Person. in&id. D@ Vol. 19, No. 6, pp. 827439.

1995
Copyright 01995 Elsevier Science Ltd
Pergamon 0191~t&369(95)00104-2 Printed in Great Britain. All rights reserved
0191.8869/95 $9.50 + 0.00

OPTIMISM AND PESSIMISM: ASSOCIATIONS TO COPING

AND ANGER-REACTIVITY

Kenneth E. Hart* and James B. Hittnert

Center for Alcohol and Addiction Studies, Brown University, Box G-BH,
Providence, RI 02912, U.S.A.

(Received 14 March 1995)

Summary-The purpose of the present study was to examine episodic coping and situational
anger-reactivity in relation to optimism and pessimism while correcting for some of the limitations of
previous research. In the present investigation, optimism and pessimism were operationalized as separate
constructs. In addition, an attempt was made to standardize the nature of the stressor that subjects reported
on, and to assess short-term emotional reactivity. Semipartial correlational analyses that corrected for
multicollinearity among the coping subscales indicated that, independent of variability in the contextual
features of the stressors, scores on the optimism subscale of the Life Orientation Test (LOT) were positively
correlated with confrontive coping and accept-responsibility coping, and inversely related to
escape-avoidance coping. Pessimism scores were positively related to escape-avoidance coping and
inversely related to confrontive coping and planful problem solving coping. Neither optimism nor pessimism
were related to levels of stress-induced anger reactivity. These data suggest that optimism and pessimism
are distinct constructs, that each construct is associated with different coping strategies, and that coping
differences between optimists and pessimists are not necessarily associated with differences in emotional
reactivity. These findings have implications for future research examining dispositional optimism and
pessimism as stress-moderating personality factors.

INTRODUCTION

In an attempt to clarify more precisely why optimists appear to be healthier than pessimists,
biobehavioral researchers have begun to test the notion that optimists and pessimists differ in terms
of the coping strategies they use when faced with stressful life events (Scheier, Weintraub & Carver,
1986). Preliminary evidence supports the notion that these two groups of people cope with stress
differently and that this difference might explain, in part, why optimism is linked to positive
well-being. Specifically, previous research has shown that compared to pessimists, optimists cope with
stress by using greater amounts of potentially adaptive strategies such as problem-focused coping
(Aspinwall & Taylor, 1992; Friedman, Nelson, Baer, Lane, Smith & Dworkin, 1992; Smith, Pope,
Rhodewalt & Poulton, 1989) and seeking social support (Scheier et uf., 1986). Relative to pessimists,
optimists also appear to rely significantly less on potentially maladaptive coping strategies such as
escape/avoidance (Aspinwall & Taylor, 1992; Friedman et al., 1992; Scheier et al., 1986, Study 2;
Smith et al., 1989; Strutton & Lumpkin, 1992; Zeidner & Ben-Zur, 1993; Zeidner & Hammer, 1992)
and venting emotion (Scheier et al., 1986; Zeidner & Ben-Zur, 1993; Zeidner & Hammer, 1992).
Although these preliminary findings are provocative, they must be interpreted with caution because
they are derived from investigations that suffer from a number of methodological limitations. One
problem concerns the fact that most previous studies failed to assess short-term adaptational outcomes
and simply assumed that optimists enjoy more benign stress-related outcomes than pessimists. For
example, most previous research has implied that optimism is associated with dampened emotional
reactivity in the face of stressful events, and that coping accounts for this (presumed) hyporesponsivity.
However, in the two studies that have actually examined stress-related outcomes, little or no evidence
was found to support the idea that optimists show dampened emotional reactivity in the face of stress
(Zeidner, 1993; Zeidner & Ben-Zur, 1993).
A second methodological limitation which requires that caution be exercised in interpreting the
results of previous coping research involving the Life Orientation Test (LOT) concerns recent findings
suggesting that the LOT assesses two distinct constructs rather than a single bipolar dimension as
previously thought (Dember, Martin, Hummer, Howe & Melton, 1989; Marshall, Wortman, Kusulas,
*To whom all correspondence should be addressed.
tCurrent address: Department of Psychology, College of Charleston, Charleston, SC 29424, USA.

