Psychological Trauma: Theory, Research, Practice, and Policy © 2015 American Psychological Association

2016, Vol. 8, No. 3, 259 –266 1942-9681/16/$12.00 http://dx.doi.org/10.1037/tra0000073

Shattering World Assumptions: A Prospective View of the Impact of

Adverse Events on World Assumptions
Eric R. Schuler and Adriel Boals
University of North Texas

Shattered Assumptions theory (Janoff-Bulman, 1992) posits that experiencing a traumatic event has the
potential to diminish the degree of optimism in the assumptions of the world (assumptive world), which
could lead to the development of posttraumatic stress disorder. Prior research assessed the assumptive
world with a measure that was recently reported to have poor psychometric properties (Kaler et al., 2008).
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

The current study had 3 aims: (a) to assess the psychometric properties of a recently developed measure
This document is copyrighted by the American Psychological Association or one of its allied publishers.

of the assumptive world, (b) to retrospectively examine how prior adverse events affected the optimism
of the assumptive world, and (c) to measure the impact of an intervening adverse event. An 8-week
prospective design with a college sample (N ⫽ 882 at Time 1 and N ⫽ 511 at Time 2) was used to assess
the study objectives. We split adverse events into those that were objectively or subjectively traumatic
in nature. The new measure exhibited adequate psychometric properties. The report of a prior objective
or subjective trauma at Time 1 was related to a less optimistic assumptive world. Furthermore,
participants who experienced an intervening objectively traumatic event evidenced a decrease in
optimistic views of the world compared with those who did not experience an intervening adverse event.
We found support for Shattered Assumptions theory retrospectively and prospectively using a reliable
measure of the assumptive world. We discuss future assessments of the measure of the assumptive world
and clinical implications to help rebuild the assumptive world with current therapies.

Keywords: trauma, subjective trauma, shattered assumptions, prior trauma history, psychometrics

Shattered Assumption theory (Janoff-Bulman, 1989, 1992) cen-
ters on the destructive nature of a trauma on an individual’s belief
system or assumptive world that focuses on beliefs related to the
world as benevolent, the world as meaningful, and the self as
worthy. Exposure to a traumatic event may provide information
about the nature of the world that is incongruent with the pre-
trauma assumptions, which has the potential to shatter these beliefs
(Janoff-Bulman, 1992). Individuals may attempt to rebuild their
assumptive worlds to incorporate the traumatic experience (e.g.,
the world is mostly good, but some people cannot be trusted);
however, it is during this process that the symptoms of posttrau-
matic stress disorder (PTSD) are believed to arise (Janoff-Bulman,
1992). Prior research assessing if a trauma affects the assumptive
world are highlighted along with methodological and measurement
issues. The present study empirically assesses the psychometrics of
a new measure of the assumptive world and tests whether an
intervening traumatic event can affect the beliefs in a prospective
design.1
Previous Research into Shattered Assumptions
The most widely used measure of the assumptive world is the
World Assumptions Scale (WAS; Janoff-Bulman, 1989), which
consists of 32 items that are divided into eight subscales that
include: benevolence of the world, benevolence of people, con-
trollability, justice, self-worth, randomness, self-control, and luck.
Prior research that has assessed the hypothesis that a trauma affects
the assumptive world has reported mixed findings, which may be
due to measurement and design issues (Kaler et al., 2008). Indi-
viduals with prior trauma exposure reported less optimistic as-
sumptive worlds compared with individuals with no prior traumas
in a cross-sectional study (Janoff-Bulman, 1989). In addition,
breast cancer survivors (diagnosed 5 years earlier) reported less
control and that the world was more random compared with the
matched control group (Tomich & Helgeson, 2002). However,
because these studies were cross-sectional, the change in the world
assumptions pre- and posttrauma could not be assessed.
Subsequently, some longitudinal research has attempted to as-
sess if changes in the world assumptions were caused by either the
experience of a trauma or by the development of PTSD symptoms.
Ginzburg (2004) had participants who were recovering from a
myocardial infarction and a matched control group completed the
WAS and again 7 months later. Participants who both experienced
a myocardial infraction and developed PTSD at Time 2 reported
lower levels of luck, self-worth, and randomness compared with
both the matched control and those who did not develop PTSD 7
months later (Ginzburg, 2004). The changes in assumptions were

This article was published Online First July 27, 2015.

Eric R. Schuler and Adriel Boals, Department of Psychology, University
of North Texas.
Correspondence concerning this article should be addressed to Adriel
Boals, Department of Psychology, University of North Texas, 1155 Union
Circle #311280, Denton, TX 76203. E-mail: adriel@unt.edu
1
The term prospective design is used here in regards to assessing the
impact of an intervening traumatic event after accounting for the impact of
prior traumas. It is important to note that our design is not a truly
prospective design because our sample does contain individuals with prior
trauma exposure.

