Assessing posttraumatic cognitive processes: the Event Related
Arnie Cann*, Lawrence G. Calhoun, Richard G. Tedeschi, Kelli N. Triplett,
Tanya Vishnevsky and Cassie M. Lindstrom
Department of Psychology and the Interdisciplinary Health Psychology Doctoral Program,
University of North Carolina Charlotte, 9201 University City Blvd., Charlotte, NC 28223, USA
(Received 6 May 2010; final version received 4 October 2010)
Cognitive processes in the aftermath of experiencing a major life stressor play an
important role in the impact of the event on the person. Intrusive thoughts about
the event are likely to be associated with continued distress, while deliberate
rumination, aimed at understanding and problem-solving, should be predictive of
posttraumatic growth (PTG). The Event Related Rumination Inventory (ERRI),
designed to measure these two styles of rumination, is described and validation
information is provided. Using a college student sample screened for having
experienced highly stressful life events, data were obtained (N?323) to conduct
an exploratory factor analysis that supported the two factors of the ERRI.
Separate confirmatory factor analyses (CFA) on two additional samples
(Ns ?186 and 400) supported a two-factor model. The two ERRI factors were
validated by comparison with related variables and by assessing their contribu-
tions to predicting distress and PTG in two samples (Ns ?198 and 202) that had
been combined to conduct the second CFA. Data indicate the ERRI has solid
psychometric properties, captures variance not measured by stable differences in
cognitive styles, and the separate factors are related to posttraumatic distress and
growth as predicted by existing models of PTG.
Keywords: posttraumatic growth; posttraumatic distress; rumination styles;
Highly stressful and traumatic events can lead to distressing posttraumatic
symptoms, but the struggle with major life crises can also lead to the experience
of posttraumatic growth (PTG; Joseph & Linley, 2008; Tedeschi & Calhoun, 1995).
Two of the main elements that can lay the foundation for the possibility of growth
are the degree to which the event presents a significant challenge to, or invalidates
components of, the individual’s assumptive world, and the degree to which this
challenge initiates cognitive processes that focus on the life crisis and its implications
(Cann et al., 2009; Janoff-Bulman, 1992).
When a major stressful event seriously challenges individuals’ beliefs about how
the world works and their place in the world, they have lost their framework for
understanding. However, through the process of attempting to understand the event
and rebuilding those core beliefs about the world individuals are provided with the
opportunity for realizing growth. Effective cognitive work that confronts the
*Corresponding author. Email: email@example.com
Anxiety, Stress, & Coping
Vol. 24, No. 2, March 2011, 137?156
ISSN 1061-5806 print/1477-2205 online
# 2010 Taylor & Francis
challenged beliefs can help restore or revise the assumptive world and allow the
person to appreciate how they have been challenged and changed by the experience
of a major crisis (Janoff-Bulman, 2006).
The term rumination has acquired a negative connotation in recent years,
particularly in the clinical literature on depression and posttraumatic stress disorder
(PTSD), where the word has been used by some to mean only self-focused negative
thinking about symptoms (e.g., Nolen-Hoeksema, McBride, & Larson, 1997) or it
has been defined as closely related to worry (Ehring, Frank, & Ehlers, 2008; Michael,
Halligan, Clark, & Ehlers, 2007). However, rumination as it is generally defined
simply means repetitive thought, pondering or meditating on information, essentially
a cognitive ‘‘chewing the cud.’’ Multiple authors have recognized that ruminative
thought can take alternative forms and attempts have been made to identify stable
differences in ruminative thinking (Scott & McIntosh, 1999; Segerstrom, Stanton,
Alden, & Shortridge, 2003; Treynor, Gonzalez, & Nolen-Hoeksema, 2003; Wyer,
1996). Ruminative thoughts precipitated by a highly stressful event can include a
variety of different types of recurrent thinking; these can include intrusive thoughts
that are often undesired, like those commonly associated with rumination as a
symptom of distress, but they can also include more controlled thoughts focused on
making sense of the experience, problem-solving, reminiscence, and anticipation
(Martin & Tesser, 1996; Watkins, 2008). We propose that in order to understand the
processes related to the experience of PTG and posttraumatic distress, it is important
to distinguish two major types of rumination: intrusive and deliberate. Furthermore,
assessing rumination more neutrally, rather than with a focus on negative content,
will help clarify its role in PTG.
