, Rheaume, Letarte &
Dugas (1994). The IU measures emotional and behavioral reactions to
ambiguous situations, the implications o f being uncertain, and attempts to
control the future. Results o f psychometric analysis show that nonanxious
controls have a mean score o f 43.8 (SD = 10.8) and clinical GAD participants
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
17
have a mean score o f 81.3 (SD = 24.5). The scale contains 27 items and has
been shown to have good internal consistency and test-retest reliability after a
five-week period as well as good convergent validity.
9) Subjective units o f distress (SUDS). Participant rates on 0-100 scale.
Procedure
The study was conducted in one 60-65 minute session. Upon arrival at the
laboratory, participants were asked to read and sign an informed consent form.
Second, they filled out copies of BIS/BAS, FNES, BDI, PSWQ, PANAS I (i.e.,
PANAS at Time 1) and demographics questionnaire and rated their SUDS. Third,
participants listened to a classical 70-second classical music tape. The purpose o f the
music tape was to help induce a neutral mood state. Fourth, just prior to the
presentation of the first ATSS vignette, they filled out copies of PANAS2 (i.e,
PANAS at Time 2). Fifth, the experimenter provided a brief overview of the ATSS
procedure and answered any questions or concerns that were raised. Then the
participants were introduced to the ATSS procedure via tape-recorded instructions.
The instructions let the participants know that the researchers are interested in
people’s thoughts. They were asked to listen to two tape-recorded scenarios (see
Appendix L and M for scenario transcripts) and to imagine themselves actually being
involved in the depicted situations. The described goal was to tune into what is on
their minds and say these thoughts out loud in as much detail as possible.
Following the instructions, the participant listened to a practice tape (step five
in the procedure), after which any remaining questions or concerns were addressed.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
18
After ensuring that the participant fully understood the procedure, the ATSS
procedure began and the experimenter left the room to ensure privacy and the
likelihood o f uncensored responding. During the ATSS procedure each participant
listened to (a) social situation scenario (step six in the procedure), and (b) the
unknown professor scenario (step eight in the procedure). In step seven, participants
filled out copies of PANAS3 (i.e., PANAS at Time 3).
(1) Scenario about a social/interpersonal situation: The data from the Hazlett
& Craske (in preparation) study was used to select the social/interpersonal theme of
the worry induction scenario. Responses to “What worries you about social
relations?” were coded into categories (romantic, family, friends, negative
evaluation, fear of dependency, fear of rejection, loneliness, fear o f harming others,
fear of initial contact, roommate, lack of trust, discrimination, violence and personal
safely). Worrying about having friends was the most frequent topic o f social worry
(28 out of 63, or 43%). The social situation scenario presents the possibility of
making friends at a social function on campus and creates the opportunity for several
interpretations, positive and worrisome ones. An optimistic interpretation (i.e.,
positive appraisal) may be that the person is looking forward to the event, hopes to
meet interesting people who subsequently may develop into friends. A negative
interpretation (i.e., negative appraisal) may be that the person won’t meet anybody
interesting, that nobody will be interested in them, that any interactions at the event
may not lead to further interactions and friendships, that they will end up alone and
lonely. (Appendix A)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
19
(21 The unknown professor scenario: In this scenario the participant was
asked to imagine that he or she overhears a discussion between two fellow students
about a boring professor. This scenario has been used in other studies in Dr.
Davison’s lab as a control scenario. In the case of a GAD sample, topics related to
academic achievement may have very well triggered worry and negative affect.
Overhearing a conversation about a boring professor, may, however, only remotely
bring up ideas of participant’s own academic life. This scenario can be
conceptualized as being an ambiguous stimulus as well, on the theme of academic
performance, which only in a more indirect way may have tapped into worries
related to academic achievement. (Appendix B).
After the ATSS procedure, participants filled out the PANAS4 and responses
to the manipulation check questions (step nine) and were played the same classical
music tape again in an effort to diminish any negative affect that may be generated
by the ATSS scenarios (step ten). Data from production approaches may have
meaning ascribed to them that participants did not intend (Chamberlain & Haaga,
1999). In order to get the participants’ own assessment o f their response, subjects
were asked if they worried as a result o f being exposed to the two scenarios.
Debriefing (step eleven) All participants were thoroughly debriefed at the end
of the procedure. They were reassured that all the scenarios were entirely fictional
and the experimenter answered any questions they may have had about the study. In
addition, referral to the USC Student Counseling Center was offered to all
participants.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ATSS Verbalization Ratings
The social situation and unknown professor audiotapes were transcribed and
divided into individual statements for each of the seven segments respectively, and
for each o f the two scenarios. Statement counts per scenario for each participant
were obtained to facilitate the subsequent creation o f proportions. Each individual
statement was rated independently by two raters in order to extract the following
dependent variables: 1) appraisal (negative or positive), 2) mood (high and low
positive and negative activation), 3) temporal orientation (past, present, future) and
4) degree of shifting from one topic to another. Appraisal was classified as negative
if the presence o f threat or negative expectancy was detected, or positive if the
presence o f positive expectancy was detected.
Mood was classified into high-arousal negative affect (high-NA) (e.g.,
feeling fearful) or low-arousal negative affect (low-NA) (e.g., feeling calm and
relaxed) and high-arousal positive affect (high-PA) (e.g., feeling enthusiastic and
excited) or low-arousal positive affect (low-PA) (e.g., feeling fatigued and sluggish)
categories, consistent with the Watson & Tellegen (1985) two factor structure of
affect. In other words, each statement received an NA and a PA rating or neither.
Temporal orientation was rated as past, present or future on the basis o f the
tense o f each verb found in each statement.
The degree of shifting was coded at the segment level instead o f at the
statement level. Two raters indicated the extent (i.e., intensity) to which an entire
segment involved shifting from one topic to another, using a 5-point Likert scale (1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
21
no shifting, 3 = some shifting, 5 = a lot of shifting), separately for the social and
unknown professor scenarios.
The raters, which were blind to group status, coded the social situation and
unknown professor transcripts. The unit of coding and the unit of analysis differed
by rating category. For appraisal and mood, each statement was first rated as either
present or absent, then frequency ratings were summed up for each transcript. In
other words, the appraisal and mood variables that were entered into the analysis