827
828 Kenneth E. Hart and James B. Hittner

Hervig & Vicekers, 1992; Mook, Kleijn & Van der Ploeg, 1992; Mroczek, Spiro, Aldwin, Ozer &
Bosse, 1993; Plomin, Scheier, Bergeman, Pedersen, Nesselroade & McClearen, 1992). To date,
episodic coping research that has utilized the LOT has failed to address the possibility that optimism
and pessimism may indeed be orthogonal constructs, and that coping and adaptational outcomes may
have differential associations to each of the two constructs.
Interpretation of the results of previous research has also been hampered by the use of inappropriate
statistical analyses. It is likely that, in the vast majority of these studies, the coping predictor variables
were more strongly correlated with each other than with the LOT criterion variable. This being the
case, use of zero-order correlations would yield highly misleading findings due to multicollinearity.
Nevertheless, most previous studies base their conclusions and inferences on simple correlations.
A fourth problem with earlier research is that it has neglected to examine the possibility that there
may be systematic differences between optimists and pessimists in terms of the nature of the stressors
they report on. As noted by Folkman (1992) and McCrae (1992) this neglect of contextual
embeddedness is a general problem in most coping research, and is therefore not unique to studies
involving optimism. Nevertheless, in many of the coping studies involving the LOT Ss were given
free choice in the selection of the type of stressful event they could report on. As a result, some Ss
may have reported on financial stressors, whereas others may have reported on interpersonal stressors.
Still others may have reported on illness-related, or work-related stressors. In the light of theory and
research (Fleishman, 1984; Lazarus & Folkman, 1984; McCrae, 1984, 1992) suggesting that different
types of stressors elicit different coping patterns, and the possibility that individuals scoring high on
dispositional optimism may have elected to report on stressors that were systematically different than
the stressors faced by their lower scoring counterparts, previously reported coping differences between
optimists and pessimists might have stemmed from differences in characteristics of the stressors-
rather than characteristics of the disposition per se.
A closely connected fifth problem is that some of the coping items that appear on episodic coping
inventories are simply not applicable in the context of certain types of stressful situations (Ben-Porath,
Wailer & Butcher, 1991; Folkman, 1992; Stone & Kennedy-Moore, 1992). Thus, if the coping
strategies of optimists and pessimists are contextually embedded in different physical and social
environments, it is possible that the corresponding scale scores may be artifactually distorted,
depending on the degree to which items are or are not relevant to the stressful situation being reported
upon. This, in turn, could lead researchers to misinterpret the meaning of results showing different
coping patterns for optimists vs pessimists.
As noted by Stone and Kennedy-Moore (1992), there are a variety of possible solutions to the
problem of variability in contextual relevance. In the case of the present study, this problem was
addressed by studying coping and its correlates within a single type of stressor category. Specifically,
an attempt was made to restrict the heterogeneity of the stressors that individuals reported on by
instructing all Ss to describe a work-related stressor that provoked feelings of anger. While this
method still allows for a certain amount of variability in terms of stressor severity or type, it ensures
relatively more standardization than the typical method of unconstrained free recall of unknown
stressors.
The decision to study negative emotional arousal in the context of occupational settings was based
on research showing that work-related stress is a highly prevalent and very costly phenomenon
(Cooper & Payne, 1988). Moreover, anger-provoking situations were targeted because previous
research shows that anger-arousal increases the probability of aggression and violent crime (Maiuro,
Cahn, Vitaliano, Wagner & Zegree, 1988; Rule & Nesdale, 1976), heightens the risk of premature
death and physical illness (Appel, Holroyd, & Gorkin, 1983; Chesney & Rosenman, 1985; Friedman,
1992; Siegman & Smith, 1994), and is involved in other dysfunctional syndromes (Novaco, 1986).
Moreover, research shows that anger-coping strategies may influence the risk of hypertension and
mortality (Appel et al., 1983; Chesney & Rosenman, 1985; Friedman, 1992; Holroyd & Gorkin, 1983;
Johnson, 1990; Julius, Harburg, Cottington & Johnson, 1986; Siegman & Smith, 1994). In this
connection, both theory (Roskies, 1980) and research (Suarez & Williams, 1989) have suggested that
variability in anger-reactivity is associated with different degrees of hemodynamic and neuroen-
docrine responsivity. Thus, it is possible that optimism may be inversely related to illness because
optimists show dampened emotional arousal and concomitant low levels of physiological and
neuroendocrine reactivity when provoked by stressful circumstances.
 Coping 829

In the light of the above considerations, the purpose of the present study was to examine the
relationship between coping and optimism/pessimism after correcting for each of the five major
problems identified in previous research. In particular, an attempt was made to circumscribe the milieu
in which coping took place and to statistically disentangle the collinearity among the coping subscales
via semipartial correlations. Use of this statistical approach allows for a determination of the unique
contribution of each coping predictor. Moreover, a direct measure of situational anger-reactivity was
obtained. Finally, the present study also treated optimism and pessimism as separate personality traits.

METHOD

Subjects
Ss consisted of 43 adults employed at local banks on Long Island, New York. The 13 males and
30 females completed a research questionnaire packet as part of a work-related seminar offered in
the School of Business at a midsized, private University. Their mean age was 36.8 yr (SD = 4.6).

Overview r$ procedure
In a mass testing setting, Ss completed a number of inventories. First, they completed the Life
Orient&m Test (LOT). Next, they responded to the Stressful Scene Reconstruction Inventory (Hart
& Cardozo, 1986; Hart, Turner & Cardozo, 1987), a newly developed measure using both open and
closed-ended formats to obtain data about the ecological features of a recently encountered work
situation that provoked feelings of anger. With special reference to the anger-provoking work stressor,
Ss then completed the 1986 version of the Ways of Coping Inventor?; (Folkman, Lazarus, Gruen &
DeLongis, 1986), and a IO-item Anger-Reactivity scale to indicate how angry they felt while dealing
with the stressful situation.

Measures of optimism and pessimism

Optimism and Pessimism were assessed via the Life Orientation Test [LOT (Scheier & Carver,
1985)], a 12-item self-report measure originally designed to assess dispositional optimism or
generalized expectancies that good things will happen. Since four items are fillers, the LOT total score
is based on only eight of the 12 items. Example LOT items include Im a believer in the idea that
every cloud has a silver lining and I hardly ever expect things to go my way. In the present study,
items on the LOT were responded to on the basis of O-4 point Likert agreement scale with the anchor
phrases strongly agree and strongly disagree.
Preliminary convergent, discriminant and predictive validity evidence for the LOT have been
reported in a number of studies (Carver & Gaines, 1987; Scheier & Carver, 1985, 1987; Scheier et
al.. 1986; Strack, Carver & Blaney, 1987). Recent data, however, suggest that the LOT may actually
assess two distinct constructs rather than a single bipolar dimension (Dember et al., 1989; Marshall
et al.. 1992: Mook et al., 1992; Mroczek et al., 1993; Plomin et al., 1992).
Thus, for the purpose of the present study, two subscale scores were computed: one for optimism
and another for pessimism. The Optimism subscale contains four items (e.g. Im a believer in the
idea that every cloud has a silver lining ). The Pessimism subscale also contains four items (e.g.
I hardly ever expect things to go my way). Both scales were scored using the sum score approach,
such that higher values were indicative of greater construct endorsement.

Assessing the contextual features of the stressor

Folkman (1992) has stated that when coping research does not employ prospective designs, our
responsibility is . to make sure that retrospective techniques are designed carefully to allow as much
accurate recall as possible.. . (p. 220). In the present study, an attempt was made to enhance the
accuracy of retrospective reports by using visual imagery. Specifically, Ss completed the newly
developed Stressful Scene Reconstruction Znventory (Hart & Cardozo, 1986; Hart et al., 1987), a two
part inventory that first primes Ss for recall of angry affect and then uses visual imagery to facilitate
mental simulation of a previously encountered anger-provoking work-stressor. Finally, the inventory
830 Kenneth E. Hart and James B. Hittner