It is important to recognize and appreciate the difference between these styles of
ruminative thought and the symptoms of posttraumatic distress that have been
labeled by some as rumination. Although repetitive thinking about psychological
symptoms and the sources of the symptoms (e.g., Nolen-Hoeksema, 1991) or worry
associated with a negative experience (e.g., Michael et al., 2007) can be a contributor
to continued distress, all rumination is not negative. Even intrusive thoughts in the
aftermath of a significant life event, which are often seen as symptoms of distress, are
an expected and normal response to the event. We would expect intrusive rumination
even in cases where the significant life event was a positive experience. A first
encounter with someone you find highly desirable, for example, can result in intrusive
thoughts as you experience difficulty in ‘‘getting him/her out of your head.’’ While it
is important to understand how some forms of rumination might be detrimental to
well-being, it also is important to recognize that all forms of rumination are not
negative, and that some intrusive thoughts are quite natural after significant
Intrusive ruminations are, by definition, unsolicited invasions of one’s cognitive
world-thoughts about an experience that one does not choose to bring to mind.
Deliberate ruminations about events, however, are engaged in voluntarily and can be
focused purposefully on trying to understand events and their implications. In one of
the primary models of PTG (Calhoun, Cann, & Tedeschi, 2010), it has been
proposed that intrusive and deliberate thoughts play different roles in influencing
outcomes following a highly stressful experience. Although there is limited empirical
evidence examining the two styles of ruminative thought, what is available suggests
that the distinction between deliberate and intrusive rumination is important not
A. Cann et al.
only conceptually, but also empirically. These different forms of ruminative thought
have been found to differ in their relationships with PTG and with posttraumatic
distress. Event-related deliberate rumination, that is not negatively focused, is more
likely to be related to eventual PTG, and event-related intrusive rumination, which is
not controlled by the individual, is more likely to be related to various kinds of
posttraumatic distress (Affleck & Tennen, 1996; Calhoun, Cann, Tedeschi, &
McMillan, 2000; Cann, Calhoun, Tedeschi, & Solomon, 2010; Taku, Calhoun,
Cann, & Tedeschi, 2008). In addition, it is assumed that the level of intrusive
thoughts will be predictive of the level of deliberate thoughts, since intrusive thoughts
are a precursor leading the individual to seek a better understanding of the stressful
experience (Calhoun et al., 2010). Identifying the impact and correlates of these two
styles of event-related thought is important for the understanding of posttraumatic
adaptation processes, especially those leading to the experience of PTG.
Given that intrusive thoughts are expected to occur as a normal by-product of a
major life event, it is desirable to have an instrument that assesses the presence of
these thoughts without implying posttraumatic symptoms. For example, when
intrusive thoughts are viewed as symptoms of distress they are typically associated
with the re-experiencing of the event, having flashbacks, and creating difficulty in
sleeping (Elwood, Hahn, Olatunji, & Williams, 2009). A commonly used measure of
PTSD symptoms, the Impact of Event Scale ? Revised (IES-R; Weiss & Marmar,
1997), includes a factor assessing intrusive event-related rumination; however, it
includes items that involve difficulties with sleeping and unpleasant dreams, so it
does not provide the more neutral assessment of intrusive thoughts that could occur
naturally without implying PTSD symptoms. In addition, the IES-R does not assess
deliberate, more reflective rumination and the scale items assessing rumination are
presented in the context of others focused on other symptoms of distress. Our goal is
to devise a measure that captures the two rumination styles, without evaluative bias
and without mixing in thoughts that specifically reflect PTSD symptoms. Such a
measure would likely have some overlap with measures like the IES-R, but would
capture some unique variance as well.
There are available scales that measure stable dispositions of the tendency to
ruminate in different ways, including more reflective or deliberate rumination. The
Ruminative Responses Scale, for example, measures the tendency to negatively brood
or to neutrally ponder about one’s thoughts or feelings in response to negative mood
or depression (Treynor et al., 2003). The Rumination-Reflection Questionnaire
(Trapnell & Campbell, 1999) assesses the general disposition to be ruminatively self-
focused, dwelling on past threats or losses, or reflectively self-focused, seeking
meaning or more global understanding of the self. The content of these two scales,
and others like them, successfully distinguish between the kinds of habitual repeated
thought that are essentially deliberate or intrusive; however, the scales do not assess
transient, event-provoked thinking about a specific highly stressful or disruptive
event. It would seem that when the focus of investigation is on repeated thinking
about a specific set of circumstances, such as a major life crisis, an instrument that is
focused on transient event-related processes will provide more useful and specific
information than scales that measure general response tendencies or dispositions.