were based on sums o f statement frequencies for a given scenario. For temporal
orientation, frequency ratings of each verb were summed up for each transcript. For
degree of shifting, segment ratings were also summed for each transcript. In other
words, while the unit o f coding varied by rating category (statement, verb or entire
segment), the unit of analysis was always at the scenario level.
Intraclass correlation coefficients (ICCs) (Shrout & Fleiss, 1979; Whitehurst,
1984; McGraw & Wong, 1996) were calculated for each rating category. The ICCs
were .91 and .92 for positive and negative appraisal for the social situation scenario
and .80 and .89 for positive and negative appraisal for the unknown professor
scenario. The ICCs were .94, .86, .88 and .94 for high and low positive activation
(PA) and negative activation (NA) for the social situation scenario and the ICCs
were .93, .80, .85 and .92 for high and low PA and NA for the unknown professor
scenario. Two out o f three temporal orientation ratings showed acceptable interrater
agreement levels: present, .96 and .98 for scenarios 1 and 2 respectively, and future,
.84 and .82 for scenarios I and 2 respectively; only the past ratings failed to achieve
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
22
an acceptable ICC above .80. Difficulty in obtaining acceptable levels of interrater
agreement for the past rating was, in part due to its very low frequency of occurrence
compared to the other two. As a result, the past rating was dropped from the main
analysis. The ICC for degree of shifting was also below the .80 acceptable level for
both scenarios. Difficulty in obtaining acceptable levels of interrater agreement for
the degree of shifting rating was, in part due to the fact that the coding system relied
on an intensity scale and a molar approach by selecting a whole segment (instead of
a statement) as the unit o f coding. This may have resulted in too much subjectivity
on the part of raters. The ratings for degree of shifting were also dropped from the
main analysis.
Means of the raters’ judgments were used in subsequent analyses for all the
ratings that achieved ICC greater than .80. Finally, to control for individual
differences in total number o f statements (i.e., degree of verbalization), proportions
o f each category (relative to total number o f statements made) were calculated for
each scenario (e.g., negative appraisal proportion = sum of negative appraisal
statements in a given scenario/total number of statements in the same scenario).
These proportions served as dependent variables in the analyses for the ratings of
appraisal and mood. For temporal orientation ratings, the means o f the rater’s
judgments served as dependent variables in the analysis. Proportions were not
created for temporal orientation, as statements include sometimes multiple verbs.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
23
Computer Word Search
Words referring to potential cognitive distortions were tallied and categorized
in three groups: I) rigid rules {have to/should), absoluteness {always/never), and
catastrophe {awful/terrible/what if) in a similar manner to the one used by Molina et
al. (1998) (see Appendix V).
Results
Preliminary analyses
An analysis o f variance on the effect o f group (GAD vs. control) and
recruitment method (pay vs. class credit) on participant age revealed a significant
effect of group, F = 14.1, p < .001, a significant effect of recruitment method, F =
44.55, g < .001 and a significant interaction effect between group and recruitment
method, F = 12.1, p < .001. A subsequent t-test for the effect of recruitment method
on participant age for the GAD participant revealed a significant effect, t = 5.89, p <
.001. A similar test for the effect o f recruitment method on participant age for the
control participants, t = 3.1,p< .001. A t-test for the effect o f group status on age
for paid participants revealed a significant effect, t = 2.9, p < .001. The mean age o f
the paid GAD participants was 34.94, while the mean age o f the class credit GAD
participants was 20.4. The mean age of the paid control participants was 24.56, while
the mean age of the class credit control participants was 20. In order to determine if
recruitment method should be carried forward in the analysis in addition to group
status, two multivariate analyses o f variance (MANOVA) were done with
recruitment as the IV for the GAD and control groups respectively on clinical
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
24
measures o f importance, GAD-Q-IV sum total score, PSWQ, BIS, FNES, Positive
Activation (PA) and Negative Activation (NA). The MANOVA for the GAD group
did not show a multivariate effect on these dependent variables (DVs) of clinical
interest. Similarly the MANOVA for the control group did not reveal any
multivariate significance. As a result, paid and class credit participants were merged
into their respective GAD and control groups and recruitment method was not
carried forward in the analysis as a separate independent variable (IV).
Since group assignment was done on the basis GAD self-report measure of
symptomatology, a MANOVA was carried out to test the effect o f group on
additional measures of anxiety, specifically PSWQ (trait worry), IUS (intolerance of
uncertainty) and BIS (a measure of trait anxiety). These three measures are highly
correlated with each other, suggesting good convergent validity, as (see Table 2).
PSWQ GAD.SUM BIS BDI FNES IUS
PSWQ
GAD.SUM .370
BIS .721 .333
BDI .384 .203 .231
FNES .489 .223 .469 .390
IUS .583 .141 .442 .510 .451
GAD and nonanxious population norms have previously been established for
PSWQ and IUS. In this study the obtained PSWQ means o f 62.12 and 41.29 (see
Table 1) for the GAD and control groups compare favorably with reported norms of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
25
63.24 and 44.27 for GAD and nonanxious controls. Similarly, the obtained IUS
means o f 75.81 and 51.78 for GADs and controls compared well with norms for the
two groups of 81.3 and 43.8. Trait anxiety was included in this analysis based on
prior research, which showed that trait anxiety is a related (but different) construct to
worry (Molina & Borkovec, 1994). The MANOVA showed a multivariate
significant effect of group, F = 426, p < .001. As a result, we concluded that the
groups were formed as intended. The effect o f group on level of verbalization was
tested for the two ATSS scenarios. Verbalization was operationalized as the overall
word count produced in response to the first and second stimulus scenarios
respectively. T-tests showed no significant differences between the groups on word
count.
Research findings have consistently shown that social phobia and depression
are frequent comorbid conditions with GAD. As a result, a MANOVA explored the
effect of group on the measures of social phobia and depression, FNES and BDI and
found a significant multivariate effect, F = 22.06, p < .001. Subsequently Roy-
Bargman Stepdown F-tests were interpreted instead of univariate F-tests (Tabachnick
& Fidell., 1996) due to the size o f the pooled-within group correlation between these
two measures (.368). In other words, the size of the shared variance between FNES
and BDI (more than 10%) required that stepdown, instead o f univariate results be
interpreted. When BDI was entered first in the model, we found stepdown
significance for it, F = 43, p < .001, but not for FNES. In a second analysis when
FNES was entered first, both FNES and BDI were significant, F = 12.95, p < .001
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
26
and F = 27.26, p < .001 respectively. As a result, only BDI was carried forward as