259
 260 SCHULER AND BOALS
reportedly due to the development of PTSD rather than trauma
exposure (Ginzburg, 2004). Ginzburg (2004) noted that not having
a pre-event measure of the assumptive world and the specificity of
the trauma may limit generalizability. Prior stressful life events
were reported to impact the assumptive world of mothers whose
children were undergoing a bone morrow transplant (Rini et al.,
2004). Mothers may have known their child’s diagnoses before
Time 1; as such, their assumptions may have already been altered
(Rini et al., 2004). Without pre-event measures of the assumptive
world, trauma’s impact on these beliefs could not be fully as-
sessed.
One methodological concern that could account for inconsistent
findings is the use of cross-sectional, retrospective designs (Kaler
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
et al., 2008). Cross-sectional designs can be problematic when
This document is copyrighted by the American Psychological Association or one of its allied publishers.
assessing for changes in the assumptive worlds in that pre-event
measures cannot be collected (Kaler et al., 2008). Retrospective
studies ask individuals to determine how much a trauma has
violated their assumptive world after the fact and makes the
assumption that these pretrauma views can be accessed by the
individual (Edmondson et al., 2011). However, when individuals
are asked to make temporal comparisons of their pretrauma selves,
there may be a derogation of the earlier self (McFarland & Alvaro,
2000). The previously discussed studies implemented retrospec-
tive designs, which could explain some of the inconsistent find-
ings.
The aforementioned studies utilized cross-sectional and longi-
tudinal designs to assess whether experiencing a traumatic event
can affect the assumptive world. A stronger research design would
be to measure the assumptive worlds before and after a traumatic
event. College students were given the WAS at Time 1 and again
approximately 8 weeks later to assess if they reported experiencing
a traumatic event in the intervening time (Kaler et al., 2008). Kaler
et al. (2008) posited that participants who were exposed to multiple
traumatic events would evidence less optimistic assumptions. This
was based on the finding of Breslau, Chilcoat, Kessler, and Davis
(1999) that experiencing multiple traumas was associated with an
increased likelihood of PTSD. Of the 655 participants who com-
pleted Time 2, 56 reported an intervening trauma (Kaler et al.,
2008). A majority of the intervening events were a sudden death of
a loved one, an assault, or a serious injury (Kaler et al., 2008).
However, there were no differences in the assumptions of victims
of an intervening trauma and nonvictims (Kaler et al., 2008). Kaler
et al. (2008) suggested that the null findings were believed to be
from the unreliable properties of the WAS.
Most research that has attempted to test the hypothesis that a
traumatic event shatters an individual’s assumptive world has
relied on the WAS (Janoff-Bulman, 1989), including the afore-
mentioned prospective study (Kaler et al., 2008). The WAS had
small correlations with the number of lifetime traumas, distress,
depression, and PTSD symptom severity (Kaler et al., 2008).
Higher scores on the WAS (more optimism) should be related to
less distress, PTSD symptom severity, and depression. Because of
the small correlations, there was little support for the relationship
between violations to the assumptive world and PTSD symptoms.
The WAS was reported to have low internal reliability (ranged
from .60 to .83) and test–retest reliability (ranged from .38 to .65)
in the control sample (Kaler et al., 2008). With the lack of temporal
reliability, assessing the shattering effect may be hindered by
measurement error. The WAS had small to near-zero correlations
with measures of optimism, satisfaction in life, and self-esteem,
reflecting poor convergent validity (Kaler et al., 2008). The use of
a measure with low reliability may diminish the accuracy of the
results and may account for the “inconsistencies in the empirical
literature” (Kaler et al., 2008, p. 332). On the basis of the theory,
there would be substantive differences in optimism (and large
effect sizes) after a trauma or when comparing the degree of
optimism between trauma victims and a control group (Janoff-
Bulman, 1989, 1992). Prior research has reported low effect sizes;
however, given the poor psychometric properties of the WAS, the
effect sizes may have been underestimated. As such, it is unclear
what the expected effect sizes are.
In an unpublished dissertation, Kaler (2009) developed the
World Assumptions Questionnaire (WAQ) to provide researchers
with a reliable and valid measure of the assumptive world. The aim
of the Kaler (2009) study was to create and assess the psychomet-
rics of the WAQ. The WAQ comprises four subscales (Controlla-
bility of Events, Comprehensibility and Predictability of People,
Trustworthiness and Goodness of People, and Safety of the World)
that are consistent with Shattered Assumptions theory. Kaler
(2009) implemented the study design and procedures from the
Kaler et al. (2008) study. Undergraduate participants (N ⫽ 270)
completed Time 2 of the study, and 36 individuals reported an
intervening traumatic event and less optimistic assumptions of the
world (Kaler, 2009). The WAQ subscales were reported to have
adequate internal reliability across time points (ranged from .74 to
.82) and test–retest reliability (ranged from .68 to .74) (Kaler,
2009, p. 47). The WAQ was hypothesized to have small to mod-
erate positive correlations with similar constructs and small to
moderate negative correlations with measures related to distress
(Kaler, 2009). The WAQ exhibited adequate convergent validity
with self-report measures of the constructs of satisfaction with life
(rs ranged from .11 to .36) and positive affect (rs ranged from .02
to .34) (Kaler, 2009, p. 47). Convergent validity for the specific
subscales were as follows: perceived control over the events was
correlated with the Control of Events subscale (r ⫽ .38), attach-
ment was uncorrelated with the Comprehensibility and Predict-
ability of People subscale (rs ranged from ⫺.07 to .01), Trustwor-
thiness and Goodness of People was correlated with a measure of
trust (r ⫽ .69), and the Safety of the World subscale was nega-
tively correlated with a measure of the future likelihood of expe-
riencing a negative event (r ⫽ ⫺.19) (Kaler, 2009, p. 47). The
WAQ subscales had small to moderate negative correlations with
negative affect (rs ranged from ⫺.42 to ⫺.15), distress (rs ranged
from ⫺.46 to ⫺.17), and PTSD symptom severity (rs ranged
from ⫺.36 to ⫺.24) (Kaler, 2009, p. 47). Kaler (2009, p. 47)
reported that the WAQ subscales had adequate discriminate valid-
ity based on the near-zero correlations with social desirability (rs
ranged from ⫺.09 to .04). Kaler (2009) reported that the four
factor structure had adequate model fit. On the basis of Kaler
(2009), the WAQ has good psychometric properties and is able to
measure the core assumptions.
The WAQ was recently used in an 8-week prospective design to
assess if an intervening traumatic event would cause reliable
changes in the assumptive world, PTSD symptom severity, mental
and physical health, and distress (Anders, Frazier, & Shallcross,
2014). A potentially traumatic event was any highly distressing
event that was nominated on a traumatic history that included
events that do not meet the A1 criteria (e.g., being bullied) based
 SHATTERED ASSUMPTIONS
on a pilot. Anders et al. (2014) had participants (N ⫽ 1,082)
complete measures at Time 1 and Time 2. At Time 2, 153 partic-
ipants reported experiencing a highly distressing, potentially trau-
matic event since Time 1 on the measure of potentially traumatic
events for the 8-week interim (Anders et al., 2014). There was a
decrease in the Control WAQ subscale, but no differences or
reliable change on the other subscales (Anders et al., 2014).
However, there were changes in distress and mental and physical
wellbeing (Anders et al., 2014). The WAQ subscales were reported
to have good internal reliability (Anders et al., 2014); however,
further psychometric assessments are needed.
Most prior research has focused on the impact of Diagnostic and
Statistical Manual of Mental Disorders, Fourth Edition—Text
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Revision (DSM–IV–TR; American Psychiatric Association [APA],
This document is copyrighted by the American Psychological Association or one of its allied publishers.
2000) defined traumatic events on the assumptive world. Adverse
events (e.g., the dissolution of a serious relationship) may not meet
the A1 and A2 criteria of the DSM–IV–TR; however, they could be
traumatic for the individual if it became an core part of the
individual’s life story, called subjective traumas (Boals &
Schuettler, 2009; Boals, Southard-Dobbs, & Blumenthal, 2015). In
the previously mentioned study by Anders et al. (2014), potentially
traumatic events that were high in distress but may not meet the
DSM–IV–TR criteria for PTSD could be considered subjective
traumas. A way to assess if an event was subjectively traumatic is
to see if the event has become a central part of the life story, called
event centrality (Berntsen & Rubin, 2006). A subjective trauma is
one that has become highly central to their life story, reflected by
high scores on the Centrality of Events Scale (CES; Berntsen &
Rubin, 2006). Highly central events were determined by a quartile
split on the CES, with the highest 25% reflecting subjective
traumas (Boals et al., 2015). Highly central events were reported to
correspond to higher levels of PTSD symptom severity and dis-
tress (Berntsen & Rubin, 2006). Subjective and objective traumatic
events could potentially shatter the world assumptions.
Present Study
In the present study, we tested Shattered Assumptions theory
through the use of several methodological improvements over past
studies. We replicated the design used in Anders et al. (2014) and
Kaler et al. (2008) to obtain a baseline measure of the assumptive
world and had participants complete Time 2 approximately 8
weeks later to assess if there was an intervening adverse life event
and if the assumptive world was affected. Although 8 weeks is a
relatively short time for a potentially traumatic event to occur, this
time frame was selected to replicate the methodology of Anders et
al. (2014) and Kaler et al. (2008). The 8-week interim allowed 4
weeks to collect Time 1 data at the beginning of the semester and
3 weeks to collect Time 2 data before the participant pool closed.
At Time 1, participants completed the WAQ and measures of
similar constructs and measures of distress to assess convergent
and discriminant validity. Second, we used the WAQ to assess the
assumptive world, which appears to be a reliable measure of the
assumptive world. However, the WAQ was only used in a couple
of studies (Anders et al., 2014; Kaler, 2009) and further psycho-
metric assessment is needed. Internal reliability and test–retest
reliability were assessed to determine if WAQ had adequate psy-
chometric properties. In the current study, the WAQ total scores
were used instead of the four subscales because the underlying
261
structure of the WAQ has not been assessed using a confirmatory
factor analysis outside of the unpublished dissertation. Third, we
examined objective and subjective traumatic events and their im-
pact on world assumptions. In the current study, an objective
trauma was an event nominated on a trauma history questionnaire
without necessarily meeting the DSM–IV–TR’s A1 and A2 criteria
(APA, 2000, p. 463). The data were collected while the DSM–IV–
TR’s (APA, 2000) criteria for PTSD were still in use; however, at
the end of data collection, the field shifted to the DSM 5th edition
(DSM-5; APA, 2013), which has a narrower criterion for the
diagnosis of PTSD. As such, the DSM–IV–TR criteria were imple-
mented. Subjective traumas are events in the top quartile of CES
scores (Boals et al., 2015). There were three specific hypotheses:
Hypothesis 1: The WAQ would exhibit adequate psychomet-
ric properties related to internal reliability, test–retest reliabil-
ity, and convergent validity. To assess the convergent validity,
the WAQ would be related to less depression, negative affect,
event centrality, negative thoughts about the world, and PTSD
symptoms, but be related to more life satisfaction and positive
affect.
Hypothesis 2: Individuals who have experienced a prior sub-
jective or objective traumatic event would have a less opti-
mistic assumptive world (lower WAQ scores) compared with
individuals with no prior traumatic experience.
Hypothesis 3: Individuals who experienced an intervening
traumatic event (either objective or subjective) between Time
1 and Time 2 would evidence a less optimistic assumptive
world compared with individuals who did not.
Method
Participants
Participants (N ⫽ 882) were recruited from a midsized univer-
sity in the southern United States. Participants received partial
course credit in their psychology courses for participating. There
were no exclusionary criteria for this study. The demographic data
of gender, ethnicity, and age for Time 1 and Time 2 are reported
in Table 1.
Measures
The WAQ (Kaler, 2009) consists of 25 items that correspond to
the three core assumptions posited by Janoff-Bulman (1992). The
participants rate their agreement toward statements about their
assumptions of the world using 6-point scales ranging from 1
(strongly agree) to 6 (strongly disagree), with seven items being
reverse coded. Lower scores on the WAQ reflect less optimistic
assumptions. An example item is “People often behave in unpre-
dictable ways.” The WAQ subscales been reported to have ade-
quate internal reliability (ranges from .64 to .83) (Anders et al.,
2014), good test–retest reliability (subscales ranged from .68 to
.74), and adequate convergent and discriminate validity (Kaler,
2009). In the current study, the WAQ total scores had an internal
reliability of .82 at Time 1 and ␣ ⫽ .84 at Time 2 and had a
test–retest reliability of .67.
 262 SCHULER AND BOALS