also solicits descriptive data pertaining to the ecological features of the stressor by posing three groups
of open- and closed-ended questions.
During Part 1 of the SSRI, instructions encourage subjects to re-create, through recall prompts
and visual imagery, a phenomenologically significant anger-provoking work-related situation that
they had encountered in the last 2-3 months. In other words, Ss were encouraged to report on a critical
incident involving anger-arousal. To prime Ss memories and facilitate easy recall of the situation,
Ss first completed the Novaco Provocation Inventory [NPI (Novaco, 1975)]. The NPI provides Ss with
a list of 80 situations that might provoke feelings of irritation, anger or rage. Theory and research on
mood-congruent memory (Blaney, 1986) suggests that use of such a priming strategy should enhance
recall and increase the accuracy of subsequent reports of real-life anger-provocations.
Toward the end of Part 1 of the SSRI, Ss were given detailed mental imagery instructions designed
to enhance their ability to visualize the anger-provoking situation. In particular, Ss were asked to close
their eyes and remain quiet for 2-3 min and to use this time to project a mini-mental movie of the
situation onto the screen of their mind. As noted by Taylor and Schneider (1989), this type of mental
simulation represents a useful tool for stress researchers because it most likely yields retrospective
reports that are more accurate than traditional methods.
In Part 2 of the SSRI, three groups of open-and closed-ended questions were posed in order to obtain
descriptive data pertaining to the ecological features of the stressful encounter. For the purposes of
the present study, closed-ended questions pertaining to the situations temporal, social, and
psychological context were analyzed.
Assessing the temporal context. In an attempt to better understand the temporal context of the
situation, Ss were asked to respond with a yes or no answer to the following five statements: (1) the
situation is still going on right now; (2) the situation has happened already and is resolved; (3) the
situation has happened already but has not yet been resolved; (4) the situation has happened and
will likely be resolved soon; and (5) the situation has happened and will not likely be resolved
anytime soon.
Assessing the social context. The second group of questions posed in Part 2 of the SSRI asked for
a yes or no response to nine statements regarding the situations social context. These statements were:
(1) At least one other person was directly involved in the situation; (2) One other person was
directly involved in the situation; (3) Two other people were directly involved in the situation; (4)
Three other people were directly involved in the situation; (5) Four other people were directly
involved in the situation; (6) The primary other person was my boss/supervisor; (7) The primary
other person was a co-worker; (8) The primary other person was a subordinate; and (9) The
primary other person was a customer.
Assessing the psychological context. The third group of questions posed in Part 2 of the SSRI probed
for information regarding the psychological context of the stressful situation, and were taken from
previous research by Folkman et al. (1986). Ss were asked to give yes/no answers to the following
four statements: (1) I could change the situation or do something about it; (2) I had to accept or
get used to the situation; (3)I needed to know more about the situation before acting; and (4) I
had to hold myself back.

Measurement of coping
Coping was assessed via the Ways Of Coping Inventory (WCI) which is an episodic measure of
coping designed to elicit information about a number of cognitive and behavioural coping efforts that
people may use to deal with specific situations that tax or exceed response capabilities (Cohen, 1987;
Lazarus & Folkman, 1984). In the present study, Ss completed the 1986 version of the WC1 (Folkman
et al., 1986) with special reference to how they coped with the anger-provoking work situation
described in the Stressful Scene Reconstruction Inventory.
Names of the eight subscales on the 1986 WCI, together with sample items and Cronbach US(as
reported in Folkman et al., 1986), are as follows: (1) confrontive coping (I stood my ground and fought
for what I wanted, I expressed anger to the person who caused the problem; c1= 0.70); (2)
distancing (I made light of the situation, or refused to get too serious about it, I tried to forget the
whole thing; 0:= 0.61); (3) se lf- control (I tried to keep my feelings from interfering with other things
too much, I tried to keep my feelings to myself; c1= 0.70); (4) seeking social support (I talked
 Coping 831

Table I. Descriptive data for coping, personality and anger variables

Number of Cronbach
Mean SD items ci

Confrontive
coping 5.7 3.6 6 0.64
Distancing
coping 6.0 4.0 6 0.68
Self-control
coping 9.1 4.5 0.74
Seeking support
coping 6.7 4.3 6 0.80
Acceot
resbonsibility
coping 3.4 3.3 0.79
Escape-avoidance
coping 4.2 4.4 0.78
Planful problem
solving coping 9.0 4.6 0.81
Positive reappraisal
coping 6.7 5.0 0.80
Optimism 10.2 2.4 4 0.67
Pessimism 5.6 2.5 4 0.64
Anger-reactivity 21.4 5.2 10 0.80

to someone to find out more about the situation, I talked to someone about how I was feeling;
a = 0.70); (5) uccept responsibility (I criticized or lectured myself, I realized I brought the problem
on myself; a = 0.66); (6) escape-avoidance (I wished that the situation would go away or somehow
be over with, I tried to make myself feel better by eating, drinking, smoking or using drugs or
medication; a = 0.72); (7)plunfulproblem solving (I made a plan of action and followed it, I just
concentrated on what I had to do next; a = 0.68); and (8) positive reappraisal (I changed or grew
as a person in a good way, I rediscovered what is important in life; a = 0.79).
A summed frequency (i.e. raw frequency) score was computed for each subscale by summing
the Likert ratings for all items loading on that subscale.

Measurement of anger-reactivity
Short-term negative emotional reactions resulting from exposure to the stressor described in the
SSRI were assessed via the Anger-Reactivity Scale. In the present study, this scale consisted of a
lo-item adaptation of the state version of the State-Trait Anger Inventory [STAI (Spielberger,
Jacobs, Russell & Crane, 1983)], and was used as a situation-specific measure of the extent to which
Ss felt angry emotions when dealing with the stressful provocation. In the present study, the State
portion of the STAI was adapted by having Ss respond to the items in the context of how they felt
while coping with the stressful situation at work. The scale was introduced with the following written
instruction: For each of the statements listed below, indicate how angry you felt while you were
dealing with the anger-provoking situation that you described earlier. Each of the 10 items on the
scale was responded to on the basis of a four point Likert scale that ranged from (1) not at all to
(4) very much so.