potential covariate in subsequent analyses, as its unique variance and the variance
shared with FNES were related to the significant group differences. However, the
obtained BDI mean score o f 16.8 (see Table 1) is more closely aligned with the GAD
norm o f 14.46 versus the depression norm of 26.37. Furthermore, a frequency
analysis o f BDI scores in the GAD group revealed that 88% o f the GAD group
participants had a mean BDI score less than the depression mean of 26. In other
words, less than 15% o f the GAD participants had BDI scores at or above the
depression mean.
Screening o f Self-Report Data
Prior to the analysis, trait worry (PSWQ), social phobia (FNES), trait anxiety
(BIS), intolerance o f uncertainty (IUS), depression (BDI), positive mood activation
at times 1 though 4 (PA1, PA2, PA3, PA4), negative mood activation at times 1
through 4 (NA1, NA2, NA3, NA4) were examined for missing values and fit
between their distributions and the assumptions o f univariate and multivariate
analyses.
Two cases missed one item each on IUS and one case missed one item on
PA4. For these three cases the mean was substituted for the missing response. One
case refused to respond to item 9 (suicidal ideation) on the BDI and at the end o f the
experiment seemed agitated and requested the phone number for the student-
counseling center. For this case the value 2 was substituted for the missing response,
assuming suicidal ideation.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
27
One case was a univariate outlier on BDI (z = 3.48). This score was left
unchanged. One case was a univariate outlier on NA4 (z = 5.03) and another was a
univariate outlier on NA3 (z = 3.87). The Windsor method (Guttman, 1973) was
applied to these outliers on the NA4 and NA3 measures, converting them to one unit
above the next closest unit. By using Mahalanobis distance with p < .001, no
multivariate outliers were identified. The data were not modified as a result of the
modest violations of skewness and kurtosis for NA3 and NA4, which were found.
The means and standard deviations for self-report measures are reported in Table 1
and 3.
A review of the standardized values associated with the PANAS scale at
times 1-4 revealed two univariate outliers. NA2, NA3 and NA4 in the control group
showed moderate skewness (z = 4.5-5.0) and NA4 also showed a violation of
kurtosis (z = 5).
Table 3. Repeated self-report ratings of mood (PANAS
GAD
(n = 41)
M/(SD)
Control
(n = 44)
M/(SD)
Effect sizes
between subjects
(Cohen’s d)
Time 1
PA1 23.87/(7.0) 27.97/(8.98) .51*
NA1 19.36/(7.88) 13.68/(3.12) 1.03***
Time 2
PA2 23.17/(8.42) 26.56/(9.43) .38'
NA2 16.14/(6.31) 12.31/(3.13) .81***
Time 3
PA3 21.82/(7.88) 26.04/(9.88) .47*
NA3 18.09/(6.57) 13.70/(4.77) .77***
Time 4
PA4 21.96/(9.16) 25.20/(9.40) .34
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
28
Table 3. Continued
NA4 16.65/(5.97) 12.13/(2.33) 1.08***
Note: * = e < -05; t = trend, p < .10; *** = jj < .001
Screening o f ATSS Verbalization Ratines
Prior to the analysis, ratings o f positive and negative appraisal, mood (high
and low PA and NA), temporal orientation (past, present, future), degree of shifting
and cognitive distortions (rules, absoluteness and catastrophe) were examined for
missing values and fit between their distributions and the assumptions of univariate
and multivariate analyses.
There were no missing data. Each of the three out of the four ratings of
appraisal had one case of a univariate outlier. The Windsor method (Guttman, 1973)
was applied to these outliers converting them to one unit (.01) above the next closest
unit. One rating, the proportion of positive appraisals in response to scenario 2, in
the GAD group had a violation of skewness (z = 4.74). Due to the adequate sample
size (n = 44 for GAD group), the impact of departure from zero skewness is
diminished and no transformation was performed (Tabachnick & Fidell, 2001).
The ratings of high and low PA for scenario 2 had a violation o f skewness (z = 4.41).
Each o f these ratings of mood had one case of a univariate outlier. The Windsor
method (Guttman, 1973) was applied to these outliers converting them to one unit
(.01) above the next closest unit. O f the four ratings o f temporal orientation (present
and future for scenarios 1 and 2), the rating of future for scenario 2 showed a
violation o f kurtosis (z = 6.26) and skewness (z = 5.47). Univariate outliers were
again windsorized. Ratings of cognitive distortions done with computer word search
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
29
showed a very low base rate o f the phenomenon o f interest. As a result these ratings
were dropped from the analysis. By using Mahalanobis distance with p < .001, four
cases were identified as multivariate outliers with g < .001. All four cases were
deleted. The means, standard deviations and effect sizes for all ATSS verbalization
ratings are reported in Table 4.
Table 4. ATSS Verbalization Ratings
GAD
(n = 41)
M/(SD)
Control
(n = 44)
M/(SD)
Effect size
(Cohen’s d)
Appraisal
Positive/Scenario 1
(Posl)
.1442/(.1061) .1763/. 1257) .3
Negative/Scenario I
(Nesl)
.1119/(.09877) .06477/( .06942) .56**
Positive/Scenario 2
(Pos2)
.03002/(.04012) .02177/(.02239) .26
Negative/Scenario 2
(Neg2)
.07606/(.0699) .05479/(.06067) .31
Affect
High PA/Scenario 1
(HiPAl)
.08907/(.06929) .1102/(.08179) .28
High NA/Scenario 1
(HiNAl)
.06310/( .04502) .05000/(.04965) .28
Low PA/Scenario 1
(LoPAl)
.06267/(.045l5) .05000/(.04965) .26
Low NA/Scenario 1
CLoNAl)
.08888/(.0694) .1099/(.08l74) .27
High PA/Scenario 2
(HiPA2)
.03877/(.04398) .06175/(.0541) . 52*
High NA/Scenario 2
CHiNA2)
.06901/(.7037) .06739/(. 6592) Nill
Low PA/Scenario 2
(LoPA2)
.06721/(.07367) .066l8/(.06676) Nill
Low NA/Scenario 2
(LoNA2)
.03877/(.04398) .06175/(.05410) . 52*
Temporal Orientation
Present/Scenario 1
(Presentl)
23.25/(9.29) 24.34/(9.89) Nill
Present/Scenario 2
(Present2)
27.13/(10.00) 28.43/(11.62) Nill
Future/Scenario 1
(Futurel)
7.89/(5.24) 5.58/(3.89) . 51*
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
30
Table 4. Continued
Future/Scenario 2 4.13/(2.65) 3.13/(3.14) .34 j
(Future2) |
Note: * = p < .05; ** = p < .01.
Major analyses
To evaluate the association of GAD group status with positive and negative
appraisal for each of the two stimulus scenarios, we conducted t-tests for each
measure of appraisal comparing the appraisals in the GAD group with those in the
nonanxious control group. Multivariate tests were not appropriate given the low
correlations between positive and negative appraisals (POS1/2; NOG1/2) in the two
scenarios.
Univariate effects were significant only for negative appraisal in scenario 1,
t(83) = 2.56, p < .01. Univariate effects were not significant for positive appraisal in
scenario 1 and positive and negative appraisal for scenario 2. We suspect that low
power was the reason we did not find significance for these three effects. According
to Cohen (1997), power is about only .15 for a “small” effect, a group sample size of
about 45 and alpha = .052. The mean proportion of statements containing a negative
appraisal was 11.19% (SD = 9.9%) for GADs, which was significantly higher than
the mean proportion of statements containing a negative appraisal for the nonanxious
participants (M = 6.47%, SD = 6.9%).
We evaluated next BDI and FNES as potential covariates for negative
appraisal. According to Tabachnnick & Fidell (2001), covariates should be
significantly correlated with the dependent variable to compensate for the loss of
degrees o f freedom resulting from the use of ANCOVA. This was not the case:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
31
negative appraisal had .155 correlation with BDI and .279 correlation with FNES, so