Although groups of items that intercorrelate have been developed to assess event-
related intrusive and deliberate ruminations (Calhoun et al., 2000), we are aware of
no published scale that has solid psychometric properties, ensuring that the two
Anxiety, Stress, & Coping
styles are being separately assessed as they apply to repeated thought about a specific
A scale focused on repeated thinking about a specific event would allow for the
investigation of the role of more transitory cognitive states, both in the development
of PTG and of posttraumatic distress. No matter what one’s stable predisposition to
engage in intrusive or deliberate thinking, a major life crisis is likely to temporarily
provoke both (Calhoun et al., 2000; Taku, Cann, Tedeschi, & Calhoun, 2009). We
describe here the development of such a scale: the Event Related Rumination
Inventory (ERRI). An important general question for the development of an
inventory designed to assess specific event-related rumination, is whether such a scale
does indeed add information about post-event processes beyond what the scales that
assess stable ruminative styles already provide. In addition, the scale should clearly
show the expected relationships with posttraumatic distress and PTG. Intrusive
thoughts are expected to be more strongly related to levels of ongoing distress than to
levels of PTG, especially if they occur well after the experienced event. Alternatively,
deliberate rumination that is not negatively focused, but reflects potential problem-
solving and meaning-seeking, should be more predictive of PTG. In addition, levels
of intrusive thoughts should be positively related to levels of deliberate thoughts,
since it is believed that the intrusive thoughts in the aftermath of a stressful
experience serve to stimulate attempts to engage in more deliberate processing of
Study 1: development of the Event Related Rumination Inventory (ERRI)
The original set of event-related rumination items was developed for a study of
religious beliefs, cognitive processing, and growth (Calhoun et al., 2000). The items
were drawn from avariety of sources, including items modified from published scales
(e.g., Barrett & Scott, 1989; Sanavio, 1988) and items created based on ideas
suggested in the trauma literature (e.g., Tennen & Affleck, 1998; Yalom &
Lieberman, 1991). The items were created or reworded to reflect either intrusive or
deliberate forms of repetitive thinking about a highly stressful event. A revision of
the items was required after their evaluation with a larger sample indicated that not
all the original items loaded well on the expected dimensions (Taku et al., 2009).
In formulating the current versions of the items, an effort was made to exclude
any clearly positive or negative implications of the different thoughts. With the
original items as a starting point, various sets of items were tested to insure clarity
and to insure internal consistency. Ultimately, based on several pretest results and the
judgments of the three lead researchers and a group of doctoral students engaged in
research on psychological responses to highly stressful or traumatic events, 10 items
were chosen for each of the two rumination styles (see Table 1). Separate instructions
preceded each set of items to reinforce the participants’ appreciation for the type of
cognitive activity that was being assessed. Participants rated the degree to which the
thoughts occurred during a specified time frame on a 4-point scale (not at all ?0 to
3?often). In the initial samples used to assess the factor structure of the ERRI, the
time frame specified was the ‘‘weeks immediately after the event.’’
A. Cann et al.
To determine, and to confirm, the factor structure of the ERRI, the two sets of 10
items were administered to participants as part of separate online studies of PTG
processes not relevant to the current report. In each of these cases, participants
based on the exploratory factor analysis.
Items included in the Event Related Rumination Inventory with factor loadings
After an experience like the one you reported, people sometimes, but not always, find
themselves having thoughts about their experience even though they don’t try to think
about it. Indicate for the following items how often, if at all, you had the experiences
described during the weeks immediately after the event.
.169 I thought about the event when I did not mean to.
.241 Thoughts about the event came to mind and I could not stop thinking
.242 Thoughts about the event distracted me or kept me from being able to
.238 I could not keep images or thoughts about the event from entering my
.280 Thoughts, memories, or images of the event came to mind even when I did
not want them.
.232 Thoughts about the event caused me to relive my experience.
.260 Reminders of the event brought back thoughts about my experience.
.225 I found myself automatically thinking about what had happened.
.302 Other things kept leading me to think about my experience.
.315 I tried not to think about the event, but could not keep the thoughts from
After an experience like the one you reported, people sometimes, but not always,
deliberately and intentionally spend time thinking about their experience. Indicate for
the following items how often, if at all, you deliberately spent time thinking about the
issues indicated during the weeks immediately after the event.
.687 I thought about whether I could find meaning from my experience.
.731 I thought about whether changes in my life have come from dealing with
.621 I forced myself to think about my feelings about my experience.
.742 I thought about whether I have learned anything as a result of my
.682 I thought about whether the experience has changed my beliefs about the
.583 I thought about what the experience might mean for my future.
.681 I thought about whether my relationships with others have changed
following my experience.
.692 I forced myself to deal with my feelings about the event.
.569 I deliberately thought about how the event had affected me.
.568 I thought about the event and tried to understand what happened.
Note: ‘‘Int’’ refers to the intrusive dimension and ‘‘Del’’ to the deliberate dimension. N?323.