Table 1 To assess the participant’s past traumatic history, the Stressful

Descriptive Statistics of Gender, Ethnicity, and Age Reported at Life Events Screening Questionnaire (SLESQ; Goodman, Corc-
Time 1 and Time 2 oran, Turner, Yuan, & Green, 1998) was used. The SLESQ con-
sists of 11 specific categories of traumatic events and two general
Time 1 Time 2 events to assess the lifetime exposure to traumatic life events in
Variable N % N % nontreatment-seeking participants. The SLESQ has been found to
have adequate convergent validity and test–retest reliability, and it
Gender
Men 255 28.90 150 29.40
can differentiate between A1 and non-A1 criteria for PTSD diag-
Women 627 71.10 361 70.60 nosis in a college sample of women (Goodman et al., 1998). A
Total 882 100.00 511 100.00 modified version of the SLESQ (items that ask about childhood
Ethnicity traumas were removed) was used to assess for an intervening
Caucasian 451 51.10 277 54.20 trauma at Time 2.
African American 129 14.60 65 12.70
Asian 58 6.60 29 5.70 The PTSD Checklist–Specific (PCL-S; Weathers, Litz, Herman,
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Hispanic 200 22.70 119 23.30 Huska, & Keane, 1993) was used to assess the severity of PTSD
This document is copyrighted by the American Psychological Association or one of its allied publishers.