RESULTS

Descriptive data for the personality, coping and affective measures appear in Table 1. Table 1 shows
that the four-item optimism scale had an a of 0.67. The four-item pessimism scale also had an
acceptable, but somewhat lower level of internal consistency reliability (a = 0.64). These findings are
somewhat lower than those reported by Mroczek et al. (1993) who found that the four-item optimism
and pessimism subscales had as of 0.76 and 0.79, respectively. It is likely that the higher as reported
by Mroczek et al. (1993) can be attributed to their relatively large sample size (N = 1192). Thus, the
comparatively lower as in the present study are not surprising considering that there are only four items
on each of the subscales and that the sample consisted of 43 Ss.
Although not reported in Table 1, the eight-item (combined) optimism-pessimism scale had an a
of only 0.49. This finding, which indicates that all eight items together lack homogeneity, is consistent
832 Kenneth E. Hart and James B. Hittner

Table 2. Ecological data describing the contextual features of the anger-provoking stressful situation

Agreement
w)*

Temporal Context
I. Situation is still going on right now 14.0
2. Situation has happened already and has been resolved 23.3
3. Situation has happened already but has not yet been resolved 16.3
4. Situation has happened already and will likely be resolved soon 2.3
5. Situation has happened and will not likely be resolved any time soon 27.9

Social Context
6. At least one other person was involved 90.7
7. One other person was directly involved 53.5
8. Two other people were directly involved 16.3
9. Three other people were directly involved 2.3
10. Four other people were directly involved 7.0
II. Primary other person was boss/superior 14.0
12. Primary other person was co-worker 23.3
13. Primary other person was subordinate 20.9
14. Primary other person was customer 7.0

1.5. Could change or do something about the situation 53.5

16. Had to accept or get used to situation 41.9
17. Needed to know more about the situation before acting 23.3
18. Had to hold self back 55.8

*Some percentages do not total to IO0 because of missing data.

with (eight-item) CISof 0.58 and 0.60 reported by Zeidner (1993) and Zeidner and Hammer (1992)
respectively. As can be seen in Table 1, as varied across the various coping scales from a low of 0.64
(confrontive coping) to a high of 0.8 1 (planful-problem solving). Considering that coping is a process
variable and that each subscale contains only a few items, CLS in this range are acceptable (Folkman,
1992).
Table 1 also shows that Ss experienced substantially elevated levels of anger in the stressful
situation reported upon. This conclusion is based on a comparison of the Anger-Reactivity score of
21.4 with normative data showing a mean of approx. 14.5 for Ss responding to the original State version
of the STAI (Spielberger et al., 1983).

Contextual features of the stressor

Table 2 shows ecological data describing the contextual features of the stressful situation. Briefly,
results pertaining to the temporal context of the situation indicate that only a minority of Ss (14%)
reported on a stressful situation that was currently ongoing. Thus, the vast majority of Ss reported on
stressful situations that had occurred in the past. Results of a oneway ANOVA that compared mean
optimism scores for the five temporal categories was statistically nonsignificant. Similarly, an
ANOVA comparing mean pessimism scores across the five temporal categories was also
nonsignificant.
Regarding the social context, most Ss reported on interpersonal anger provocations (90.7%). This
finding is consistent with results reported by Averill (1982), suggesting that anger is an interpersonal
emotion. Similar to the temporal context analyses, oneway ANOVAs indicated nonsignificant
differences in optimism and pessimism scores across the nine social context categories.
Table 2 also summarizes the situations psychological context. It is interesting that although the
majority of Ss (53.5%) felt that the anger-provoking situation was potentially controllable (i.e. item
15), most believed they needed to hold themselves back (55.8%) and wait for more information
before acting (23.3%). Thus, the majority of Ss perceived a need to be cautious and not act too hastily.
Finally, a sizable percentage of the Ss indicated a sense of resigned acceptance (41.9%). Oneway
analyses of variance failed to reveal significant differences in optimism and pessimism scores across
the psychological context categories.

Zero-order correlations linking optimism and pessimism to coping and anger

Table 3 presents zero-order Pearson correlations relating optimism and pessimism scores to the
various coping subscales and anger-reactivity. Optimism was positively correlated with confrontive
 Coping 833

Table 3. Pearson correlations of optimism and pessimism scores to

coping and anger-reactivity

Optimism Pessimism

Confrontive coping 0.44

Distancing coping
Self-control coping
0.17
0.22
-:FlZ
0.19
Seeking support coping 0.23 0.13
Accepting responsibility coping 0.29 0.39*
Escape-avoidance coping 0.08 0.41*
Planful problem solving coping 0.33 0.15
Positive reappraisal coping 0.27 0.19
Anger-reactivity 0.20 0.22

*P < 0.05
All probability levels are two-tailed.
The accept-responsibility and escape-avoidance subscales were
logarithmically transformed to attain normality.

coping (r= 0.44, P< 0.05). Other data, not shown in Table 3, revealed that greater use of con-
frontive coping was associated with HIGHER levels of anger-reactivity (r= 0.40, P< 0.01).
Pessimism scores were positively correlated with both accepting responsibility coping (r = 0.39,
P < 0.05) and escape-avoidance coping (r = 0.41, P < 0.05).

Interrelationships among the coping scales

Table 4 shows the intercorrelations among the coping subscales. Inspection of these data indicates
that 19 of the 28 correlations (68%) were statistically significant. Moreover, the correlations exceeded
0.50 in 10 of the 19 cases. Because the majority of the intercorrelations were so large, predictor
multicollinearity represented a statistical problem that required some resolution. In order to deal with
the multicollinearity issue, semipartial or part correlations were calculated to determine the unique
association between each coping scale and optimism and pessimism. The part correlation coefficient
squared (rr*) is a coefficient of determination and represents the amount of criterion variance uniquely
accounted for by each coping scale predictor (Darlington, 1990; Stevens, 1986). In the present study,
part correlations were generated by regressing optimism and pessimism on the eight coping subscales.
The regression statistics themselves, however, were not interpreted due to the small sample size
employed (N = 43) and the insufficient subject-to-predictor variable ratio (Darlington, 1990).

Semipartial correlations relating optimism and pessimism to coping

Table 5 shows the semipartial correlations (rp) linking each of the eight coping scales to optimism
and pessimism. As previously noted, each of these coefficients represents the unique association
between a particular coping strategy (after removing common variance associated with other coping
scales) and the personality traits of optimism and pessimism. Table 5 also presents the corresponding
coefficients of determination (i.e. rp2values) which index the strength or relative importance of each
relationship. Following the recommendation of Darlington (1990), r,* values will be used in
interpreting the significance or importance of the relationships. Thus, unlike previous research,
measures of statistical significance (i.e. P values) will not be examined.