Anxiety, Stress, & Coping
were responding to a specific highly stressful event that they had experienced. The
ERRI data were collected for the purpose of examining the psychometric
properties of the ERRI, so the other variables included in the surveys are not
Participants ? exploratory factor analysis
Participants were students enrolled in psychology courses. At the beginning of the
semester, all eligible students could take a common pretest. As part of that pretest,
students indicated if they had experienced any of a preselected set of nine highly
stressful events (death of close other, serious medical issue for self, or for close other,
accident resulting in serious injury, residence seriously damaged, victim of assault,
victim of robbery, been stalked, and divorce) within a specified time period (within 3
years). Only students who indicated ‘‘yes’’ to one of the events were eligible to
participate in the research. If more than one event had been experienced, participants
were instructed to respond based on the most stressful event experienced in the
designated time period. The final sample used for the exploratory factor analysis
(N?323) included 221 women and 102 men with an average age of 21.45 years
(range from 18 to 60 years) who were predominantly Caucasian (70%). The stressful
event had taken place an average of 223.8 days ago. The most common stressful
events that were reportedwere: serious medical problem (42%), death of a close other
(34%), and victim of assault (11%).
Participants ? confirmatory factor analysis (CFA)
A separate sample of participants was used for a confirmatory factor analysis (CFA).
Participants were students enrolled in psychology courses. Using the same pretesting
procedure described above, students were screened for exposure to a stressful event in
order to be eligible. The sample (N?186) included 141 women and 45 men, with an
average age of 21.76 years (range from 19 to 58 years) who were predominantly
Caucasian (68%). The stressful event had taken place an average of 334.4 days ago.
The most common stressful events that were reported were: death of a close other
(47%), serious medical problem (25%), victim of assault (10%), and serious accident
with injury (7%).
Exploratory factor analysis
The 20 items were subjected to an exploratory factor analysis with a varimax
rotation. Using parallel analysis for identifying the number of components to
extract (O’Connor, 2000), it was determined that the results supported the expected
two-factor solution, with all items loading best on the expected factor. In addition,
only the first two factors produced eigenvalues above 1. The items and factor
loadings are presented in Table 1. The two factors accounted for 57% of the
variance, with the intrusive factor accounting for 32% (eigenvalue ?6.394) and the
deliberate factor accounting for 25% (eigenvalue ?4.980) of the variance based on
the rotated factors. As would be expected, the internal consistencies were strong
A. Cann et al.
(intrusive a?.94, deliberate a?.88). Although the two separate factors clearly
emerged, they were, as expected, correlated (r?.60), indicating that those who tend
to experience more intrusive thoughts also tend to engage in more deliberate
Confirmatory factor analysis (CFA)
The anticipated two-factor model was compared to a single-factor model. No other
alternative model was suggested by theory or by the exploratory factor analysis. The
single-factor model indicated a potential fit, x2(170) ?678.90, pB.01, with the fit
indices above .90 (Normed Fit Index (NFI)?.91; Comparative Fit Index
(CFI)?.93), but the root mean square error of approximation (RMSEA) (.18, 90%
CI [.17?.19]) was relatively high. The two factor model produced a far superior fit, x2
(169) ?305.09, pB.01, with all fit statistics meeting recommended levels (NFI?.96;
CFI?.98; RMSEA?.061, 90% CI [.048?.073]). Thus, the CFA supports the
presence of two factors; one capturing intrusive thoughts and the other more
deliberate rumination. Once again, the two factors were correlated (r?.61).
Study 2: evaluation of the Event Related Rumination Inventory (ERRI)
To further evaluate the ERRI, data were collected from two additional samples.
Having established that the ERRI has two factors and good psychometric properties,
the next step was to show that it accounts for variance in predicting outcomes after
exposure to a highly stressful event. Along with the ERRI, participants completed
other measures relevant to the posttraumatic cognitive processes, especially as
hypothesized in the current models (Calhoun & Tedeschi, 2006; Janoff-Bulman,
1992, 2006; Linley & Joseph, 2004). In addition, the ERRI was compared to an
existing measure of stable tendencies to ruminate, to personality measures that might
be related to the two rumination styles, and to stable differences in styles of coping. It
was expected that the ERRI would be only weakly related to stable personality
differences, stable differences in rumination styles, and stable differences in coping
styles since it focuses on event-stimulated cognitive processing, that should occur
regardless of existing stable differences.
Deliberate rumination, in the aftermath of a highly stressful experience, should
be most strongly, and positively, related to PTG, while intrusive rumination should
be positively related to symptoms of distress (Affleck & Tennen, 1996; Calhoun &
Tedeschi, 2006; Janoff-Bulman, 1992). It also is possible that continued high levels of
deliberate rumination long after the event could be related to distress, since that
pattern would suggest a failure to completely process and deal with the experience.
The measures obtained from the two samples were identical except for the time
frame requested for the ERRI ratings. In one sample, participants were instructed to
rate their thoughts in ‘‘the weeks immediately after the event.’’ In the other sample,
participants reported on their thoughts ‘‘during the last couple of weeks.’’ Although
the ideal situation would be to track changes in cognitive processes longitudinally,
these two samples allow for a preliminary investigation of the usefulness of the ERRI
for assessing thought processes currently versus retrospectively.