they had less than 10% shared variance with negative appraisal. Nevertheless,
ANCOVA resulted in finding no significant effect for negative appraisal when either
BDI or FNES were entered as covariates. A regression analysis was done to evaluate
which covariate (CV) best predicts the DV (Tabachnick & Fidell, 2001). (The DV in
this case is negative appraisal in scenario I). FNES showed an adjusted R square =
.11, g < .001, while BDI had an adjusted R square of .07, g < -05, thus suggesting
that neither one is a particularly good covariate.
To evaluate the association o f GAD status with ATSS ratings o f mood
verbalization for each of the two stimulus scenarios, we conducted t-tests and for
each measure of mood comparing the ratings in the GAD group with those in the
nonanxious control group. The means, standard deviations and effect sizes are
presented in Table 4. Univariate effects were significant only for High PA (HiPA2)
and Low NA (LoNA2) in scenario 2, t (83) = 2.139, g < .05. All other ratings of
mood were not significant, although some ratings showed small effects sizes in the
hypothesized direction (see Table 4). As the correlations between the two ratings of
mood, which were found to be significant (i.e., HiPA2 and LoNA2), and BDI and
FNES respectively were small (.01-.05), the use of ANCOVA was not deemed
justified.
To evaluate the association o f GAD status with ATSS ratings o f temporal
orientation for each o f the two stimulus scenarios, we conducted t-tests for each
measure of temporal orientation comparing the ratings in the GAD group with those
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
32
in the nonanxious control group. The means, standard deviations and effect sizes are
presented in Table 4. Univariate effect was significant only for future orientation, t
(83) = 2.2, p < .05. All other ratings of temporal orientation were not significant,
although future orientation for scenario 2 showed small-to-medium effects size in the
hypothesized direction (see Table 4).
An evaluation of the pooled within-cell correlations (Table 6) for the repeated
measures o f affect (PANAS 1 though PANAS4) showed that PAI-PA4 as a group
and NAI-NA4 as a group have very low correlations with each other as predicted by
Watson & Tellegen’s (1985) two factor model of mood. A repeated measures
MANOVA was performed on the four ratings of negative mood activation, NA1,
NA2, NA3, and NA4. There was a significant multivariate interaction effect
between group and time using Wilks’ criterion, F(3, 81) = 3.0, p < .05; a significant
multivariate effect of group, F( 1, 83) = 2 1.42, p < .001; and a significant multivariate
effect of time, F(3, 81) = 28.07, p < .001. Both the GAD and the control group
respectively had a significant multivariate effect o f time, F(3, 81) = 3.0, p < .001. A
between-subjects simple effects analysis was performed on NA1-NA4 respectively.
The results o f these four t-tests were t(83) = 4.426, p < .001 for NA1, t(83) = 3.576,
p < .001 for NA2, t(83) = 3.544, p < .001 for NA3, t(83) = 4.652, p < .001 for NA4.
The effect sizes and their associated statistical significance results o f the within-
subjects comparisons o f ratings o f NA for the GAD and nonanxious groups
respectively, between times I and 2, 2 and 3 and 3 and 4 are summarized in Table 5.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
33
Table 5. Within-subiects effect sizes for repeated self-report ratines of mood ("PANAS!___________
GAD/PA GAD/NA Control/PA Control/NA
Tim e 1 -2 nill .45*** nill .44***
Time 2 -3 nill .3** Nill .35*
Time 3 - 4 nill .23* nill .44*
Note: * = p < .05; ** = p < .01; *** = p < .001
PA1 PA2 PA3 PA4 NAl NA2 NA3
PA1
PA2 .877
PA3 .831 .842
PA4 .751 .746 .802
NAl -.138 -.141 -.217 -.123
NA2 -.139 -.141 -.258 -.199 .895
NA3 .016 -.020 -.174 -.008 .595 .630
NA4 -.007 .010 -.050 .035 .663 .603 .685
A repeated measures MANOVA was performed on the four ratings of
positive mood activation, PA1, PA2, PA3, and PA4. The interaction between group
and time using Wilks’ criterion was not significant; there was a significant
multivariate effect of group, F(l, 83) = 4.44, p < .05; and a significant multivariate
effect of time, F(3, 81) = 5.60, p < .01. The GAD group trended towards a
significant multivariate effect o f time, F(3, 81) = 2.42, p < .1; and the control group
had a significant multivariate effect of time, F(3, 81) = 3.635, p < .05. A between-
subjects simple effects analysis was performed on PA1-PA4 respectively. The
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
34
results o f these four t-tests were t(83) = -2.334, p < .05 for PA1, t(83) = -1.746, p <
.1 for PA2, t(83) = -2.164, p < .05 for PA3, t(83) = -1.605, ns for PA4. The within-
subjects comparisons of ratings of PA between times I and 2, 2 and 3 and 3 and 4
did not yield significant findings (Table 5).
The pattern of correlations between negative appraisal (NEG1/NEG2) and
ATSS ratings o f negative and positive mood activation (HiNA/LoNA, HiPA/LoPA)
and self-report ratings o f negative mood activation (NA at Time 3) is shown in
Tables 6 and 7. The correlations between NEG1 and HiNAl/LoPAl were .660,
while the correlation between NEG1 and NA3 was .322. These data suggest that
while negative appraisal and the two measures o f negative mood activation covary,
they are different constructs.
Table 7. Pooled within-cell correlations between ATSS verbalization ratings of negative appraisal
NEG1 NEG2 NAl NA2 NA3
NEG1
NEG2 .229
NAl .356 .131
NA2 .376 .151 .895
NA3 .322 .186 .591 .630
NA4 .227 .149 .663 .603 .685
As shown in Table 3, there was a significant difference between the groups at
time 2 in negative mood activation (NA2). As a result an analysis was performed to
determine whether an ANCOVA is appropriate in order to control for the effect of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
35
group differences in NA2 on negative appraisal from scenario 1 (NEG1). In other
words, should the shared variance between NA2 and NEG1 be covaried out for the
test of effect of group status on NEG1. While NA2 and NEGl have a pooled within-
cell correlation of .376 as shown in Table 7, a regression analysis showed that the
adjusted R square for NA2 predicting NEGl was .18 (p < .05). As a result,
ANCOVA in this case was not deemed justified.
Discussion
In summary, we found support for the prediction o f interpretative bias in the
cognitive processing of people who report GAD symptoms. The GAD group
expressed significantly more negative appraisal (NEGl) in response to the social
situation scenario than did the nonanxious control group. The magnitude o f this
difference between the two groups was medium size (d = .56; Cohen, 1977). This
finding is consistent with the Ellis (1969) and Beck (1976) theories o f anxiety, and
with the more recent experimental findings o f Hirsch & Mathews (1997). Although
there was no statistical significance found for the other three ratings o f appraisal -
positive appraisal for scenario 1 (POS1) and positive/negative appraisal for scenario
2 (POS2/NEG2), the obtained effects were in the direction hypothesized for POS1
and for NEG2. The directionality o f the POS2 effect however, was not in the
hypothesized direction. With power being approximately .15-.20 for these effects
(Cohen, 1977), the chance o f Type II error for these non-significant effects is about
80-85%. Scenario 1 elicited a stronger negative appraisal effect than scenario 2 (a d
o f .56 versus .31). A likely explanation is that scenario 1 was designed especially for
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
36
this study and around the most prevalent object o f worry in the social functioning
domain, while the unknown professor scenario, which had been used previously in