Table 4. Intercorrelations among the coping subscales

2 3 4 S 6 7 8

I Confrontive 0.32 o&t** 0.31 0.33 0.47** 0.50** 0.47**

2. Distancing 0.56* 0.15 0.17 0.43*** 0.29 0.47**
3. Self-control 0.55* 0.60* 0.67* 0.61* 0.58*
4. Seeking support 0.37*** 0.30 0.4.5** 0.45**
5. Accepting responsibility 0.74* 0.30 0.52**
6. Escape-avoidance 0.24 0.601
7. Planful-problem solving 0.53*
8. Positive reappraisal

P<0.001; **p<o.o1: ***P<o.os.

All P values are two-tailed.
The accept-responsibility and escape-avoidance coping subscales were logarithmically transformed to attain normality.
834 Kenneth E. Hart and James B. Hittner

Table 5. Semipartial correlations and coefficients of determination linking

optimism and pessimism to coping

Optimism Pessimism

P r,* P rppz

Confrontive 0.31 0.096 - 0.23 0.053

Distancing 0.19 0.036 -0.18 0.032
Self-control - 0.07 0.005 -0.13 0.017
Seekingsupport 0.04 0.002 - 0.03 0.001
Accept-responsibility 0.39 0.152 0.10 0.010
Escape.-avoidance - 0.33 0.109 0.28 0.078
Planful problem solving 0.01 0.000 - 0.23 0.053
Positive reappraisal 0.04 0.002 0.01 0.000

rp = semipartial or part correlation; rp* = coefficient of determination.

Only rpz values 2 0.050 (5% unique explained variance) were interpreted as
having research significance.

There are a number of advantages to relying on the coefficient of determination in interpreting the
value and worth of findings. First, use of this effect size index focuses attention on variance accounted
for and thus helps keep researchers honest by reducing the temptation to inappropriately inflate or
deflate the importance of findings. Unfortunately, these tendencies are sometimes all too common in
studies that rely on the interpretation of P values alone. In addition, because rp2 serves as an index
of relative strength, its use allows researchers to meaningfully compare and contrast the effect sizes
of various predictors. For example, if rp is 0.10 for Predictor 1 and 0.050 for Predictor 2, one can
conclude that the magnitude of the relationship to the criterion is twice as strong for Predictor 1 as
compared to Predictor 2. In the former case, 10% of the variance is accounted for, while in the latter
case only 5% of the variance is accounted for.
In using the coefficient of determination as a guide to establishing research significance, a decision
was made to only interpret as meaningful r,* values that were greater than 0.050. Thus, at least 5%
of the criterion variance had to be uniquely accounted for in order for a coping scale to be interpreted
as having research significance.
Prediction ofoptimism. Inspection of Table 5 reveals that three of the coping scales predicted unique
variance in the criterion of optimism. The strongest effect size was for accept-responsibility coping,
which was positively correlated with optimism (rp = 0.39, r, * = 0.152). Thus, accept-responsibility
coping accounted for 15.2% unique variance in optimism. The next strongest effect size was for
escape-avoidance coping (rp = - 0.33, rp2= 0.109) which accounted for 10.9% of the unique variance
in optimism. Finally, confrontive coping also emerged as a significant predictor (rp = 0.31,
r,* = 0.096) accounting for 9.6% unique variance in optimism.
Prediction of pessimism. Results shown in Table 5 also indicate that three of the coping scales
predicted unique variance in the criterion of pessimism. The strongest effect size was for
escape-avoidance coping, which was positively correlated with pessimism (r-r = 0.28, rp2= 0.078).
Thus, escape-avoidance coping accounted for 7.8% unique variance in pessimism. Confrontive coping
and planful problem solving coping demonstrated equivalent effect sizes. However, the direction of
their associations were opposite. Confrontive coping was inversely related to pessimism (rp = - 0.23)
and planful problem solving was positively related to pessimism (rp = 0.23). Each of these predictors
accounted for 5.3% unique variance in the outcome of pessimism.

DISCUSSION

The major objective of the present investigation was to examine the relationship between coping
and optimism/pessimism after correcting for five major problems identified in previous research. In
addition to employing a direct measure of short-term negative emotional reactivity, the present study
also treated optimism and pessimism as separate personality traits. Moreover, an attempt was made
to circumscribe the milieu in which coping took place, and to statistically disentangle the
intercorrelations among the coping predictors. Specifically, the study utilized semipartial correlations
so as to determine the unique contribution of each coping predictor.
Results of the semipartial correlational analyses indicated that optimism was significantly and
 Coping 835