Anxiety, Stress, & Coping
To begin, 938 students completed the pre-screening measure and of those, 465
reported having experienced one of the designated highly stressful events in the past
6?8 months. Of those 465 who were eligible 208 volunteered to complete the online
survey, but 10 were excluded from the analyses. Nine of the 10 chose to report on an
event that had occurred longer ago than the designated time frame when they
completed the survey, and one rated the event as not at all stressful. The usable sample
of participants (N?198), who reported about rumination immediately after the
event, included 129 women and 69 men, with amean age of 21.64years (range from 18
to 52 years). The sample was predominantly Caucasian (66.7%; 21.2% African-
American, 6.1% Asian American, and 6% other). The critical event had occurred
within the past 240 days. To assess the immediate impact of the event, participants
rated the event on its stressfulness at the time and the extent to which the event
provoked extreme fear or horror on 7-point scales (1 not at all to 7 extremely). On
both scales, the ratings averaged above 5 (stressfulness: M?5.94, SD ?1.25; fear or
horror: M?5.20, SD ?1.71),indicating that the eventswerestressfullife experiences.
For the second sample 1065 students completed the pre-screening measure, and
of those 506 reported having experienced one of the designated stressful events in the
past 6?8 months. Of those, 215 completed the online survey, but 13 were excluded
from the analyses. Eight of the 13 chose to report on an event that had occurred
longer ago than the designated time frame, three rated the event as not at all
stressful, one failed to describe the event they were reporting about, and one listed
two events rather than one. The usable sample of participants (N?202), who were
reporting about rumination recently (in the last couple of weeks), included 117
women and 85 men, with a mean age of 21.26 years (range from 18 to 50 years). The
sample was predominantly Caucasian (60.4%; 19.8% African-American, 5.9% Asian
American, 5.4% Latina/Latino, and 8.4% other). The critical event had occurred
within the past 210 days. The ratings of stressfulness of the event and the fear or
horror experienced averaged at least 5 (stressfulness: M?5.81, SD ?1.22; fear or
horror: M?5.00, SD ?1.66).
The pre-screening assessment was changed for use with these two samples to
include 12 possible events. The events, along with the number of participants
reporting on that type of event, were: unexpected or violent death of close other (96),
personally experienced very serious medical problem (31), close other experienced
very serious medical problem (176), accident leading to serious injury to you or close
other (40), place of residence seriously damaged (2), exposed to threat of death or
serious bodily harm (14), witnessed severe assault of close other (5), victim of severe
physical or sexual assault (4), intimate partner violence (6), victim of crime such as
robbery or mugging (15), you were stalked (10), and deployed with the military to an
active combat zone (1).
Materials in addition to the Event Related Rumination Inventory (ERRI)
Participants provided demographic information such as age, gender, and ethnicity.
Participants indicated which of the 12 events they were reporting about, the
A. Cann et al.
estimated days since the stressful event, and also rated the stressfulness of the event
and experience of horror or fear on the 7-point scales described above.
Posttraumatic Growth Inventory (PTGI)
Participants completed the Posttraumatic Growth Inventory (PTGI; Tedeschi &
Calhoun, 1996), a 21-item measure assessing the extent to which individuals believe
they have changed positively as a result of a stressful experience. Scoring for the
PTGI consists of a total score although the scale does capture growth on five
dimensions: spiritual change, new possibilities in life, feeling stronger as a person,
having a greater appreciation for life, and positive changes in the way survivors relate
to others. The response scale ranges from 0 (I did not experience this change) to 5
(I experienced this change to a great degree). Scores are presented as means across
the 21 items. The PTGI has been found to have appropriate internal consistency
(a?.90), test?retest reliability over a two-month interval (r?.71), scores are not
correlated with social desirability and responses tend to be corroborated by others
(Shakespeare-Finch & Enders, 2008). Within the current samples the reliability was
very good (a?.94).
Core Beliefs Inventory (CBI)
Participants completed the Core Beliefs Inventory (CBI; Cann et al., 2010), a 9-item
scale designed to measure the degree to which a specific stressful event challenged
participants’ core beliefs about their world. In longitudinal and cross-sectional
research, the CBI has been shown to be strongly related to subsequent PTGI scores.
Responses are made on a scale from 0 (not at all) to 5 (to a very great degree). The
measure has shown good internal reliability (a?.82) and acceptable test?retest
reliability over a 60 to 74 day time period (r?.69) (Cann et al., 2010). Across the
current samples the reliability was acceptable (a?.85).
The Impact of Events Scale ? Revised (IES-R)
The IES-R (Weiss & Marmar, 1997) was used to assess the level of distress associated
with posttraumatic symptoms experienced during the past 7 days. The scale includes
22 items and can be used to produce a total score or scores can be calculated for three
symptom subscales: intrusion, avoidance, and hyperarousal. Responses are on a 0
(not at all) to 5 (often) scale. The internal consistency of the measure ranges from .79
to .92 for the three subscales with test?retest reliability ranging from .51 to .94 (Weiss
& Marmar, 1997). In the current samples the total score for the IES-R had strong
internal reliability (a?.94), and all subscales had reliabilities of .85 or higher.