Native American 5 0.60 4 0.80
Other 39 4.40 17 3.30
882 100.00 511 100.00
Mean SD Mean SD
Age (in years) 19.89 3.68 19.87 3.52
Note. Self-reported demographic information collected from the partici-
pants at Time 1 and Time 2.
The Positive and Negative Affect Schedule (PANAS; Watson,
Clark, & Tellegen, 1988) assesses participants’ degrees of positive
and negative emotions on two separate subscales, each consisting
of 10 single-word items. Participants are asked how often they
experienced the emotion in the last 2 weeks using 5-point scales
ranging from 1 (very slightly or not at all) to 5 (extremely). Watson
et al. (1988) reported adequate psychometric properties for the
PANAS. In the current study, the internal reliability for the posi-
tive affect subscale was ␣ ⫽ .89 at Time 1 and ␣ ⫽ .92 at Time
2. The internal reliability for the negative affect subscale of the
PANAS was ␣ ⫽ .87 at Time 1 and ␣ ⫽ .86 at Time 2.
The Satisfaction with Life Scale (SWLS; Diener, Emmons,
Larsen, & Griffin, 1985) is a measure of how much a person feels
happy with how things are in their life. The scale consists of five
items in which participants rate their agreement using 7-point
scales ranging from 1 (the statement is absolutely untrue) to 7 (the
statement is absolutely true). The SWLS was reported to have
good convergent and discriminate validity (Pavot & Diener, 1993).
In the current study, the internal reliability of ␣ ⫽ .87 at Time 1
and was ␣ ⫽ .90 at Time 2.
The Depression, Anxiety, and Stress Scales (DASS-21; Lovi-
bond & Lovibond, 1995) is a 21-item scale that is composed of
three subscales (depression, anxiety, and stress). Each of the three
subscales was measured with seven items. Participants rate how
frequently in the last 2 weeks they experienced each of the items
using 4-point scales ranging from 0 (does not apply) to 3 (very
much so). The DASS-21 has previously been reported to have
good internal consistency and concurrent validity (Antony, Biel-
ing, Cox, Enns, & Swinson, 1998). For the current study, the two
items about suicide ideation were removed because of the online
nature of the study and not having a protocol in place if partici-
pants reported high suicide ideation. Using this modified DASS
affects the scoring of the measure with a lower maximum and
means. The modified DASS had an internal reliability of .94 at
Time 1 and .96 at Time 2.
symptoms. Participants nominated their most traumatic event and
then rated the frequency they have experienced each item of the 17
items using a 5-point scale that ranged from 1 (not at all) to 5
(extremely). Wilkins, Lang, and Norman (2011) synthesized the
psychometric properties of the PCL-S from 19 studies. The PCL-S
was reported to have adequate test–retest reliability (.66), conver-
gent validity with other measures of PTSD symptom severity, and
an internal consistency (⬎.75) (Wilkins et al., 2011). In the current
study, the internal reliability was ␣ ⫽ .95 at Time 1 and ␣ ⫽ .96
at Time 2.
The CES (Berntsen & Rubin, 2006) is a measure of the degree
to which a nominated event has become an integral part of the life
story. Participants were asked to use their most adverse life event
in reference to the items. The CES consists of 20 items that
participants rated how strongly they agree with each statement
using a 5-point scale ranging from 1 (totally disagree) to 5 (totally
agree). Prior research has reported that the CES has high internal
reliability across college samples (range: .93 to .95) and was
positively correlated with PTSD symptom severity (r ⫽ .38) and
depression symptoms (r ⫽ .23) (Berntsen & Rubin, 2006). In the
current study, the internal reliability was ␣ ⫽ .97 at Time 1 and
␣ ⫽ .98 at Time 2.
The Negative Cognitions about the World (NCW) subscale of
the Posttraumatic Cognitions Inventory (PTCI; Foa, Elhers, Clark,
Tolin, & Orsillo, 1999) was used to assess the specific negative
cognitions related to an adverse event. The NCW subscale contains
seven items in which participants rate their agreement using a
7-point scale ranging from 1 (totally disagree) to 7 (totally agree).
Beck et al. (2004) reported that the PTCI had adequate structural
and convergent validity. In the current study, the internal reliability
at Time 1 was ␣ ⫽ .92 and ␣ ⫽ .92 at Time 2.
In the demographic questionnaire, participants were asked to
click and/or type in answers that corresponded to their gender, age,
and ethnicity.
Procedure
Participants completed the measures as part of a larger online
study. At Time 1, participants completed all questionnaires in the
order listed above. At Time 1, if the participant nominated expe-
riencing multiple traumatic events on the SLESQ, then they were
asked which was their most traumatic and to use that event in
reference to completing the CES and PCL-S. If there was no prior
trauma nominated on the SLESQ, then participants were asked
what their most stressful life event was and to use that event in
 SHATTERED ASSUMPTIONS 263

reference to the CES and PCL-S. To replicate the methodology of
Anders et al. (2014) and Kaler et al. (2008), Time 2 was completed
approximately 8 weeks after Time 1. During the completion of the
Trauma History Questionnaire at Time 2, participants were asked,
“Since the completion of the Time 1 survey, have you experienced
a traumatic or negative life event?” If participants reported an
intervening event, then that event was used in reference to the
PCL-S and CES. If there were no reported intervening traumatic
events, then participants were asked if they experienced a stressful
life event. This procedure was approved by the Institutional Re-
view Board of the University of North Texas.
Results
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
ability at both time points (.82 and .84, respectively) and test–retest
reliability (.67). To assess convergent validity at Time 1, WAQ
scores were correlated with the DASS, PCL-S, PANAS-Negative
Affect subscale, PANAS-Positive Affect subscale, SWLS, CES,
and the NCW subscale of the PTCI (see Table 2). The number of
objective traumas was predicted to be negatively correlated with
Time 1 WAQ scores, suggesting that additional traumas could
further diminish the assumptive world (Kaler et al., 2008). There
was a small correlation between Time 1 WAQ scores and the
number of objective traumas at Time 1. The WAQ evidenced good
convergent validity based on the correlational table (see Table 2).
The WAQ had moderately negative correlations with negative
affect, NCW, and PCL-S while having a positive relationship with

the SWLS. The correlations of Time 1 WAQ scores and measures

This document is copyrighted by the American Psychological Association or one of its allied publishers.