Dr. Davison’s lab, only indirectly tapped into achievement-related cognitions and
emotions.
The GAD group expressed more future-oriented verbalizations (Future I) in
response to the social situation scenario than the nonanxious control group. Indeed
there was a medium size effect and statistical significance for Future 1 (d = .51).
These results are consistent with the prior finding reported by Borkovec et al. (1983)
that the majority o f worrisome content refers to the future or past, and with the
Molina et al. (1998) content analysis o f worrisome streams. Although the group
association with Future2 was not statistically significant, the effect was in the
direction hypothesized. The same pattern of effect sizes seen before with appraisal
ratings between stimulus scenarios was found here again; scenario 1 effect sizes
were larger than those found for scenario 2 (d of .51 versus .34). There was no
difference between the frequency o f references to the present between groups for
both scenarios.
All the effect sizes associated with the frequency of mentioning NA and PA
affect words in ATSS verbalizations were in the hypothesized direction. However,
only HiPA2 and LoNA2 achieved statistical significance. In other words, GADs
articulated fewer high-positive-affect words and fewer low-negative-affect words
than their control counterparts. The frequency o f words reflecting potential
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
37
cognitive distortions occurred at a low base rate and no differences were found
between the groups.
The repeated measurement of self-report ratings of mood revealed an
interesting pattern o f results. For time I measurement (which occurred prior to the
start of the experimental manipulation), GADs reported a significantly higher level
o f negative affect activation and lower level o f positive activation. The effect sizes
associated with these significant between-groups were large (d = 1.03) for NA and
medium size for PA (d = .51). So even before the start o f the experimental
manipulation, GAD reported significantly more negative mood and less positive
mood. The time 2 measurement occurred after participants listened to a 70-second
music tape. The time 2 measurements results are consistent with the view that
participants in both groups experienced a mood change in the direction o f reducing
their negative activation (NA). While the differences between the groups remained
significant at time 2 for NA (the differences trended towards significance for PA),
their effect sizes were smaller. Furthermore, the GAD and control groups,
respectively, experienced a significant, within-group, reduction in their ratings of
NA, while there was no difference in their within-group PA ratings between time I
and 2 measurements. While the music appears to have had a significant relaxing
effect on both groups in reducing their within-group negative mood, the GADs
continued to report significantly more negative mood. The time 3 measurement
occurred after the exposure and response to the social situation scenario. This time,
the within-group changes in NA for the two groups showed a significant increase in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
38
NA from time 2 levels, while maintaining a significant between-groups difference in
NA (GAD > Control). The time 4 measurement occurred after the exposure and
response to the unknown professor scenario. Similar to the effect o f music earlier,
there was a significant decrease in within-group NA ratings for both groups, while
the between group difference in NA remained significant and sizable. From Time I
to 4, there were no significant within-group changes in either group in PA ratings. In
summary, the GAD and control groups had mood differences at base line in the
direction predicted by theory. Throughout the experimental manipulation, ratings o f
negative affect changed with the strength o f the effect size in negative appraisal. In
other words, scenario 1, which resulted in a stronger effect for negative appraisal
than scenario 2, was associated with an increase in negative mood activation, while
scenario 2 was associated with a decrease in negative mood activation. Finally, the
pattern o f correlations between negative appraisal and both ATSS and self-report
measures o f mood activation, shows that while the two covary, they are different
constructs. Depending on the type of mood measure, negative appraisal and mood
share. 10-.43 common variance (based on scenario 1 results).
These findings were the result o f a mixed-design (quasi experiment) study
where participants could not be randomly assigned to GAD status. As a result, we
can only observe and report covariation but cannot infer causality between GAD
status and dependent variables. It is possible that participants differed on some other
important variable that may account for the predictive significance o f GAD-status.
For example it is possible that a condition co-morbid with GAD, such as social
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
39
phobia or depression, may be also related to the negative appraisal finding. However
the regression analysis of these covariate candidates (social phobia and depression)
as predictors o f negative appraisal indicated that it is unlikely that social phobia and
depression are alternative explanations for the presence o f negative appraisal in this
study. Another possibility that was examined is that the difference between groups in
negative mood activation prior to the start o f the experimental manipulation is
related to the resulting difference in negative appraisal. However, this difference in
negative mood predicted only 18% o f the variance in negative appraisal, thus making
this alternative explanation not very likely.
It should be noted that in the present study the sample was an analogue
population o f college undergraduates. It is possible that this cohort may differ in
some way from an older community sample of diagnosed participants. However,
undergraduate participants selected with the GAD-Q (Roemer, Borkovec, Posa, &
Borkovec, 1995) did not differ from clinical community samples on a variety of
externally valid criteria such as worry frequency and content (Roemer et al., 1991;
Shadick, Roemer, Hopkins, & Borkovec, 1991), attachment styles (Cassidy, 1995),
and deficiency o f parasympathetic tone (Lyonfields, Borkovec, & Thayer, 1995), or
their scores on the Penn State Worry Questionnaire (Meyer et al., 1990) ( Roemer et
al., 1991). In this study, participants were selected with GAD-Q-IV (Newman et al.,
in press), which was shown to have 17% false positives when compared with ADIS-
IV criteria. Additionally, 35 out o f the 89 study participants were from a community
sample. These participants had a mean age o f 30 versus 20 years for the 54 college
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
40
undergraduates. These older participants did not differ, however, from their college
counterparts on clinical measures of anxiety.
Both stimulus scenarios were designed with a college population in mind, yet
the sample included a significant number of non-college age participants who were
paid for their participation. This may raise the question of degree of relevance o f the
two stimuli for the paid participants. To evaluate this question, responses to
debriefing questions 3, 4 and 5 (see Appendix R) were compared for the GAD group
and the control group separately. For the GAD group, there was no difference
between paid and for course-credit participants on responses to debriefing questions
3 ,4 and 5. The same pattern of results emerged for the control group. So even
though the stimulus topics were about college life, the older participants appear to
have found the two scenarios just as relevant as their younger counterparts.