Rumination-Reflection Questionnaire (RRQ)
The Rumination-Reflection Questionnaire (RRQ; Trapnell & Campbell, 1999)
assesses stable tendencies to engage in reflective versus ruminative thinking.
Reflective thinking is defined as intellective, involving curiosity or searching for
understanding and is related to the personality trait of openness. Ruminative
thinking is characterized as more neurotic, motivated by perceived threats or
Anxiety, Stress, & Coping
potential losses. There are 12 items used to assess each style, and respondents
indicate their agreement with each on a 5-point scale (strongly disagree  to 
strongly agree). Scores are reported as means for the items assessing each style. The
two scales have good internal reliability (Trapnell & Campbell, 1999; a?.90 for
rumination and .91 for reflection). In the current samples the internal reliabilities
were also adequate (reflection a?.78, rumination a?.88).
Need for Cognition (NCog)
The Need for Cognition ? Short Form (NCog-SF; Cacioppo, Petty, & Kao, 1984)
measures individuals’ stable tendencies to engage in and to enjoy thinking. Those
high on NCog tend to process information and ponder it even when the information
is not highly relevant, but they also are active processors of information with high
personal relevance. High NCog people have been shown to engage in more effortful
processing when trying to make sense of the world (Cacioppo, Petty, Feinstein, &
Jarvis, 1996). The NCog has 18 items, and responses are made on a 5-point scale
(extremely uncharacteristic  to  extremely characteristic). Scores are reported as
means. In a review of research using the NCog and NCog-SF, the internal reliability
was consistently above .85, and the validity of the measure as an indicator of
cognitive effort was verified (Cacioppo et al., 1996). In the current samples the
internal reliability was .86.
Private self-consciousness (PSCS)
The Self-Consciousness Scale (Fenigstein, Scheier, & Buss, 1975) assesses three
dimensions: private self-consciousness (PSCS), public self-consciousness, and social
anxiety. Only the 10 PSCS items were used in this study. PSCS captures stable
tendencies to engage in thinking about inner feelings, and self-analysis. Responses
are made on a 5-point scale (extremely uncharacteristic  to  extremely
characteristic). Test?retest reliability for the PSCS is adequate (.79) over a two-
week period (Fenigstein et al., 1975). In the current samples the internal reliability
was adequate (a?.68).
The COPE Inventory (Carver, Scheier, & Weintraub, 1989) provides a theoretically
based measure of differences in responses to stressors. The 60-item COPE provides
scores for 15 variations in coping strategies. Responses are made on a 4-point scale (I
usually don’t do this  to  I usually do this a lot). The instructions indicated that
participants should answer based on ‘‘what you usually do when you experience a
stressful event.’’ The internal reliabilities for the 15 scores were generally sufficient
given that each is based on only four items. Test?retest reliabilities were reasonable
over 6 and 8 week intervals (Carver et al., 1989). In the current samples, the internal
consistencies ranged from .52 to .95, with mental disengagement (.52) and
suppressing competing thoughts (.54) the only factors with a under .65.
A. Cann et al.
All procedureswere approved by the university institutional review board.Students in
eligible undergraduate psychology courses have access to an online research
participation website. When they log into the web page, they can see a list of studies
for which they are eligible. When students selected the current study, they were
directed to an online survey. The Informed Consent statement appeared on the first
page and, after agreeing to continue, participants began completing the measures.
Measures that dealt with stable differences (COPE, PSCS, NCog, and RRQ) were
presented first, invaried orders, before participants were asked to indicate the specific
event they had experienced. Demographic information came next, and as part of
those questions participants indicated and described the stressful event they had
experienced. The remaining event-specific measures (CBI, PTGI, ERRI, and IES-R)
were presented in random order for each participant. All responses were anonymous.
Confirmatory factor analysis (CFA)
To further verify the two-factor structure of the ERRI, a CFAwas conducted on the
new sample, combining the two separate samples (N?400). Once again, the
anticipated two-factor model was compared to the alternative single-factor model.
The single-factor model indicated a potential fit, x2(170) ?1144.83, pB.01, with the
fit indices above .90 (NFI?.94; CFI?.94), but the RMSEA (.17, 90% CI [.16?.18])
was relatively high. The two-factor model produced a far superior fit, x2
(169) ?362.07, pB.01, with allfit
(NFI?.98; CFI?.99; RMSEA?.055, 90% CI [.047?.062]). Thus, these data
further support the presence of two factors; intrusive thoughts and deliberate
rumination. As in the previous analyses, the two factors, while representing distinct
differences, also are correlated (r?.67).
statistics meeting recommendedlevels
Relationships with stable individual difference measures
Table 2 contains the correlations of the intrusive and deliberate scores with the
personality and individual difference measures. The two ERRI styles appear to be
unrelated to NCog, and moderately correlated with PSCS and the RRQ dimensions.