positively related to confrontive coping. Despite the fact that the present study operationalized
optimism differently than previous research and that different statistics were used, the present finding
pertaining to confrontive coping is consistent with the results of at least six other studies linking higher
LOT scores to greater use of active coping or problem-focused coping (Aspinwall & Taylor, 1992;
Friedman er al., 1992; Scheier et al., 1986, Study 2; Smith et al., 1989; Strutton & Lumpkin, 1992;
Zeidner & Ben-Zur, 1993).
Visual inspection of the confrontive coping items indicates that greater optimism was associated
with a greater willingness to express angry feelings or discharge negative emotions to the source of
frustration. This observed tendency toward greater emotional expressiveness is inconsistent with the
results of at least three previous studies, all of which have reported a negative relationship between
optimism and the discharge of emotions (Scheier et al., 1986; Zeidner & Ben-Zur, 1993; Zeidner &
Hammer, 1992). In these prior studies, zero-order correlations that ranged from - 0. I7 to - 0.24 were
found for expressing feelings, ventilation, and discharging emotions. One possible resolution of
the apparent inconsistency between the present finding and previous results lies in the distinction
between the hostile expression of angry affect and the instrumental expression of angry affect.
Specifically, it is possible that, compared to nonoptimistic people, individuals scoring higher on
optimism are relatively more likely to express anger to others when it has instrumental value, but are
less inclined to express anger when it might potentially have unproductive or even destructive effects.
Of course, methodological differences might also explain the inconsistency between the present results
and previous findings. Clearly, this is an issue that deserves more research.
As can be seen in Table 5, pessimism scores were also related to the tendency to engage in
problem-oriented confrontive coping. However, the relationship was negative in direction. The inverse
association mirrors the finding for optimism and is also consistent with previously published findings
linking lower scores on the LOT to lesser use of active coping or problem-focused coping
(Aspinwall & Taylor, 1992; Friedman et al., 1992; Scheier et al.. 1986, Study 2; Smith et al., 1989;
Zeidner & Ben-Zur, 1993).
The tendency among previous studies has been to simply assume that a positive correlation between
LOT scores and active/problem-focused coping supports the notion that optimists engage in more
adaptive coping than pessimists. in the present study, the issue of coping efficacy was addressed
empirically by examining the semipartial correlation between confrontive coping and short-term
negative emotional reactivity. Results of this analysis showed a moderate positive association
(rP = 0.34, r,Z = 0.116). Thus, greater use of confrontive coping did not appear to result in positive
adaptational outcomes. On the contrary, greater use of confrontive problem-focused coping appeared
to be maladaptive, at least in terms of short-term emotional well-being. The issue of whether or not
confrontive coping (or any other coping scale for that matter) truly mediated the relationship between
personality and negative emotional reactivity could not be examined as neither optimism (r = 0.15)
nor pessimism (r = 0.07) were correlated with anger-reactivity.
The general idea that higher optimism (or lower pessimism) may not necessarily be associated with
greater coping effectiveness has also received empirical support from a study by Zeidner and Hammer
(1992) who examined perceived coping effectiveness in the context of SCUD missile attacks, and
Deshamais, Godin, Johin, Valois and Ross (1990) who studied perceived coping effectiveness in post
myocardial-infarction patients. In both studies, LOT scores were unrelated to perceived coping
effectiveness. Of course, it is possible that the coping strategies used by optimists in these studies may
have had delayed positive effects, or salubrious effects on adaptational outcomes that were not
assessed, such as physiological, social and behavioral adjustment.
As can be seen in Table 5, other results showed that optimism scores were positively related to
the tendency to cope by accepting responsibility for the situation. The tendency for greater optimism
to be associated with greater self-blame conflicts with previous research showing LOT scores to be
significantly and inversely related to self-blame (Smith et al., 1989). The source of this discrepancy
is presently unknown. However, one could speculate that methodological differences may have been
responsible.
In any case, the finding that greater optimism is associated with a more pronounced tendency to
accept-responsibility might be interpreted as indicating a corresponding tendency to make internal
attributions for negative life events. This line of thinking is consistent with previous research showing
that higher scores on the LOT are associated with a greater internal locus of control (Guarnara &
836 Kenneth E. Hart and James B. Hittner

Williams, 1987). In this connection, it is also possible that high optimism scorers may use higher levels
of self-blame coping in order to help bolster their positive outcome expectancies by creating what
Folkman (1984) has termed an illusion of control. Greater use of self-blame, however, does not
appear to confer an adaptational advantage with respect to short-term emotional reactivity. This
speculation is based on the results of a semipartial correlation showing that anger-responsivity was
unrelated to scores on the accept-responsibility scale (rp = 0.11, rp2= 0.012).
As shown in Table 5, results indicated that optimism was inversely related to escape-avoidance
coping. Similar inverse correlations have also been reported in a number of previous studies that
utilized the unidimensional LOT scale (Aspinwall & Taylor, 1992; Friedman et al., 1992; Scheier
etal., 1986, Study 2; Smith etal., 1989, Study 2; Zeidner & Ben-Zur, 1993; Zeidner 8z Hammer, 1992).
Thus, the inverse association between optimism and escape-avoidance coping appears to be a rather
robust phenomenon.
The adaptational significance of the inverse association between optimism and escape-avoidance
coping was examined by conducting a semipartial analysis in which residualized escape-avoidance
coping scores were correlated with variability in anger-reactivity. The near zero correlation (rp = 0.08)
indicated that, at least in the short-run, escape-avoidance coping was not associated with positive
emotional benefits. It is possible, however, that salubrious effects may have been delayed, or that they
may have occurred in the form of unmeasured outcomes.
Other results indicated a positive semipartial correlation between escape-avoidance coping and
scores on the pessimism subscale. This finding is consistent with the results of previous studies that
have utilized the unidimensional LOT (Aspinwall & Taylor, 1992; Friedman er al., 1992; Scheier et
al., 1986, Study 2; Smith et al., 1989, Study 2; Zeidner & Ben-Zur, 1993; Zeidner & Hammer, 1992).
However, there was no evidence to support the notion that greater use of escape-avoidance coping
by high pessimism scorers had maladaptive emotional consequences. This conclusion was based on
the results of a semipartial analysis linking residualized escape-avoidance coping scores to
anger-reactivity (rp = 0.08).
Further inspection of the results in Table 5 indicates that pessimism scores were also inversely
related to planful problem solving coping. These results are consistent with previously reported
findings showing lower scores on the LOT to be associated with lesser use of active coping or
problem-focused coping (Aspinwall & Taylor, 1992; Friedman et al., 1992; Scheier et al., 1986,
Study 2; Smith et al., 1989; Zeidner & Ben-Zur, 1993).
Neither optimism nor pessimism was related to variability in anger-reactivity. Thus, the negative
short-term emotional effects of exposure to anger provoking circumstances do not appear to be
moderated by optimism or pessimism. This lack of association to emotional arousal is consistent with
previous research using unidimensional LOT scores which has found that anxiety reactions to SCUD
missile attacks are unrelated to the bipolar optimism-pessimism scale (Zeidner, 1993). In a related
study (Zeidner & Ben-Zur, 1993) pessimism was found to be a relatively weak predictor of stress
reactions and outcomes (p. 218). The present results, however, conflict with those reported by
Thomas and Donnellan (1991) who found a significant inverse association between LOT scores and
a trait measure of somatic anger reactivity in the face of provocation. It is possible that methodological
differences account for the discrepancies among these studies. In any case, the findings of the present
investigation suggest that it is incorrect to simply assume that individuals scoring high on optimism
become less emotionally distressed than their lower scoring counterparts, and that coping accounts
for this presumed difference.
The findings of the present study are also consistent with previous research suggesting that
optimism and pessimism are orthogonal constructs rather than bipolar opposites of a single dimension
(Dember et al., 1989; Marshall et al., 1992; Mook et al., 1992; Mroczek et al., 1993; Plomin et al.,
1992). In fact, the zero-order correlation between optimism and pessimism in the present study was
F-= - 0.05. Furthermore, optimism and pessimism scores were both positively (albeit non-
significantly) related to anger reactivity. Theoretically, one would expect that if the two constructs
were indeed reverse images of each other, these two correlations would have been in opposite
directions. The fact that both correlations were in the same direction further supports the notion that
optimism and pessimism are distinct constructs. Other discriminant validity evidence for the two
personality scales comes from results showing that while optimism accounted for over 15% of the
variance in accept-responsibility coping, pessimism accounted for only 1%. Similarly, pessimism
 Coping 837