At Time 1, 882 participants completed the online study. Of the

initial sample, 243 participants dropped out of the study (27.70%
attrition rate) at Time 2. At Time 2, 126 participants nominated a
prior traumatic event instead of an intervening event; this was
based on the fact that the dates of the events reported had occurred
before the completion of Time 1. These participants who did not
follow the instructions at Time 2 were removed from the Time 2
data analyses. During data cleaning, two participants were re-
moved from the data set because of selecting the same response on
every measure. This left a sample of 511 participants who com-
pleted Time 2. To assess for differences between completers and
noncompleters in the Time 1 data that could account for the
attrition rate, a series of independent t tests and ␹2 tests were
conducted on the measures of WAQ, CES, gender, and ethnicity.
There were no statistically significant differences between com-
pleters and noncompleters on CES, t(878) ⫽ 1.20, p ⫽ .231; on
WAQ, t(879) ⫽ .96, p ⫽ .338; or based on gender, ␹2(1) ⫽ .29,
p ⫽ .600. There was a nonsignificant trend of Caucasian partici-
pants being less likely to attrite, ␹2(5) ⫽ 11.03, p ⫽ .051. On the
basis of these analyses, there were no differences between com-
pleters and noncompleters at Time 2.
The measures of interest were assessed for internal reliability at
both time points to determine the stability of the measures. The
means, standard deviations, and correlations for the WAQ and the
measures of interest are reported in Table 2. To assess the first
hypothesis that the WAQ would have adequate internal reliability,
Cronbach’s ␣ was examined. The WAQ had good internal reli-
of SWLS, Positive Affect, number of prior traumas, and CES were
significant, but below the expected r ⱖ |.30|. Table 3 reports the
frequencies and types of prior and intervening objective traumatic
events that were nominated. In summary, the WAQ evidenced
adequate psychometric properties.
To test the second hypothesis that individuals who have expe-
rienced a prior adverse event will have a less optimistic assumptive
world (lower WAQ Time 1 scores) compared with individuals
with no prior traumatic event, we conducted two independent-
sample t tests. We classified adverse events into one of two
nonindependent groups: (a) objective traumatic events (reported
one of the events on the SLESQ) or (b) subjective traumatic events
(evidenced a CES total score ⬎73.00, the threshold for the top
quartile split). Table 4 reports the sample sizes, means, standard
deviations, and t test results for both of the groups. As noted in
Table 4, there were significant differences in WAQ scores between
the groups for both classifications.
The third hypothesis of the study was that individuals who
reported an intervening adverse event (using both nonindependent
classifications) between Time 1 and Time 2 would have less
optimistic assumptions compared with those without an interven-
ing event. At Time 2, 75 participants reported an intervening
objective trauma and 44 experienced a subjective trauma. Changes
in the WAQ between time points and groups were assessed by two
separate multiple regressions. The intervening variables of trauma
for each category were coded as 0 (no event) or 1 (self-reported

Table 2
Means, Standard Deviations, and Correlational Matrix of the Measures of Convergent Validity

M SD 1 2 3 4 5 6 7 8 9 10 11

1WAQ 75.71 12.25 —

1SWLS 23.21 6.74 .26ⴱⴱⴱ —
1DASS 12.62 11.02 ⫺.34ⴱⴱⴱ ⫺.39ⴱⴱⴱ —
1NA 22.55 7.69 ⫺.37ⴱⴱⴱ ⫺.32ⴱⴱⴱ .62ⴱⴱⴱ —
1PA 33.71 7.95 .10ⴱⴱ .37ⴱⴱⴱ ⫺.21ⴱⴱⴱ ⫺.06 —
1PCL 34.61 16.07 ⫺.35ⴱⴱⴱ ⫺.30ⴱⴱⴱ .53ⴱⴱⴱ .44ⴱⴱⴱ ⫺.02 —
1CES 56.77 21.80 ⫺.17ⴱⴱⴱ ⫺.11ⴱⴱ .27ⴱⴱⴱ .22ⴱⴱⴱ .08ⴱ .51ⴱⴱⴱ —
1NCW 70.92 34.17 ⫺.40ⴱⴱⴱ ⫺.30ⴱⴱⴱ .42ⴱⴱⴱ .34ⴱⴱⴱ ⫺.08ⴱ .50ⴱⴱⴱ .35ⴱⴱⴱ —
PT 1.67 1.60 ⫺.18ⴱⴱⴱ ⫺.21ⴱⴱⴱ .27ⴱⴱⴱ .19ⴱⴱⴱ .00 .41ⴱⴱⴱ .32ⴱⴱⴱ .39ⴱⴱⴱ —
2WAQ 78.52 12.31 .67ⴱⴱⴱ .25ⴱⴱⴱ ⫺.32ⴱⴱⴱ ⫺.28ⴱⴱⴱ .15ⴱⴱ ⫺.24ⴱⴱⴱ ⫺.16ⴱⴱⴱ ⫺.38ⴱⴱⴱ ⫺.14ⴱⴱ
2CES 44.16 21.38 .09ⴱ ⫺.13ⴱⴱ .20ⴱⴱⴱ .13ⴱⴱ .02 .29ⴱⴱⴱ .34ⴱⴱⴱ .16ⴱⴱⴱ .16ⴱⴱⴱ ⫺.16ⴱⴱⴱ —
Note. 1 ⫽ Time 1, 2 ⫽ Time 2, NA ⫽ PANAS–Negative Subscale; PA ⫽ PANAS–Positive Subscale; NCW ⫽ Posttraumatic Cognitions Inventory:
Negative Cognitions of the World; PT ⫽ Number of prior traumas.
ⴱ
p ⬍ .050. ⴱⴱ p ⬍ .010. ⴱⴱⴱ p ⬍ .001.
 264 SCHULER AND BOALS