It is possible that Type I error might have been inflated by the reported
moderate violations in normality (specifically skewness and kurtosis) and the
relatively high number of inferential tests that were made. First, researchers have
reported that a two-tailed t-test is reasonably robust to violations of normality
(Tabatchick & Fidell, 2001; Shaddish & Campbell, 2002). Second, MANOVA was
used whenever dependent variables were sufficiently related to each other (Maxwell
& Delaney, 1990 as cited in Shaddish & Campbell, 2002). Third, for the largest
effect size obtained (Cohen’s d = .56), the power never exceeded .60. An
indiscriminate application o f the Bonferroni correction might have resulted in
overlooking these theoretically important effects.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
41
The psychometric status o f cognitive production methods (e.g., ATSS,
thought listing, etc.) is still not as well established as the more popular pencil and
paper methods. Although Davison et al. (1997) reported promising predictive and
concurrent validity for ATSS, much remains to be done regarding convergent,
discriminant and criterion-related validity (Clark, 1997) and reliability. Another
major issue is the lack o f convergence between different forms o f cognitive
assessment (Chamberlain & Haaga, 1999; Clark, 1997). However, production
methods, such as ATSS, can provide useful information, as we saw in this study, by
focusing on cognition that occurs at a precise moment in the stream o f
consciousness.
Possible order effects may limit the generalizability o f the findings from the
unknown professor scenario, which was presented after the social situation scenario.
The presentation o f the two scenarios was intentionally not counter-balanced.
It appears that content analysis of think aloud streams o f consciousness of
anxious individuals can be a valuable line of research. An interesting additional
direction might be to compare the appraisal ratings of verbalization responses to
standardized stimuli, such as ATSS, with those from open-ended worry induction
paradigms, such as the one used in the Molina et al., (1998). In the latter case,
participants are asked to worry in their usual way about a topic most worrisome to
them. This research would allow for the comparison o f appraisal effects between the
two classes o f stimuli.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
43
Reference List
Andrews, V. H., & Borkovec, T. D. (1988). The differential effects of
inductions o f worry, somatic anxiety, and depression on emotional experience.
Journal of Behavior Therapy & Experimental Psychiatry. 19(11. 21-26.
Beck, A. T., Ward, C. H., Mendelson, M., Mock, J., & Erbaugh, J. (1961).
An Inventory for Measuring Depression. Archives of General Psychiatry. 4. 561-
571.
Beck Aaron T. (1976). Cognitive therapy and the emotional disorders .
Beck, A. T., & Clark, D. a. (1997). An information processing model o f
anxiety: automatic and strategic processes. Behaviour Research & Therapy. 35(1).
49-58.
Borkovec, T. D. (1994). The nature, functions, and origins of worry.
Worrying: Perspectives on Theory. Assessment and Treatment.. 5-33.
Borkovec, T. D., Robinson, E., Pruzinsky, T., & Depree, J. a. (1983).
Preliminary exploration of worry: some characteristics and processes. Behaviour
Research & Therapy. 21(1). 9-16.
Bradley, B. P., Mogg, K., Falla, S. J., & Hamilton, L. R. (1998). Attentional
bias for threatening facial expressions in anxiety: manipulation of stimulus duration.
Cognition & Emotion. 12(61. 737-753.
Bradley, B. P., Mogg, K., White, J., Groom, C., & De Bono, J. (1999).
Attentional bias for emotional faces in generalized anxiety disorder. British Journal
o f Clinical Psychology. 38(31. 267-278.
Brown, T. A., Antony, M. M., & Barlow, D. H. (1992). Psychometric
properties o f the penn state worry questionnaire in a clinical anxiety disorders
sample. Behaviour Research & Therapy. 30(11. 33-37.
Carver Charles S., & White Ten L. (1994). Behavioral inhibition, behavioral
activation, and affective responses to impending reward and punishment: The
BIS/BAS Scales. Journal of Personality & Social Psychology. 67(21. 319-333.
Cassidy, J. (1995). Attachment and generalized anxiety disorder. Emotion.
Cognition, and Representation (pp. 343-370, xiii, 437).
Chamberlain John, & Haaga David A. (1999). Convergent validity o f
cognitive assessment methods. Behavior Modification. 23(21. 294-315.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
44
Craske, M. G., Rapee, R. M., Jackel, L., & Barlow, D. H. (1989). Qualitative
dimensions o f worry in dsm-iii--r generalized anxiety disorder subjects and
nonanxious controls. Behaviour Research & Therapy. 27(4). 397-402.
Davison Gerald C., Robins Clive, & Johnson Marcia K. (1983). Articulated
thoughts during simulated situations: A paradigm for studying cognition in emotion
and behavior. Cognitive Therapy & Research. 7(1). 17-39.
Davison Gerald C., Vogel Ralph S., & Coffman Sandra G. (1997). Think-
aloud approaches to cognitive assessment and the articulated thoughts in simulated
situations paradigm. Journal of Consulting & Clinical Psychology. 65(61. 950-958.
Eckhardt Christopher I., Barbour Krista A., & Davison Gerald C. (1998).
Articulated thoughts of maritally violent and nonviolent men during anger arousal.
Journal of Consulting & Clinical Psychology. 66(2). 259-269.
Ellis Albert. (1969). Rational-Emotive Therapy. Journal of Contemporary
Psychotherapy. 1(2). 82-90.
Freeston, M. H., Rheaume, J., Letarte, H., Dugas, M. J., & And Otherset Al.
(1994). Why do people worry? Personality & Individual Differences. 17(61. 791-802.
Glass Carol R., & Amkoff Diane B. (1997). Questionnaire methods o f
cognitive self-statement assessment. Journal of Consulting & Clinical Psychology.
65(6), 911-927.
Grey Susan, & Mathews Andrew. (2000). Effects o f training on interpretation
o f emotional ambiguity. Quarterly Journal of Experimental Psychology: Human
Experimental Psychology. 53(41. 1143-1162.
Haaga David A., Davison Gerald C., McDermut Wilson, Hillis Stephen L., &
and others. (1993). "States-of-mind" analysis o f the articulated thoughts of
exsmokers. Cognitive Therapy & Research. 17(51. 427-439.
Haaga David A. F. (1997). Introduction to the special section on measuring
cognitive products in research and practice. Journal of Consulting & Clinical
Psychology. 65(61. 907-910.
Hirsch, C., & Mathews, A. (1997). Interpretive inferences when reading
about emotional events. Behaviour Research & Therapy. 35(121. 1123-1132.
Ingram Rick E. (1990). Self-focused attention in clinical disorders: Review
and a conceptual model. Psychological Bulletin. 107(21. 156-176.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
45
Ingram Rick E., Kendall Philip C., Ingram Rick E., Kendall Philip C., Smith
Timothy W., Donnell Christina, & and others. (1987). The cognitive side o f anxiety:
Cognitive specificity in emotional distress. Cognitive Therapy & Research: Journal
o f Personality & Social Psychology. 11. 53(5.4), 523-536, 734-742.
Lyonfields, J. D., Borkovec, T. D., & Thayer, J. F. (1995). Vagal tone in
generalized anxiety disorder and the effects o f aversive imagery and worrisome
thinking. Behavior Therapy. 26(3). 457-466.
MacLeod, C., & Cohen, I. L. (1993). Anxiety and the interpretation o f
ambiguity: a text comprehension study. Journal o f Abnormal Psychology. 102(2).
238-247.
Mathews, A. (1990). Why worry? The cognitive function o f anxiety.
Behaviour Research & Therapy. 28(61. 455-468.
Meyer, T. J., Miller, M. L., Metzger, R. L., Borkovec, T. D., & And Otherset
Al. (1990). Development and validation of the penn state worry questionnaire.
Behaviour Research & Therapy. 28(61. 487-495.