accounted for 5.3% of the variance in planful-problem solving whereas optimism accounted for no
variance at all. In summary, the overall pattern of findings reported here supports the notion that
optimism and pessimism are separate constructs and not bipolar opposites of the same construct.
Other results showed that neither optimism nor pessimism scores were associated with the
ecological features of the stressors that were reported on. This finding suggests that the observed
associations linking coping to optimism and pessimism were relatively independent of the
confounding influence of contextual variability in the nature of the stressors that were reported on.
Thus, it is unlikely that the observed relationships linking personality to coping can be explained
away as methodological artifacts arising from differential exposure to different types of stressors.
Rather, the coping differences reported in the present study are more likely attributable to dispositional
characteristics associated with the traits of optimism and pessimism. This conclusion should be viewed
as tentative, however, in light of the fact that low statistical power due to small sample size was a
problem in many of the ecological analyses.
To summarize, the present investigation examined the relationship between coping and
optimism/pessimism after correcting for limitations identified in previous research. In addition to
employing a direct measure of situational emotional arousal, optimism and pessimism were also
treated as separate personality traits. Moreover, an attempt was made to delimit the circumstances in
which coping took place, and to statistically disentangle the intercorrelations among the coping
predictors. Semipartial correlations, that assessed the unique contribution of each coping predictor,
showed that both optimism and pessimism are associated with a variety of different coping strategies.
Moreover, because all Ss reported on the same type of stressor, it is unlikely that these findings were
confounded by personality-related variability in the contextual features of the stressors reported on.
Finally, results showed that neither optimism nor pessimism were directly related to levels of
stress-induced anger-reactivity.
In conclusion, the present results should be interpreted in light of several methodological
limitations. The most obvious limitation stems from the fact that the cross-sectional design precludes
causal inferences. In addition, since information was collected about coping and adaptational
outcomes in a fairly small sample of employed individuals exposed to a single type of stressor, one
must resist the temptation to overgeneralize and assume that results would be replicable under different
conditions. Moreover, only a single adaptational outcome was examined, and additional potential
confounds such as neuroticism were not assessed. Hopefully, future researchers will examine a broader
array of adaptational outcomes and control variables, using multiple methods, in studies involving
larger sample sizes and prospective repeated-measures designs.

Acknowledgement-The authors gratefully acknowledge the assistance of Dr Debra Comer, who made the data
collection possible.

REFERENCES

Appel, M. A., Holroyd, K. A. & Gorkin, L. (1983). Anger and progression of physical illness. In Temoshok, L., VanDyke,
C. & Zegans, L. S. (Eds) Emorions in health and illness: Theoretical and research foundations (pp. 73-87). New York:
Grune & Stratton.
Aspinwall, L. G. & Taylor, S. E. (1992). Modeling cognitive adaptation: A longitudinal investigation of the impact of
individual differences and coping on college adjustment and performance. Joumaf OfPersonality and Social Psychology,
63.989-1003.
Averill, J. R. (1982). Anger and aggression: An essay on emotion. New York: Springer.
Ben-Porath, Y. S., Wailer, N. G. & Butcher, J. N. (1991). Assessment of coping: An empirical illustration of the problem
of inapplicable items. Journal of Personality Assessment, 57, 162-176.
Blaney, P. H. (1986). Affect and memory: A review. Psychological Bulletin, 99, 229-246.
Carver, C. S. & Gaines, J. G. (1987). Optimism, pessimism and physical well-being: The influence of generalized outcome
expectancies on health. Journal of Personality, 55, 169-210.
Chesney, M. A. & Rosenman, R. H. (1985).Anger and hostility in cardiovascular and behavioral disorders. Washington,
DC: Hemisphere.
Cohen, F. (1987). Measurement of coping. In Kasl, S. V. & Cooper, C. L. (Eds), Stress and health: Issues in research
methodology. New York: John Wiley.
Cooper, C. L. & Payne, R. (1988). Causes, coping and consequences ofstress at work. New York: John Wiley.
Darlington, R. B. (1990). Regression and linear models. New York: McGraw-Hill.
Dember, W. N., Martin, S. H., Hummer, M. K., Howe, S. R. & Melton, R. S. (1989). The measurement of optimism and
pessimism. Current Psychology: Research & Reviews, 8, 102-I 19.
838 Kenneth E. Hart and James B. Hittner