To see more clearly how the ERRI styles relate to these stable individual differences,
while controlling for the shared variance between the two ERRI types of thought, the
intrusive and deliberate scores were used in multiple regression analyses to predict
PSCS and the two RRQ styles. For both samples, the PSCS regression models were
significant, F(2, 197)?6.22 for immediate and F(2, 199) ?5.34 for recent, pB.01,
although only a small amount of variance in PSCS was accounted for by the ERRI
styles (adjR2?.05 for immediate and .04 for recent). In each model, only the
deliberate style was individually significant (deliberate style b?.189 and .226 for
immediate and recent, respectively, pB.05). Thus, it appears that people who have a
stable tendency to engage in self-analysis were more likely to engage in deliberate
rumination about a stressful experience, but the shared variance is still quite low.
For the reflection dimension of the RRQ a somewhat similar pattern was found.
For both the immediate and the recent time frames the models were significant,
Anxiety, Stress, & Coping
F(2, 197)?5.18 for immediate and F(2, 199)?7.05 for recent, pB.01, but explained
only a small amount of variance (adjR2?.04 for immediate and .06 for recent).
However, in the case of reflection, neither of the ERRI types was individually
significant in the sample that reported rumination soon after the event (b?.101 for
deliberate and .144 for intrusive, p?.13), but the deliberate style was individually
significant for the sample reporting recent ruminations about the stressful event
(b? ?.028 for intrusive, p?.74, and b?.273 for deliberate, pB.001). A stable
tendency to engage in intellective cognitive processing was associated with a higher
level of overall rumination immediately after experiencing a significant stressor, but
only with more deliberate rumination recently.
Both ERRI styles of repeated thought were associated with the RRQ rumination
dimension. In both samples the models were significant, F(2, 197)?17.32 for
immediate and F(2, 199)?12.67 for recent, pB.001, and a greater percent of
variance was explained than in the previous analyses (adjR2?.14 for immediate and
.10 for recent). In this case, as expected, the ERRI intrusive rumination style shared
more variance with the more neurotic RRQ rumination style. In each model, only the
ERRI intrusive style was individually significant (b?.331 for the immediate sample,
pB.001, and b?.367 for the recent sample, pB.001). A stable tendency to engage in
rumination reflection questionnaire factors and COPE dimensions.
Correlations of ERRI styles with private self-consciousness, need for cognition,
Time frame for ERRI responses
Immediately after the eventRecent
Need for cognition
Instrumental social support
Emotional social support
Note: The immediately after time frame sample size was 198, and the recent time frame sample size
A. Cann et al.
thinking focused on perceived threats and potential losses was associated with higher
levels of intrusive rumination recalled as having occurred in the immediate aftermath
of a stressful experience and also reported in the recent past. Again, while this was
the expected relationship, there still was only a very limited amount of shared
variance between the ERRI, designed to capture event-specific rumination styles,
and measures designed to assess stable individual differences.
The results for the relationships with differences in typical coping styles (see Table
2) indicate that in the time immediately after the stressor was experienced, those who
typically focus on their emotions, the venting emotions style, and those who use
mental disengagement, trying to distract oneself from thoughts about the stressor,
tend to experience more intrusive and more deliberate rumination. Again, because of
the shared variance between the two rumination styles, multiple regression analyses
were conducted for these two coping styles, using intrusive and deliberate styles as the
predictors. Both of the models were significant, F’s(2, 197)?7.34 for venting and
7.24 for mental disengagement, p’sB.01. In each case, the variance explained was
small (adjR2?.07 for venting, adjR2?.06 for mental disengagement), and the only
individually significant predictor was the intrusive style (for venting, intrusive
b?.251, pB.01, deliberate b?.020, p?.83; for mental disengagement, intrusive
b?.201, pB.05, deliberate b?.081, p?.38). Thus, it appears that for both of these
coping styles, the primary relationship was with higher levels of intrusive rumination.
Coping by focusing on emotions or by seeking distraction was associated with more
intrusive thoughts about the event, but not with deliberate thinking. In each case,
however, the variance explained was relatively small.
Intrusive rumination was also associated with the behavioral disengagement
coping style, which involves reducing efforts to deal with the stressor. This
relationship was present for both time frames and suggests that when one prefers
to cope by avoiding dealing with the stressor, intrusive thoughts about the event are
more common and salient. This same explanation could apply to the significant
relationships found between intrusive thoughts and trying to cope through substance
abuse. The coping style of substance abuse, like that of behavioral disengagement,
involves trying to avoid thinking about the event, so any thoughts about the event
would be intrusive, not deliberate.