Table 3
Frequencies of the Most Traumatic Prior and Intervening Events on the SLESQ

Prior trauma Intervening trauma

Type of traumatic or stressful life event nominations nominations

Had a threatening life illness 32 1

Life-threatening accident 69 4
Robbery/mugging 10 0
Death a family member or friend 151 8
Sexual assault 91 4
Physical abuse as a child 22 N/A
Physical abuse as an adult 31 3
Emotional abuse 106 44
Threatened with a weapon 12 2
Witnessing a death, injury, or assault 31 3
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

In a situation in which your life was in danger 18 1

This document is copyrighted by the American Psychological Association or one of its allied publishers.

In a situation in which you felt helpless 60 7

Other event 99 3
Note. The trauma category of physical abuse as a child was removed from the trauma history questionnaire at
Time 2 to reduce the chances of a Time 1 prior trauma being nominated. N/A ⫽ not applicable; SLESQ ⫽
Stressful Life Events Screening Questionnaire.

event). Regression models for each category were as follows: Time
2 WAQ scores being predicted by Time 1 WAQ scores and the
intervening trauma variable. If the intervening trauma variables
were statistically significant after controlling for Time 1 WAQ
scores, then reporting an event accounted for unique variance of
Time 2 scores. Experiencing an intervening objective trauma was
a significant predictor of Time 2 WAQ scores. The omnibus
regression results, coefficients, and squared semipartial correla-
tions are reported in Table 5.
Discussion
Empirical research to assess if experiencing a traumatic event
may shatter the assumptive world has reported mixed findings
(Kaler et al., 2008). The use of cross-sectional designs and mea-
sures with poor psychometrics can diminish the ability to detect
changes in the assumption after a trauma (Kaler et al., 2008). The
first hypothesis was to assess a new measure of the assumptive
world—the WAQ. This hypothesis was supported in that the WAQ
was reported to have good internal reliability and test–retest reli-
ability. Time 1 WAQ scores had moderate correlations with mea-
sures of negative affect, PTSD symptom severity, negative cogni-
tions, and depression. The correlations with the WAQ and
satisfaction with life and positive affect were smaller than antici-
pated, but they were significant. The small correlation between
positive affect and WAQ could be indicative of the assumptive
world as a trait construct and not directly affected by mood. Thus,
it appears the WAQ is a reliable and valid measure of the assump-
tive world.
The second hypothesis of the negative impact of prior subjective
or objective trauma on an individual’s assumptive world was
supported. The subjective measure of previous trauma had the
largest reduction in the assumptive world compared with the prior
objective trauma. This reported finding that prior subjective or
objective traumas can negatively impact an individual’s assump-
tive world is consistent with previous studies assessing differences
in the assumptive world based on objective measures of trauma
(Janoff-Bulman, 1989, 1992; Park, Mills, & Edmondson, 2012).
Individuals who had subjective traumas had lower optimism in
their assumptive world compared with individuals with low event
centrality. This suggests the importance of the individual’s sub-
jective experience in trauma research. Furthermore, this finding
reflects that the assumptive world can be affected by a broad range
of traumatic events.
The third hypothesis that an intervening traumatic event would
result in a reduction of the optimism of the assumptive world was
partially supported. There were reported small effects of the inter-
vening objective traumas explaining the variance of the Time 2
WAQ scores beyond that of Time 1 scores. Experiencing a sub-
jective traumatic event was trending toward statistical significance
(p ⫽ .060), which could have been due to the small number of

Table 4
Comparing Assumptive Worlds Between Prior and No Prior Events Using the Classifications of
Subjective and Objective Trauma