Desharnais, R., Godin, Cl., Johin, J., Valois, P. & Ross, A. (1990). Optimism and health-relevant cognitions after a myocardial
infarction. Psychological Reports, 67, 113 l-1 135.
Fleishman, J. A. (1984). Personality characteristics and coping patterns. Journal ofHealrh and Social Behavior, 25.229-244.
Folkman, S. (1984). Personal control and stress and coping processes: A theoretical analysis. Journal of Personality andSocial
Psychology, 46, 839-856.
Folkman, S. (1992). Commentary to part three: Improving coping assessment: Reply to Stone & Kennedy-Moore. In Friedman,
H. S. (Ed.), Hosrilhy, coping and healrh (pp. 215-223) Washington, DC: American Psychological Association.
Folkman, S., Lazarus, R. S., Gruen, R. J. & DeLongis, A. (1986). Appraisal, coping, health status, and psychological
symptoms. Journal of Personality and Social Psychology, 50, 571-579.
Friedman, H. S. (Ed.), (1992). Hostility, coping and health. Washington, DC: American Psychological Association.
Friedman, L. C., Nelson, D. V., Baer, P. E., Lane, M., Smith, F. E. & Dworkin, R. J. (1992). The relationship of dispositional
optimism, daily life stress and domestic environment to coping methods used by cancer patients. Journal of Behavioral
Medicine, 15, 127-141.
Guarnara, S. & Williams, R. L. (1987). Optimism and locus of control for health and affiliation among elderly adults. Journal
of Gerontology, 42, 594-595.
Hart, K. E. & Cardozo, S. R. (1986). Ways of appraising and coping with anger-provoking transactions among high and low
hostile/cynical individuals. Paper presented at the Convention of the Societyfor BehavioralMedicine, San Francisco, CA.
Hart, K. E., Turner, S. H. & Cardozo, S. R. (1987). Stress, personality, coping and well-being: A four-wave longitudinal
process study. Personaliq and Individual Differences, 8, 591-593.
Holroyd, K. A. & Gorkin, L. (1983). Young adults at risk for hypertension: Effects of family history and anger management
in determining responses to interpersonal conflict. Journal of Psychosomatic Research, 27, 131-138.
Johnson, E. H. (1990). The deadly emotions: The role of anger, hostilhy, and aggression in health and emotional well-being.
New York: Praeger.
Julius, M., Harburg, E., Cottington, E. M. & Johnson, E. H. (1986). Anger-coping types, blood pressure, and all-causes
mortality: A follow-up in Tecumseh, Michigan (1971-1983). American Journal ofEpidemiology, 124, 220-233.
Lazarus, R. S. & Folkman, S. (1984). Stress, appraisal and coping. New York: Springer.
Maiuro, R. D., Cahn, T. C., Vitaliano, P. P., Wagner, B.C. & Zegree, R. (1988). Anger, hostility and depression in domestically
violent versus generally assaultive men and nonviolent control subjects. Journal of Consulting and Clinical Psychology,
56, 17-23.
Marshall, G. N., Wortman, C. B., Kusulas, J. W., Hervig, L. K. & Vicekers, R.R. (1992). Distinguishing optimism from
pessimism: Relations to fundamental dimensions of mood and personality. Journal of Personality and Social Psychology,
62, 1067-1074.
McCrae, R. R. (1984). Situational determinants of coping responses: Loss, threat and challenge. Journal of Personaliry and
Social Psychology, 46, 9 19-928.
McCrae, R. R. (1992). Situational determinants of coping. In Carpenter, B. N. (Ed.), Personal coping: Theory, research and
upplicarion (pp. 65-76). New York: Praeger.
Mook, J., Kleijn, W. C. & Van der Ploeg, H. M. (1992). Positively and negatively worded items in a self-reported measure
of dispositional optimism. Psychologicul Reports, 7/, 275-278.
Mroczek, D. K., Spiro, A., Aldwin, C. M., Ozer, D. J. & Bosse, R. (1993). Construct validation of optimism and pessimism
in older men: Findings from the Normative Aging Study. Health Psychology, 12, 406-409.
Novaco, R. W. (1975). Anger control: The development and evaluation of an experimental treatment. Lexington, MA:
D. C. Heath.
Novaco, R. W. (1986). Anger as a clinical and social problem. In Blanchard, R. J. & Blanchard, D. C. (Eds), Advances in
the study ofuggression (Vol 2). New York: Academic Press.
Plomin, R., Scheier, M. F., Bergeman, C. S., Pedersen, N. L., Nesselroade, J. K. & McClearen, G. E. (1992). Optimism,
pessimism and mental health: A twin/adoption analysis. Personality and individual Differences, 13, 921-930.
Roskies, E. (1980). Considerations in developing a treatment program for the coronary-prone (Type A) behavior pattern. In
Davidson, P. 0. & Davidson, S. M. (Eds), Behavioral medicine: Changing health lifesfyles (pp. 299-333). New York:
BrunnerMazel.
Rule, B. & Nesdale, A. (1976). Emotional arousal and aggressive behavior. Psychological Bulletin, 83, 851-863.
Scheier, M. F. & Carver, C. S. (1985). Optimism, coping and health: Assessment and implications of generalized outcome
expectancies. Health Psychology, 4, 2 19-247.
Scheier, M. F. & Carver, C. S. (1987). Dispositional optimism and physical well-being: The influence of generalized outcome
expectancies on health. Journal of Personali~, 5.5, 169-210.
Scheier, M. F., Weintraub, J. K. & Carver, C. S. (1986). Coping with stress: The divergent strategies of optimists and
pessimists. Journal of Personality and Social Psychology, 51, 1257-1264.
Siegman, A. W. & Smith, T. W. (1994) (Eds.), Anger, hostiliry and the heart. Hillsdale, N.J: Lawrence Erlbaum.
Smith, T. W., Pope, M. K., Rhodewalt, F. & Poulton, J. L. (1989). Optimism, neuroticism, coping, and symptom reports:
An alternative interpretation of the life orientation test. Journal of Personality and Social Psychology, 56, 640-648.
Spielberger, C. D., Jacobs, G., Russell, S. & Crane, R. (1983). Assessment of anger: The State-Trait Anger Scale. In Butcher,
J. N. & Spielberger, C. D. (Eds), Advances in personaliry assessment (Vol 2, pp. 161-189). Hillsdale, NJ: Lawrence
Erlbaum.
Stevens, J. (1986). Applied multivariate sfaristics for rhe social sciences. Hillsdale, NJ: Lawrence Erlbaum.
Stone, A. A. & Kennedy-Moore, E. (1992). Commentary to part three: Assessing situational coping: Conceptual and
methodological considerations. In Friedman, H. (Ed.), Hostility, coping and health. Washington, DC: American
Psychological Association.
Strack, S., Carver, C. S. & Blaney, P. H. (1987). Predicting successful completion of an aftercare program following treatment
for alcoholism: The role of dispositional optimism. Journal of Personality and Social Psychology, 53, 579-584.
Strutton, D. L. & Lumpkin, J. (1992). Relationship between optimism and coping strategies in the work environment.
Psychological Reports, 71, 1179-l 186.
Suarez, E. L. & Williams, R. B. (1989). Situational determinants of cardiovascular and emotional reactivity in high and low
hostile men. Psychosomatic Medicine, 51, 4OU18.
Taylor, S. E. & Schneider, S. K. (1989). Coping and the simulation of events. Social Cognition, 7, 174-194.
 Coping 839

Thomas, S. P. & Donnellan, M. M. (1991). Correlates of anger symptoms in women in middle adulthood. American Journal
of Health Promotion, 5, 264-272.
Zeidner, M. (1993). Coping with disaster: The case of Israeli adolescents under threat of missile attack. Journal of Youth and
Adolescence, 22, 89-108.
Zeidner, M. & Ben-Zur, H. (1993). Coping with a national crisis: The Israeli experience with the threat of missile attacks.
Personality and Individual Differences, 14, 209-224.
Zeidner, M. &Hammer, A. L. (1992). Coping with missile attack: Resources, strategies and outcomes. Journal of Personal@,
60.709-740.