The other results for the recent time frame are understandable. People who try to
cope by denial refuse to engage in any constructive cognitive work about the stressor,
so they are likely to experience continued intrusive thoughts if they have not dealt
with their stressful experience. The results for deliberate rumination recently are
interesting. The only significant relationships are the positive relationships with the
two coping styles that involve seeking social support. A deliberate style of thinking,
involving more problem-solving and meaning-seeking, is associated with seeking out
others, perhaps to help in the process of making sense of the experience.
Relationships with distress and posttraumatic growth (PTG)
The ERRI dimensions should explain more variance in the outcomes associated with
a stressful experience than the stable differences in styles of thinking. To evaluate
these possibilities, the ERRI dimensions and the RRQ dimensions were used to
predict distress levels as measured by the dimensions of the IES-R. The ERRI
intrusive dimension should be related to distress reported. The three dimensions of
Anxiety, Stress, & Coping
the IES-R, as well as the IES-R total score, were used as separate indicators of
distress, to demonstrate that the potential to predict the IES-R total is not being
biased by the inclusion of the intrusions dimension within the IES-R total score. The
CBI also was included in the model, since it provides some assessment of the
cognitive impact of the stressful event ? the degree of challenge to core beliefs and
the potential cognitive work that may be needed to rebuild them.
The same model was also used to predict PTG. In this case, the ERRI deliberate
dimension should be the strongest predictor. The means for the variables used in the
regression analyses are presented in Table 3. The correlations among the variables,
for both time frames in which the ERRI was completed, are provided in Table 4. The
results of the regression analyses are presented in Table 5.
The results indicate quite clearly that the amount of event-related intrusive
rumination, both in the immediate aftermath of a stressful experience and during the
most recent few weeks, was related to the level of distress reported. For all three of
the symptoms of distress captured by the IES-R and, of course, for the total score on
the IES-R, the overall models were highly significant. However, within each of the
models, only the ERRI intrusive score was individually significant as a predictor.
The stable difference in rumination about threats and potential losses, as measured
by the RRQ, was not individually significant within the models. Using the semi-
partial correlations for ERRI intrusive from the models, an estimate of the percent
of variance explained by the ERRI intrusive score can be obtained. The lowest semi-
partial correlation (sr?.31) was for the prediction of IES-R avoidance with the
immediate time frame, and the highest (sr?.45) was for the IES-R total in the recent
time frame. On average, across the four models tested for the immediate time frame,
the ERRI intrusive score explained 15.5% of the variance in the IES scores. For the
recent time frame the average variance explained by ERRI intrusive score was
Means and standard deviations for variables used in regression analyses for both
Immediately after the event time
frame for ERRIRecent time frame for ERRI
Note: N=198 for the immediately after time frame; and N=202 for the recent time frame. ERRI scores are
on a 0 (not at all) to 4 (often) scale. RRQ scores are on a 1(stronglydisagree)to 5 (stronglyagree)scale. CBI
is on a 0 (not at all) to 5 (verygreat degree) scale. IES scores are on a 0 (not at all) to 5 (often) scale. PTGI is
on a 0 (did not experience this change) to 5 (experienced this change to a very great degree) scale.
A. Cann et al.
17.2%. Thus, as predicted, the level of event-related intrusive rumination was a Download full-text
useful predictor of currently reported distress.
The two models predicting PTG (Table 4) support the predictions for the ERRI
deliberate score. Both models were significant and, as expected, initial disruption of
core beliefs, which sets the stage for possible PTG, was individually a significant
predictor. In addition to the CBI scores, the ERRI deliberate scores were individually
significant in both models. A finding that was not predicted, but which makes
intuitive sense, was the negative relationships found in both models between the RRQ
rumination measure and PTG. Apparently, a stable tendency to focus on threats and
potential losses may be somehow inhibiting the experience of growth. The RRQ
reflection dimension was not a significant predictor in either model. So, event-specific
deliberate rumination was positively associated with PTG, and explained more
variance in growth than did a stable tendency to engage in reflective rumination.
Predicting deliberate rumination
As indicated earlier, in the current model of PTG (Calhoun et al., 2010), it is assumed
that intrusive thoughts are one of the factors that may lead a person to engage in
growth for both time frames.
Correlations of ERRI and RRQ dimensions with outcome variables of distress and
Immediately after the event time frame for ERRI
1 ERRI intrusive
2 ERRI deliberate
3 RRQ rumination
4 RRQ reflection
6 IES-R avoidance
8 IES-R intrusions
9 IES-R total
Recent time frame for ERRI
1 ERRI intrusive
2 ERRI deliberate
3 RRQ rumination
4 RRQ reflection
6 IES-R avoidance
8 IES-R intrusions
9 IES-R total
Note: N=198 for the immediately after time frame; and N=202 for the recent time frame.
Anxiety, Stress, & Coping