Group N M (SD) t df Cohen’s d

ⴱⴱ
Prior objective trauma 732 75.14 (12.39) 3.09 879 .28
No prior objective trauma 149 78.52 (11.17)
Prior subjective trauma 209 71.98 (13.25) 5.11ⴱⴱⴱ 877 .41
No prior subjective trauma 672 76.87 (11.70)
Note. Statistical analyses were independent-sample t tests.
ⴱⴱ
p ⬍ .010. ⴱⴱⴱ p ⬍ .001.
 SHATTERED ASSUMPTIONS
Table 5
Variance of Time 2 WAQ Total Scores by Time 1 Scores and Intervening Trauma Variable
DV Predictor F statistic
ⴱⴱⴱ
T2 WAQ T1 WAQ F(2, 508) ⫽ 215.65
Objective
T2 WAQ T1 WAQ F(2, 508) ⫽ 209.94ⴱⴱⴱ
Subjective
Note. DV ⫽ dependent variable; T1 ⫽ Time 1; T2 ⫽ Time 2; WAQ ⫽ World Assumptions Questionnaire;
Objective ⫽ intervening objective trauma variable; Subjective ⫽ intervening subjective trauma; Radj
R2; b ⫽ unstandardized coefficient; SE ⫽ standard error of the unstandardized coefficient; B ⫽ ␤ weight;
SSPC ⫽ squared semipartial correlation; Trauma ⫽ intervening trauma reported between Time 1 and Time 2.
ⴱⴱ
p ⬍ .010. ⴱⴱⴱ p ⬍ .001.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
intervening highly subjective traumas (n ⫽ 44). Alternatively, it
could be that not enough time had passed since the intervening
subjective trauma for it to become a highly central life event. The
effect sizes were small, but there was unique variance explained by
the intervening trauma variables. On the basis of these results,
experiencing intervening objective trauma was related to reduc-
tions in the assumptive world. This pattern of findings is consistent
with Shattered Assumptions theory.
There were some important limitations in the current study to
consider. There was a large diversity of the types of traumas;
therefore, no direct comparisons of which assumptions were shat-
tered by specific adverse events or the impact of childhood trau-
mas could be made. Although the use of an undergraduate conve-
nience sample is a limitation, individuals in this sample are not
knowingly at high risk for experiencing a traumatic event (e.g.,
soldiers and first responders may expect to encounter potentially
traumatic situations; Frazier et al., 2011). Our replication of the
8-week interim of Kaler et al. (2008) and Anders et al. (2014)
between time points may not have been enough time for partici-
pants to potentially experience an intervening trauma; as such, few
participants reported experiencing an event. This interim allowed
us to collect data from our sample during the same semester as
Time 1; however, future research could benefit from a longer
interim. At the end of data collection, the field transitioned from
the DSM–IV–TR to the DSM–5; as such, we were unable to use the
updated criteria. The fact that we found that prior subjective
traumas were reported to negatively impact the WAQ scores at
Time 1 suggested that it is important to consider how these broader
traumas can also shatter the assumptive world. An additional
limitation was that the WAQ total scores were used instead of the
four WAQ subscales. The specific nature of the trauma could
affect some assumptions whereas others are unchanged (Janoff-
Bulman, 1992), which may obscure effects when looking at vari-
ous traumas or using total scores. However, there were significant
differences in the assumptive world when using a summed total
score to assess the impact of prior and intervening adverse events.
There was no direct comparison between the WAQ and the WAS
based on the findings of Kaler et al. (2008). Future studies should
assess the WAQ factor structure and test whether the type of
trauma can affect certain assumptions. Our results supported that
an adverse event can affect the assumptive world, but the effect
sizes were smaller than what was expected based on the theory
(Janoff-Bulman, 1989, 1992). It is important to note that changes
on measures pre- and postevent in longitudinal designs are gener-
ally small (Anders et al., 2014; Bonanno et al., 2002). On the basis
265
2
Radj b SE B SSPC
ⴱⴱⴱ
.46 .67 .03 .66 .44
⫺3.22ⴱⴱ 1.13 ⫺.09 .01
.45 .67ⴱⴱⴱ .03 .66 .44
⫺2.72 1.43 ⫺.06 .00
2
⫽ adjusted
of this finding, the assumptive world may be weakened rather than
shattered by a trauma.
The current study offers empirical evidence of the shattering of
the assumptive world after an intervening objectively traumatic
event using a prospective design. There were significant differ-
ences in the assumptive world based on having a prior trauma
history, which replicated prior studies (Janoff-Bulman, 1989). It is
interesting to note that prior subjective events had a larger effect
compared with subjective intervening events. A subjective trauma
potentially has an initially small impact and grows larger over time
if it becomes a central part to the life story. This suggests that
research should consider the impact of subjective traumas on the
assumptive world. Clinicians can work with clients to specifically
rebuild the assumptive world after a traumatic event. Prolonged
Exposure (PE) therapy (Peterson, Foa, & Riggs, 2011) and Cog-
nitive Processing Therapy (CPT; Williams, Galovski, Kattar, &
Resick, 2011) can help individuals rebuild their assumptive world.
Part of PE focuses on reducing of negative cognitions (e.g.,” I
deserve what happened to me”) by having the individual become
habituated to the event (Peterson et al., 2011). CPT (Williams et
al., 2011) focuses on altering incongruent information of the
trauma with the pretrauma beliefs, which parallels Shattered As-
sumptions. Onsite clinicians could work with groups at high risk
for experiencing a trauma (e.g., first responders) to develop flex-
ibility in their beliefs to become more resilient. By better under-
standing how subjective and objective traumatic events can impact
assumptions of the world, clinical treatments can be improved to
better rebuild these shattered assumptions.
References
American Psychiatric Association. (2000). Diagnostic and statistical man-
ual of mental disorders (4th ed., text revision). Washington, DC: Author.
American Psychiatric Association. (2013). Diagnostic and statistical man-
ual of mental disorders (5th ed.). Washington, DC: Author.
Anders, S. L., Frazier, P. A., & Shallcross, S. L. (2014). Changes in
functioning following potentially traumatic life events in college stu-
dents. Psychological Trauma: Theory, Research, Practice, and Policy,
6, S99 –S106.
Antony, M. M., Bieling, P. J., Cox, B. J., Enns, M. W., & Swinson, R. P.
(1998). Psychometric properties of the 42-item and 21-item versions of
the Depression Anxiety Stress Scales in clinical groups and a community
sample. Psychological Assessment, 10, 176 –181.
Beck, J. G., Coffey, S. F., Palyo, S. A., Gudmundsdottir, B., Miller, L. M.,