Mogg, K., & Bradley, B. P. (1998). A cognitive-motivational analysis of
anxiety. Behaviour Research & Therapy. 36191. 809-848.
Molina Silvia, & Borkovec, T. D. (1994). The Penn State Worry
Questionnaire: Psychometric properties and associated characteristics. Worrying:
Perspectives on Theory. Assessment and Treatment.. 265-283.
Molina, S., Borkovec, T. D., Peasley, C., & Person, D. (1998). Content
analysis o f worrisome streams o f consciousness in anxious and dysphoric
participants. Cognitive Therapy & Research. 22(2). 109-123.
Pennebaker, J. W., Czajka, J. a., Cropanzano, R., Richards, B. C., & And
Others. (1990). Levels o f thinking. Personality & Social Psychology Bulletin. 16(4).
743-757.
Roemer, L., Borkovec, M., Posa, S., & Borkovec, T. D. (1995). A self-report
diagnostic measure of generalized anxiety disorder. Journal o f Behavior Therapy and
Experimental Psychiatry. 26141. 345-350.
Roemer, L., Molina, S., & Borkovec, T. D. (1997). An investigation o f worry
content among generally anxious individuals. Journal of Nervous and Mental
Disease. 185151. 314-319.
Tabachnick Barbara G., & Fidell Linda S. (1996). Using Multivariate
Statistics.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
46
Tallis, F., & Eysenck, M. W. (1994). Worry: mechanisms and modulating
influences. Behavioural & Cognitive Psychotherapy. 22(1). 37-56.
Watson David, & Friend Ronald. (1969). Measurement of Social-Evaluative
Anxiety. Journal of Consulting & Clinical Psychology. 33(4). 448-457.
Watson, D., & Tellegen, A. (1985). Toward a consensual structure o f mood.
Psychological Bulletin. 98(21. 219-235.
White Judith, Davison Gerald C., Haaga David A., & White Kerrin. (1992).
Cognitive bias in the articulated thoughts of depressed and nondepressed psychiatric
patients. Journal of Nervous & Mental Disease. 180(2). 77-81.
Williams Marian E., Davison Gerald C., Nezami Elahe, & DeQuattro Vincent
L. (1992). Articulated thoughts o f Type A and B individuals in response to social
criticism. Cognitive Therapy & Research. 16(1). 19-30.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
47
Appendix A
Social situation narrative
Segment 1
You have just graduated from high school and you are about to embark on a
whole new life in college at USC. Before entering college, the university has
set up an optional orientation where you could meet other students and learn
more about life on campus. You are walking around campus one day and
you notice an invitation to the freshman orientation the following week. You
start to wonder if you should attend and if this might be an opportunity to
meet people and make some new friends.
Segment 2
Later on the same evening you start thinking about the freshman orientation
and begin to imagine meeting people and interacting with them. You wonder
about what kinds o f people show up for such events and if this event might be
an opportunity to meet interesting people that you could get to know better.
Segment 3
A few days have gone by and you decided to attend the event. You wonder
what kinds of people attend such an event and whom you might get to meet.
It is not clear to you who really attends these events. You wonder if you
might meet some new friends to hang out with when college starts.
Segment 4
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
48
It is the day of the event and you are now walking into the Alumni Park
where it is taking place. You size up the crowd and try to figure out whom to
approach. You walk up to a group and start interacting with the people.
Some o f the people you are talking to have interesting things say.
Segment 5
The event is now over. You met a few interesting people and agreed to stay
in touch with them. You exchanged phone numbers and email addresses with
a couple people you found particularly interesting. As you are walking away
from the event you realize that you appreciated the interactions and wonder if
they will lead to any new friendships.
Segment 6
A few days have gone by since the event took place. One evening, you recall
the event and realized that have not heard from any of the people you met
there. You wonder who will take the first step and if it might lead to any
friendship possibilities. You wonder if you should call first. You wonder if
the other people will ever call. You wonder if they have friends.
Segment 7
The following day you walk into the cafeteria and you notice that the people
you met at the event are having lunch together. You try to catch their eyes
but they don’t notice you. You begin to wonder if you should approach
them.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
49
Appendix B
Unknown Professor narrative
Segment 1
Narrator: Imagine that you’re sitting alone at a table in the cafeteria and
are overhearing two students, whom you don’t know, talking about a
professor that you have never taken.
: Ya know, that was one o f the most boring lectures I’ve been to this whole
year.
: (laughs) the course isn’t turning out to be at all what I’d hoped it would
be
: Oh, I know
: I mean, if the subject had some life in it-but the teacher doesn’t seem to
get into it very much or show any enthusiasm at all.
Segment 2
:Oh, I know what you mean. You know, he doesn’t even seem to care
about what he’s teaching. I guess he’s been doing it for a very long time-
teaching the same old material-but it gets pretty boring after awhile.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
50
:No kidding. Besides, a lot of the professors here don’t really care about
teaching anyway. I think they’re mostly into doing their own research and
they all look on teaching as some kind o f chore that they have to do
Segment 3
: he sure hasn’t spared any effort in making up the reading list though.
: (laughs)
: he makes sure we do plenty of work even if he doesn’t
:Yeah, that’s for sure. And the reading seems to be too much. I don’t
know how he expects us to keep up. I feel half-asleep last night trying
to read some of that stuff.
: Yeah, so did I
: God, when it comes to exam time, he can’t expect us to remember it all.
He should tell us what is most important.
Segment 4
:Yeah, I wish I knew what he considers to be most important because I’ve
found that most o f what I’ve read has been pretty interesting.
: Yeah
: Ya know, I’d rather read a lot o f something interesting than something I
couldn’t relate to at all
:Uh, huh, I know what you mean-some o f his articles have been pretty
interesting
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
51
Segment 5
: It’s a shame he can’t make those lectures more interesting
:Oh, I know. It’s his style that really kills it It becomes just so much
material to be memorized instead of something you can really get involved
in.
: I know, I’d like him to get more into why these ideas are interesting and
what questions they provoke rather than just concentrating on a list of
experiments people have done and what the results were.
: yeah
Segment 6
: At least his lectures are well organized though. I’ll admit that. I mean, he
knows exactly what he’s gonna cover, in what time, and he sticks right to
it.
: Well, yeah.
: All o f his headings and subheadings and everything make it easy to follow
the course of the lecture. Some of the other professors here could leam a
few things from him, don’t ya think?
Segment 7
: w ell, yeah. From what I’ve heard, I guess the other profs think very
highly o f him as far as his research is concerned. He’s well known in his
field and he’s well respected.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
52
: Oh, yeah.
:A lot of students would give anything to do some research with him.
: Yeah
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Integrating acceptance with imaginal exposure to core schema: Process and outcome in generalized anxiety disorder