& Colder, C. R. (2004). Psychometric properties of the Posttraumatic
Cognitions Inventory (PTCI): A replication with motor vehicle accident
 266 SCHULER AND BOALS
survivors. Psychological Assessment, 16, 289 –298. http://dx.doi.org/
10.1037/1040-3590.16.3.289
Berntsen, D., & Rubin, D. C. (2006). The Centrality of Event Scale: A
measure of integrating a trauma into one’s identity and its relation to
post-traumatic stress disorder symptoms. Journal of Behavior Research
and Therapy, 44, 219 –231. http://dx.doi.org/10.1016/j.brat.2005.01.009
Boals, A., & Schuettler, D. (2009). PTSD symptoms in response to trau-
matic and non-traumatic events: The role of respondent perception and
A2 criterion. Journal of Anxiety Disorders, 23, 458 – 462. http://dx.doi
.org/10.1016/j.janxdis.2008.09.003
Boals, A., Southard-Dobbs, S., & Blumenthal, H. (2015). Adverse events
in emerging adulthood are associated with increases in neuroticism.
Journal of Personality, 83, 202–211. http://dx.doi.org/10.1111/jopy
.12095
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Bonanno, G. A., Wortman, C. B., Lehman, D. R., Tweed, R. G., Haring,
This document is copyrighted by the American Psychological Association or one of its allied publishers.
M., Sonnega, J., . . . Nesse, R. M. (2002). Resilience to loss and chronic
grief: A prospective study from preloss to 18-months postloss. Journal
of Personality and Social Psychology, 83, 1150 –1164. http://dx.doi.org/
10.1037/0022-3514.83.5.1150
Breslau, N., Chilcoat, H. D., Kessler, R. C., & Davis, G. C. (1999).
Previous exposure to trauma and PTSD effects of subsequent trauma:
Results from the Detroit Area Survey of Trauma. The American Journal
of Psychiatry, 156, 902–907. http://dx.doi.org/10.1176/ajp.156.6.902
Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The
Satisfaction With Life Scale. Journal of Personality Assessment, 49,
71–75. http://dx.doi.org/10.1207/s15327752jpa4901_13
Edmondson, D., Chaudoir, S. R., Mills, M. A., Park, C. L., Holub, J., &
Bartkowiak, J. M. (2011). From shattered assumptions to weakened
worldviews: Trauma symptoms signal anxiety buffer disruption. Journal
of Loss and Trauma, 16, 358 –385. http://dx.doi.org/10.1080/15325024
.2011.572030
Foa, E. B., Elhers, A., Clark, D. M., Tolin, D. F., & Orsillo, S. M. (1999).
The Posttraumatic Cognitions Inventory (PTCI) development and vali-
dation. Psychological Assessment, 11, 303–314. http://dx.doi.org/
10.1037/1040-3590.11.3.303
Frazier, P., Gavian, M., Hirai, R., Park, C., Tennen, H., Tomich, P., &
Tashiro, T. (2011). Prospective predictors of Posttraumatic Stress Dis-
order symptoms: Direct and mediated relations. Psychological Trauma:
Theory, Research, Practice, and Policy, 3, 27–36. http://dx.doi.org/
10.1037/a0019894
Ginzburg, K. (2004). PTSD and world assumptions following myocardial
infarction: A longitudinal study. American Journal of Orthopsychiatry,
74, 286 –292. http://dx.doi.org/10.1037/0002-9432.74.3.286
Goodman, L. A., Corcoran, C., Turner, K., Yuan, N., & Green, B. L.
(1998). Assessing traumatic event exposure: General issues and prelim-
inary findings for the Stressful Life Events Screening Questionnaire.
Journal of Traumatic Stress, 11, 521–542. http://dx.doi.org/10.1023/A:
1024456713321
Janoff-Bulman, R. (1989). Assumptive worlds and the stress of traumatic
events: Applications of the scheme construct. Social Cognition, 7, 113–
136. http://dx.doi.org/10.1521/soco.1989.7.2.113
Janoff-Bulman, R. (1992). Shattered assumptions: Towards a new psychol-
ogy of trauma. New York, NY: Free Press.
Kaler, M. E. (2009). The World Assumptions Questionnaire: Development
of a measure of the assumptive world. Doctoral dissertation. Retrieved
from ProQuest. (Order No. 3371864)
Kaler, M. E., Frazier, P. A., Anders, S. L., Tashiro, T., Tomich, P., Tennen,
H., & Park, C. (2008). Assessing the psychometric properties of the
World Assumptions Scale. Journal of Traumatic Stress, 21, 326 –332.
http://dx.doi.org/10.1002/jts.20343
Lovibond, P. F., & Lovibond, S. H. (1995). The structure of negative
emotional states: Comparison of the Depression Anxiety Stress Scales
(DASS) with the Beck Depression and Anxiety Inventories. Behaviour
Research and Therapy, 33, 335–343. http://dx.doi.org/10.1016/0005-
7967(94)00075-U
McFarland, C., & Alvaro, C. (2000). The impact of motivation on temporal
comparisons: Coping with traumatic events by perceiving personal
growth. Journal of Personality and Social Psychology, 79, 327–343.
http://dx.doi.org/10.1037/0022-3514.79.3.327
Park, C. L., Mills, M. A., & Edmondson, D. (2012). PTSD as meaning
violation: Testing a cognitive worldview perspective. Psychological
Trauma: Theory, Research, Practice, and Policy, 4, 66 –73. http://dx.doi
.org/10.1037/a0018792
Pavot, W., & Diener, E. (1993). Review of the Satisfaction with Life Scale.
Psychological Assessment, 5, 164 –172. http://dx.doi.org/10.1037/1040-
3590.5.2.164
Peterson, A. L., Foa, E. B., & Riggs, D. S. (2011). Prolonged exposure
therapy. In B. A. Moore & W. E. Penk (Eds.), Treating PTSD in military
personnel: A clinical handbook (pp. 42–58). New York, NY: Guilford
Press.
Rini, C., Manne, S., DuHamel, K. N., Austin, J., Ostroff, J., Boulad, F., . . .
Redd, W. H. (2004). Changes in mothers’ basic beliefs following a
child’s bone marrow transplantation: The role of prior trauma and
negative life events. Journal of Traumatic Stress, 17, 325–333. http://
dx.doi.org/10.1023/B:JOTS.0000038481.17167.0d
Tomich, P. L., & Helgeson, V. S. (2002). Five years later: A cross-
sectional comparison of breast cancer survivors with healthy women.
Psycho-Oncology, 11, 154 –169. http://dx.doi.org/10.1002/pon.570
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and vali-
dation of brief measures of positive and negative affect: The PANAS
scales. Journal of Personality and Social Psychology, 54, 1063–1070.
http://dx.doi.org/10.1037/0022-3514.54.6.1063
Weathers, F., Litz, B., Herman, D., Huska, J., & Keane, T. (1993). The
PTSD Checklist (PCL): Reliability, validity, and diagnostic utility. Paper
presented at the annual convention of the International Society for
Traumatic Stress Studies, San Antonio, TX.
Wilkins, K. C., Lang, A. J., & Norman, S. B. (2011). Synthesis of the
psychometric properties of the PTSD checklist (PCL) military, civilian,
and specific versions. Depression and Anxiety, 28, 596 – 606. http://dx
.doi.org/10.1002/da.20837
Williams, A. M., Galovski, T. E., Kattar, K. A., & Resick, P. A. (2011).
Cognitive processing therapy. In B. A. Moore & W. E. Penk (Eds.),
Treating PTSD in military personnel: A clinical handbook (pp. 59 –73).
New York, NY: Guilford Press.
Received May 20, 2014
Revision received June 8, 2015
Accepted June 13, 2015 䡲