PDF

Affection and conflict in family relationships

PDF

Articulated thoughts about intentions to commit anti-gay hate crimes

PDF

Hedonic aspects of conditioned taste aversion in rats and humans

PDF

Articulated thoughts regarding cognitions toward older adults

PDF

Expectancies for alternative behaviors predict drinking problems: A test of a cognitive reformulation of the matching law

PDF

A typology of maritally violent men and correlates of violence

PDF

Applying the theory of reasoned action to condom use: The effect of immediate consequences on intention-behavior consistency

PDF

"People of the (cook)book": The enculturation of American Jewish women

PDF

Head injury and dementia: A co-twin control study of Swedish twins

PDF

Depression and suicidality in Latino adolescents: A study of acculturation and gender role beliefs

PDF

Emotion regulation as a mediator between family-of-origin aggression and marital aggression

PDF

An integrative analysis of racism and quality of life: A comparison of multidimensional moderators in an ethnically diverse sample

PDF

Effects of bilateral stimulation and stimulus redundancy on performance in processing nonword letter trigrams

PDF

Depression in offspring following severe prenatal stress

PDF

Family functioning, intergenerational conflict, and psychological symptomology of Asian American college students

PDF

Insights into the nature of phonological and surface dyslexia: Evidence from a novel word learning task

PDF

Ethnic identity, acculturation, self-esteem and perceived discrimination: A comparison study of Asian American adolescents

PDF

Adopting the China route

PDF

Interrupted passage: A documentary script

Asset Metadata

Creator Sachs, Andrei David (author)

Core Title Content analysis of articulated thoughts of chronic worriers

School Graduate School

Degree Master of Arts

Degree Program Psychology

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag OAI-PMH Harvest,Psychology, clinical,sociology, individual and family studies

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Davison, Gerald C. (committee chair), [illegible] (committee member), Walsh, David A. (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-295699

Unique identifier UC11337784

Identifier 1411801.pdf (filename),usctheses-c16-295699 (legacy record id)

Legacy Identifier 1411801.pdf

Dmrecord 295699

Document Type Thesis

Rights Sachs, Andrei David

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA