THE IMPACT OF GENERAL AND SPECIFIC RATIONAL AND IRRATIONAL BELIEFS ON EXAM DISTRESS; A FURTHER INVESTIGATION OF THE BINARY MODEL OF DISTRESS AS AN EMOTIONAL REGULATION MODEL

Terry DILORENZO*1, Daniel DAVID2, 3, Guy H. MONTGOMERY 3

1Stern College for Women, Yeshiva University, New-York, USA 2Babes-Bolyai University, Cluj-Napoca, Romania 3Mount Sinai School of Medicine, New York, USA

 

Abstract

The aims of the present study are to examine the relations between both general and specific rational and irrational beliefs, personality characteristics (optimism, pessimism), context-dependent factors (response expectancies), general distress and functional and dysfunctional distress in an academic exam situation. Although previous studies have investigated the contribution of both cognitive factors (e.g., irrational beliefs, response expectancies) and personality traits (e.g., optimism, pessimism) to exam-related distress, these studies have failed to distinguish between the specific contributions of irrational and rational beliefs to functional and dysfunctional distress (consistent with a binary model of distress). In a sample of 86 students facing an impending academic exam we examined these relationships. The binary model of distress was supported by results showing that an increase in specific rational beliefs and a decrease in specific irrational beliefs from the beginning of the term to the exam period were accompanied by an increase of functional distress and a decrease of dysfunctional distress (p’s < .05). Theoretical and clinical implications of the results are discussed.

 

Keywords: irrational beliefs, rational beliefs, binary model of distress, exam distress
Introduction

 

Rational Emotive and Cognitive Behavioral Psychotherapies (REBT/CBT) are based on Albert Ellis’ ABCDE cognitive model of distress (Ellis, 1962; 1994). In this model, undesirable events activate (A) either rational (adaptive) and/or irrational (maladaptive) beliefs (B). These beliefs lead to consequences (C), which can be emotional, behavioral and/or cognitive in nature. As would be expected, the consequences of rational beliefs are adaptive, while the consequences of irrational beliefs are maladaptive. The goal of REBT/CBT is to challenge (D) irrational beliefs and acquire more efficient (E) beliefs.

Current conceptualizations of REBT/CBT (e.g., DiGiuseppe, Leaf, Exner, & Robin, 1988; Ellis, 1994; Wallen, DiGiuseppe, & Dryden, 1992) focus on four categories of irrational cognitive processes. The first, demandingness, refers to absolutist requirements expressed in form of “musts,” “should,” and “oughts” (e.g., “I must pass the exam”). Awfulizing/ catastrophizing refers to beliefs that a situation is viewed as the worst thing that could happen, and worse than it absolutely should be (e.g., “It is awful that I did not pass the exam”). The third category is low frustration tolerance in which individuals believe that they will not be able to endure situations or to be happy if what they demand does not exist (e.g., “I could not stand not passing the exam”). Finally, global evaluation and self-downing occurs when individuals overgeneralize about others, themselves, and the world. In this case, people are excessively critical of others, the self and the world and make global negative evaluations (e.g., “I am stupid and worthless because I did not pass the exam”).

Each category of irrational beliefs has a rational counterpart. The rational counterpart to demandingness is desires/preferences. Here, people express their beliefs in the form of wishes, wants, and preferences rather than dogmatic “musts,” “should,” and “oughts.” (e.g., “I really want to pass the exam, but I am aware that nowhere is it written that it absolutely must happen”). A moderate evaluation of badness is the counterpart to awfulizing, in which people evaluate negative events as bad rather than awful (e.g., “Indeed, it is very bad that I did not pass the exam, but this is not the end of the world”). The rational counterpart to low frustration tolerance is frustration tolerance, expressed by beliefs in one’s ability to tolerate discomfort (e.g., “It is not a good feeling to fail the exam, but I can stand it”). Finally, acceptance of fallibility is the counterpart to global evaluation/self-downing. Individuals here believe that people cannot be evaluated by a single global rating and that one should evaluate only his/her performance, rather than his/her overall being (e.g., “I did not pass the exam. I behaved stupidly by not preparing enough for the exam, but this does not mean that I am stupid and worthless”).

Irrational beliefs are important psychological constructs (i.e., cognitive vulnerability constructs), which can predict how individuals react emotionally to stressful events. Their role in emotional regulation has been described by David and Szentagotai (2006). Indeed, a large body of empirical evidence shows that irrational beliefs impact emotional problems such as self-reported trait anxiety, trait anger, depression (Bernard, 1998), state anxiety, depression, anger, guilt (David, Belloiu, & Schnur, 2002), major depressive disorder (David, Szentagotai, Lupu, & Cosman, 2008; McDermutt, Haaga, & Bilek, 1997; Solomon, Arnow, Gotlib, & Wind, 2003), dysforia (McDermutt, et al., 1997) and test anxiety (Malouff, Schutte, & McClelland, 1992).

Although research has investigated the contribution of irrational beliefs to exam-related distress (e.g., Malouff et al., 1992), the literature suffers from several limitations. First, much of the prior research has not examined the independent contributions of irrational and rational beliefs. Moreover, previous studies that did distinguish between irrational and rational beliefs simply conceptualized rational beliefs as low scores on measures of irrational beliefs. Given that rational and irrational beliefs are no longer considered bipolar constructs (e.g., Bernard, 1998; David, Lynn, & Ellis, 2009) these measures may not actually distinguish between irrational and rational beliefs.

A second limitation is that research generally adopts a unitary model of distress, failing to make a distinction between functional and dysfunctional distress. Ellis (1994) proposed a binary model of distress which is comprised of qualitatively different, but interrelated, functional (e.g., sad, concerned/worried) and dysfunctional (e.g., depressed mood, panicked) negative feelings. Dysfunctional distress is accompanied by functional distress (e.g., feeling depressed also involves sadness); however, the reverse is not true (e.g., one can be sad but not feel depressed). Defined in this way, dysfunctional negative feelings (i.e., dysfunctional distress) would correspond to clinically significant problems (e.g., anxiety, depression), while functional negative feelings (i.e., functional distress) (e.g., concern, sadness) correspond to normal negative reactions to stressful events (e.g., Bonnano, 2004; Ellis & DiGiuseppe, 1993). (See Figure 1). Thus, while rational beliefs are hypothesized to be associated to functional distress, irrational beliefs are hypothesized to be associated to dysfunctional distress and their functional components (e.g., if you are panicked you are also concerned). Therefore, testing the binary model of distress is difficult (for further discussion see David et al., 2009; David, Montgomery, Macavei, & Bovbjerg, 2005). For example, while an association between irrational beliefs and dysfunctional distress should be expected (e.g., a positive correlation), the correlation between rational beliefs and functional feelings is unpredictable because functional feelings are also components of dysfunctional feelings. Some empirical support for a binary model of distress has been found in studies using factor analyses and group comparisons (i.e., participants with high versus low irrational beliefs) (e.g., David, Schnur, & Belloiu, 2002; David et al., 2005). However, research based on the binary model of distress is scant, and no study has investigated the binary model of distress in the context of an impending academic exam. The exam situation is useful as a model for studying sources of distress as it provides a uniform distress-provoking stimulus in a real-life setting.

Another limitation in the literature assessing rational and irrational beliefs is the lack of distinction between the general and situation-specific nature of beliefs (David et al., 2009; Smith, 1982). It has been suggested that situation-specific rational and irrational beliefs are better predictors of specific outcomes than are general beliefs (e.g., Ellis, 1994). While some scales have been developed to assess irrational beliefs related to specific situations (e.g., Lohr,

 

Brandt, & Bonge, 1977), no study has simultaneously investigated the independent contributions of specific and general beliefs on distress.

Finally, prior research on trait characteristics and context-dependent factors and distress has not examined their relative contribution to functional and dysfunctional distress. The personality literature has supported associations between factors such as optimism and pessimism and exam-related distress (e.g., Carver & Scheier, 1992; Snyder, Harris, et al., 1991). While optimism is negatively associated with distress, pessimism is positively associated with distress. These factors typically account for 16% to 36% of the variance in exam-related distress and perceptions of scholastic competence (e.g., Snyder et al., 1997).  Optimism and pessimism have also been found to be related to rational and irrational beliefs (Ziegler, 1999); however, research has not examined the predictive value of these personality factors specifically on functional and dysfunctional distress. Such information is critical as it might enable us to separate benign personality characteristics from those more likely to lead to heightened dysfunctional distress.

Context-dependent factors, such as response expectancies (i.e., expectancies for nonvolitional outcomes such as anxiety or pain), may also be predictive of exam-related distress.  Based on response expectancy theory (Kirsch, 1999), one can hypothesize that expectations about distress prior to an exam would be highly predictive of such distress.  Although these relations have not been tested in regard to examinations, both the clinical (Montgomery, Weltz, Seltz, & Bovbjerg, 2002; Sohl, Schnur, & Montgomery, 2009) and experimental (Montgomery & Kirsch, 1997) literatures strongly support the existence of relations between response expectancies and distress. Kirsch (1990) even suggests that the relation between response expectancy and outcome is causal such that the expectation for exam distress would be sufficient to cause such distress.  Although response expectancies were found to be related to rational and irrational beliefs (e.g., Montgomery, David, DiLorenzo, & Schnur, 2007), to our knowledge, relations between expectations and functional and dysfunctional distress have never been tested.

The present study will examine the relative contributions of rational and irrational beliefs, personality characteristics and context-dependent factors to general, functional and dysfunctional distress at an academic examination. In this context we will examine the relative merit of the distinction between general and specific irrational beliefs and the unitary and binary models of distress. While we formulate specific predictions for the unitary model, most of the analyses for the binary model will be exploratory. The following specific predictions were made.

First, we propose that we will find support for both unitary and binary models of distress as related to rational and irrational beliefs. Specifically, with regard to the unitary model of distress, we predict that general and specific rational beliefs will be negatively associated with general distress, while general and specific irrational beliefs will be positively related to general distress. Regarding a binary model of distress, we predict that general irrational beliefs will be positively associated to dysfunctional distress. Concerning specific rational and irrational beliefs and binary distress, based on Ellis’s model (1994) we predict (see Figure 1) that a change in functional and dysfunctional distress levels from the beginning of the term to the exam period should be accompanied by a theoretically-related change in specific rational or irrational beliefs. For example, an increase (or decrease) in dysfunctional distress should be accompanied by an increase (or decrease) in specific irrational beliefs; an increase (or decrease) in functional distress should be accompanied by an increase (or decrease) in specific rational beliefs (if irrational beliefs do not increase or decrease in the same direction; see Figure 1). It is also expected that specific rational and irrational beliefs will make a stronger contribution to distress than general rational and irrational beliefs.

Our second hypothesis predicts that based on a unitary model of distress, optimism will be negatively associated with general distress, while pessimism will be positively related to general distress. With regard to the binary distress model, optimism will be negatively associated with dysfunctional distress, while pessimism will be positively related to dysfunctional distress. No specific predictions are made about the relation between optimism, pessimism, and functional distress.

Our third hypothesis concerns the association of response expectancies and distress. Again, we expect to find support for the unitary model of distress. Based on the unitary model, and according to response expectancy theory (Kirsch, 1999), response expectancy for anxiety will be positively associated with general distress at the exam, while response expectancy for relaxation will be negatively correlated to distress at the exam. In this context we can also formulate specific predictions for the binary distress model. Response expectancy for relaxation should be negatively associated to both functional and dysfunctional distress at the exam, while response expectancy for anxiety should be positively correlated to both functional and dysfunctional distress at the exam (feeling anxious also involves feeling concerned).

Finally, based on the prior literature, we expect that general and specific rational and irrational beliefs, personality characteristics and response expectancies will be correlated with one another. Specifically, we predict that rational beliefs will be positively correlated with optimism and response expectancy for relaxation and that irrational beliefs will be positively correlated with pessimism and response expectancy for anxiety. We will also examine the relative contribution of these variables to exam-related distress (unitary and binary models).

 

Method

 

Participants

Eighty-six undergraduate students from a private women’s college in the US who were 18 years of age or above (m=22; sd =8.21), participated. Participants were recruited from psychology classes. Written informed consent was obtained prior to participation.

 

Measures

Participants completed the following measures:

Predictor Variables

Rational and irrational beliefs:

General measure: The ABS-II is a self-report scale designed to measure irrational beliefs. It has 72 items, with a 5-point response scale and three components: (1) cognitive processes (demandingness, awfulizing, low frustration tolerance, and global evaluation/self-downing); (2) content (approval, achievement, and comfort) and (3) wording modality (items worded in rational terms vs. items worded in irrational terms). The ABS-II has been demonstrated to be a reliable and valid measure of both rational and irrational beliefs (e.g., David et al., 2002; DiGiuseppe, Leaf, Exner, & Robin, 1988). The scores of the irrational and rational belief subscales were computed by summing the scores of each irrationally-worded and rationally-worded items, respectively; higher scores indicate higher levels of either irrational or rational beliefs. Cronbach’s alphas every subscales were higher than .75

Specific measure: The Exam Beliefs Scale (EBS) is an 8-item scale constructed as a specific measure of exam-related rational and irrational beliefs for this study (see Appendix 1). The scale was developed based on Wallen, DiGiuseppe, & Dryden’s REBT/CBT guide (1992) by a group of experts in REBT to assess the 4 rational and rational processes measured by the ABS-II (DiGiuseppe et al., 1988). Four items measure exam-related irrational beliefs (demandingness, awfulizing, low frustration tolerance, and global evaluation/self-downing), and four items measure exam-related rational beliefs (preferences, moderate evaluation of badness, frustration tolerance, non-global evaluation).  Participants indicate their agreement with each statement on a 4-point scale (strongly agree to strongly disagree). Scores for each subscale are obtained by summing the scores of each item on the subscale; a low score means a high level of specific rational or irrational beliefs.

Personality characteristics:

The Life Orientation Test (LOT) (Scheier & Carver, 1985) is a 12-item measure (8 items and 4 filler items) assessing dispositional optimism and pessimism, defined in terms of generalized outcome expectancies. Four of the items are positively worded (measuring optimism) and four are negatively worded (measuring pessimism). The LOT has demonstrated reliability and validity (Scheier & Carver, 1985). Although the LOT has, in the past, been considered to be a unidimensional measure of dispositional optimism, with scores on the negatively worded items reversed and summed with scores on the positively worded items, recent studies have found that it has a bidimensional structure (Chang & Bridewell, 1994; Chang, 1997). In accordance with these recent empirical findings, we followed the  bidimensional structure and generated scores for optimism and pessimism separately.

Response expectancies: Ten centimeter visual analog scales were used to assess specific expectations for distress and relaxation before the exam. Higher scores indicated higher levels of response expectancies for exam-related distress and relaxation at the exam period. This approach has been demonstrated to be effective and efficient in multiple studies (e.g., Montgomery & Kirsch, 1997; Montgomery, Weltz, Seltz, & Bovbjerg, 2002).

Outcome Variables

Distress: The Profile of Mood States-Short Version (Shacham, 1984; DiLorenzo, Bovbjerg, Montgomery, Jacobsen, & Valdimarsdottir, 1999) was used to assess distress. This measure is a 37-item mood adjective checklist; respondents indicate on a 4-point scale how much they have experienced each of the items. The items can easily be analyzed to measure general distress by summing up the score for each item (after reverse scoring positively worded items) and to discriminate functional distress (summing the scores of following items: blue, tense, sad, uneasy, on edge, anxious, nervous, restless) from dysfunctional distress (summing up the scores of the following items: miserable, angry, bitter, resentful, furious, worthless, annoyed, peeved) (based on David et al., 2005).

Exam Grades: Because distress levels may be associated with exam performance, participants’ scores on the midterm exam were obtained. To preserve anonymity, participants were assigned study ID numbers and only the Principal Investigator had access to identifying data.

 

Procedure

Participants completed four assessments:

Time One/Baseline:  All measures were given to participants to take home and return to research personnel at the start of the semester.

Time Two: One week prior to the midterm exam, students completed the POMS-SV and the EBS.

Time Three: Two days prior to the midterm exam students completed the POMS-SV and the EBS.

Time Four (exam distress): On the day of the midterm, just prior to beginning the exam, students completed the POMS-SV and the EBS.

 

Data Analysis

Data were entered and analyzed using SPSS. The first step in exploring the hypothesized relations was to use correlational (apriopri and post hoc) analyses to examine bivariate relations between variables.  Then a series of simultaneous regression (post hoc) analyses and group (a priori) comparisons were conducted.

Results

 

Prior to conducting analyses testing our hypotheses, we examined the levels of distress experienced at the four points and found significant differences (F (3, 83) = 4.6, p<.05). Distress levels at time one (start of the semester) and time four (at the exam) were higher than distress measured at times two and three. Given that we had two peaks of distress, subsequent analyses were conducted for these two time points only. We used midterm scores as covariates in subsequent analyses. However, results were essentially the same when analyses were conducted with or without this covariate.

 

Psychometric Properties of the Exam Beliefs Scale (EBS)

Reliability

Internal consistency reliability was assessed by Cronbach’s alphas (see Table 1) for the EBS at the four time points. All alphas were higher than .67, suggesting good reliability of the scale. Test-retest reliability was assessed by computing correlations between scores at the two peaks of distress (baseline and prior to the exam). Correlations suggest good temporal stability for both specific rational beliefs (r=.51, p<.05) and irrational beliefs (r=.69, p<.05).

 

Table 1. The psychometric properties of The Exam Beliefs Scale: Cronbach’s alphas and

correlations with general rationality and general irrationality subscales of the ABS

 

	Specific Rational Beliefs at Baseline	Specific Irrational Beliefs at Baseline	Specific Rational Beliefs at Exam	Specific Irrational Beliefs at Exam
Cronbach’s Alphas	.65	.69	.70	.68
General Rationality	-.34	.30	-.29	.43
General Irrationality	.26	-.39	.26	-.49

All results are significant at p<.05

 

 

Validity

            Concurent validity. Concurrent validity of the EBS was assessed by examining the associations between the rational and irrational beliefs subscales of the EBS with the same subscales of the ABS-II, an established measure of rational and irrational beliefs. As seen in Table 1, the rational beliefs subscale of EBS (at both baseline and the exam) was associated with the rational subscale of ABS-II; the same is true for irrational beliefs subscales.

Factor analysis. A principal components analysis with a Varimax rotation showed the presence of two factors (see Table 2) at both baseline and the exam period. One factor includes items that would indicate irrationality (factor 2 at baseline and factor 1 at the exam period) and the other factor (factor 1 at baseline and factor 2 at exam period) includes items considered to be rational beliefs. This factor structure is consistent with other findings in suggesting that rational and irrational beliefs seem to be two different constructs rather than extremes of a single construct (e.g., Bernard, 1998; David et al., 2009).

In a priori analyses, as predicted based on a unitary model of distress, general irrational belief scores were positively correlated with levels of general distress at both baseline and the exam period. Also consistent with our predictions, we found that general rational beliefs were negatively correlated with general distress at both baseline and the exam period (see Table 3). Specific rational beliefs at the exam period were negatively related to general distress at the exam period (r= .35, p<.05) while specific irrational beliefs at the exam period were positively related to general distress (r=-.33, p<.05) at the exam period. Specific irrational and rational beliefs at baseline were not related to general distress at the exam period (both p>.05). While specific rational beliefs at baseline were related to general distress at baseline (r=.24, p<.05), specific irrational beliefs were not (r=-.17, p<.05).

 

Table 2. Results of the factor analysis for the Exam Beliefs Scale at Baseline and Exam Period

	Factor 1 Baseline	Factor 2 Baseline	Factor 1 Exam Period	Factor 2 Exam Period
Dem	-.28	.83	.50	-.53
Pre	.27	-.65	-.16	.86
AWF	-.02	.83	.71	-.30
Anti-AWF	.79	.075	.04	.83
LFT	-.53	.17	.56	.07
FT	.45	-.18	-.65	-.02
GE/SD	-.58	.42	.63	-.40
Anti GE/SD	.77	-.28	-.60	.42

 

 

In evaluating the binary model of distress, we found that both general rational and irrational beliefs were associated with functional and dysfunctional distress at both baseline and the exam period (see Table 3). As predicted, based on a binary model of distress, general irrational beliefs were positively related to dysfunctional distress. In a simultaneous regression with general rationality and irrationality as predictors, neither general rational nor general irrational beliefs made an independent contribution to functional and dysfunctional distress (p’s>.05).

 

Table 3. Correlations between general rational and irrational beliefs and general (POMS),

functional, and dysfunctional distress at baseline and the exam period

 

		POMS Baseline		POMS Exam		Functional Distress at Baseline		Dysfunctional Distress at Baseline		Functional Distress at Exam		Dysfunctional Distress at Exam
General Rational Beliefs	-.46		-.28		-.41		-.38		-.29		-.22
General Irrational Beliefs	.43		.24		.42		.34		.24		.23

 

Functional distress was lower at baseline than at the exam period (t(85)=-2.84, p < .05) while dysfunctional distress was higher at baseline than at the exam period (t(85)=2.1, p < .05). Consistent with the binary model of distress and our hypotheses, specific rational and irrational beliefs followed a theoretically-related pattern. Thus, specific rational beliefs were lower at baseline as compared to the exam period (t(85)=2.35, p < .05), while specific irrational beliefs were higher at baseline than at the exam period (t(85)=-2.75, p < .05) (see Figure 2).

In simultaneous regression analyses with general rational beliefs and specific rational beliefs at baseline as predictors, only general rational beliefs significantly contributed to general distress at the exam period (p<.05). However, when both specific rationality measured at the  exam period and general rationality were used as predictors, only specific rationality at the exam period significantly contributed to general distress at the exam period (p<.05).

 

When both specific irrationality at the exam period and general irrationality at baseline were used as predictors in a simultaneous regression analysis, only specific irrational beliefs at the exam period predicted distress at the exam period (p<.05). When specific irrationality at baseline and general irrational beliefs were included as predictors, neither made an independent contribution to general distress at the exam period (p’s>.05).

Post hoc analyses included a simultaneous regression analysis with general rationality and irrationality as predictors. Only general rationality predicted general distress at baseline (p<.05); people with higher levels of rational beliefs reported less general distress. Neither general rational nor general irrational beliefs had an independent contribution to general distress at the exam situation (all p’s>.05).

Our second hypothesis was that the unitary model of distress would be supported with regard to personality characteristics. Consistent with this hypothesis, in a priori analyses we found that pessimism was positively correlated with general distress at both baseline and the exam period, and optimism was negatively correlated with general distress at both baseline and the exam period (see Table 4). In a simultaneous regression with optimism and pessimism as predictors, optimism predicted general distress at both baseline and the exam period (all p’s<.05), such that less optimistic participants reported more distress. Pessimism contributed to general distress only at baseline (p<.05). As expected, lower levels of pessimism were associated with less general distress. Based on the binary model of distress (see Table 4), pessimism was positively related to dysfunctional feelings both at baseline and the exam period, and optimism was negatively related to dysfunctional feelings both at baseline and exam period.

 

Table 4. Correlations between optimism, pessimism and response expectancies and general (POMS), functional, and dysfunctional distress at baseline and the exam period

 

	POMS Baseline	POMS Exam	Functional Distress at Baseline	Dysfunctional Distress at Baseline	Functional Distress at Exam	Dysfunctional Distress at Exam
Optimism	-.54	-.39	-.42	-.45	-.36	-.26
Pessimism	.49	.34	.45	.46	.28	.35
Response Expectancy for Relaxation	–	-.34	–	–	-.31	-.20
Response Expectancy for Anxiety	–	.39	–	–	.38	.23

All coefficients are significant at p<.05.

 

 

With regard to the post hoc analyses in the case of the binary model of distress, the correlations between optimism, pessimism and functional and dysfunctional distress are presented in Table 4. When pessimism and optimism were simultaneously entered into a regression equation, high pessimism predicted both high functional and high dysfunctional distress at baseline (p’s < .05), but only high dysfunctional distress at the exam period (p<.05). High optimism predicted both low functional and low dysfunctional distress at baseline (all p’s<.05) and only low functional distress at the exam period (p<.05).

Our third hypothesis, that both the unitary and binary models of distress would be supported in association with response expectancies, was supported. In a priori analyses, as predicted based on unitary model of distress, both response expectancies for anxiety and for relaxation were related to distress measured at the exam period (see Table 4). Response expectancy for anxiety was positively correlated to distress at the exam period, while response expectancy for relaxation was negatively correlated to distress at the exam period. In a simultaneous regression with response expectancies for relaxation and anxiety as predictors, only response expectancy for anxiety made an independent contribution to general distress at both baseline and the exam period (p<.05). In regard to the binary model of distress, the correlations between response expectancies and functional and dysfunctional distress were all consistent with our predictions (see Table 4). When response expectancies for anxiety and relaxation were entered into a simultaneous regression equation, only response expectancy for anxiety had an independent contribution to functional distress at the exam period (p<.05) such that individuals with greater expectations for anxiety reported more functional distress. No response expectancy made an independent contribution to dysfunctional distress at the exam period (p’s>.05).

Consistent with our final hypothesis, we found significant correlations among most of our predictor variables (general and specific rational and irrational beliefs, optimism, pessimism, and response expectancies (relaxation and anxiety)).  In those cases where correlations were not significant, they were in the expected direction. (see Table 5).

To evaluate their relative contribution to general distress and functional and dysfunctional distress, the predictor variables (general and specific rational and irrational beliefs, optimism, pessimism, and response expectancies) were simultaneously introduced in a series of regression equations with general distress and functional and dysfunctional distress as outcomes (i.e., criteria), which takes into account the theoretical anticipated association between predictors and criteria.

Separate regressions were conducted to examine contributors to general distress at baseline and the exam period. At baseline, only optimism (high) predicted distress (low) (p<.05), while controlling for pessimism and general rational and irrational beliefs. At the exam period, only specific irrational beliefs at the exam period made an independent contribution to distress when general rational and irrational beliefs, specific rational and irrational beliefs (at both baseline and exam), optimism, pessimism, and response expectancies for relaxation and anxiety were used as predictors. As expected, the higher the specific irrational beliefs, the higher the distress.

Next, we examined predictors of functional and dysfunctional distress at baseline. No predictor in the equation (optimism, pessimism, and general rational and irrational beliefs) independently contributed to functional distress. In the regression model with dysfunctional distress at baseline as the outcome variable, only pessimism made a significant contribution (p < .05); higher levels of pessimism were associated with higher levels of dysfunctional distress. To examine the predictors of functional and dysfunctional distress at the exam period, we conducted separate simultaneous regression analyses with general rational and irrational beliefs, specific rational and irrational beliefs (at both baseline and exam), optimism, pessimism, and response expectancies for relaxation and anxiety as predictors. Only specific rational beliefs at the exam period predicted functional distress at the exam period (p<.05), and only pessimism made an independent contribution to dysfunctional distress at the exam period (p<.05).  Specifically, individuals with lower levels of rational beliefs reported greater levels of functional distress, and those with higher levels of pessimism reported more dysfunctional distress.

 

Table 5. Correlations between the personality and specific predictors of general (POMS), functional, and dysfunctional distress at baseline and the exam

	General Rational Beliefs	General Irrational Beliefs	Specific Rational Beliefs at Baseline	Specific Irrational Beliefs at Baseline	Specific Rational Beliefs at Exam	Specific Irrational Beliefs at Exam	Optimism	Pessimism	Response Expectancy for Relaxation	Response Expectancy for Anxiety
General Rational Beliefs	1	-.72**	-.34**	.30**	-.28**	.43**	.32**	-.38**	.32**	-.38**
General Irrational Beliefs		1	-.34**	-.39**	.25*	-.49**	-.30**	.37**	-.25*	.28**
Specific Rational Beliefs at Baseline			1	-.58**	.50**	-.60**	-.35**	.22*	-.21*	.15
Specific Irrational Beliefs at Baseline				1	-.42**	.68**	.24*	-.16	.20	-.22*
Specific Rational Beliefs at Exam					1	-.57**	-.45**	.29**	-.17	.10
Specific Irrational Beliefs at Exam						1	.37**	-.25*	.21*	-.22**
Optimism							1	-.60**	.30**	-.25*
Pessimism								1	-.17	.33**
Response Expect for Relaxation									1	-.53**
Response Expectancy for Anxiety										1

 

3*p<.05    **p<.01

Discussion and conclusions

 

The main strengths of the study are that we used different measures of general and specific irrational beliefs, measured rational and irrational beliefs separately, employed general distress and functional and dysfunctional distress as outcomes, and examined the relationships of these variables with other personality and context-specific predictors of distress.

That we found no differences between baseline distress (i.e., measured at the beginning of the term) and exam distress (on the day of the exam) was unexpected. However, both baseline distress and exam distress were greater than the distress measured one week and two days before the exam, showing that the peaks of the distress were at the start of the semester and on the day of the exam. This finding might be explained by studies showing that the beginning of the term can be a stressful period related to the school adaptation (Pennebaker 1998). Having the two peaks of the distress allowed us to examine our hypotheses in two different stressful contexts.

The idea that the specific rational and irrational beliefs are context-specific rather than trait-like factors was supported by two findings in the present study. First, the test-retest reliability of the EBS was moderate, indicating that the context in which it is administered influences responses. Second, the context-dependency notion is also supported by the moderate associations between specific and general measures of rational and irrational beliefs. While these findings are preliminary, they do point to the need for more research to explore the stability of these psychometric indicators in order to determine whether specific rational and irrational beliefs are indeed context-specific.

In examining the unitary model of distress, we found that general irrational beliefs were positively correlated with general distress at both baseline and the exam period, while generalrational beliefs were negatively correlated with general distress at both baseline and the exam period. These results replicate those of Malouff et al. (1993) and are consistent with research in CBT/REBT, showing the protective role of rational beliefs and the vulnerability role of irrational beliefs in stressful situations (e.g., Smith, 1982). Our post hoc analyses suggest that during stressful events (i.e., the exam period) general rational beliefs are more strongly related to general distress than are general irrational beliefs. Specific rational beliefs at the exam period were negatively correlated to distress at the exam period and specific irrational beliefs at the exam period were positively correlated to distress at the exam period. These results suggest that specific irrational and rational beliefs play vulnerability and protective roles, respectively, in stressful situations. Moreover, results indicate that these roles can only be detected when assessed during the stressful event. These findings replicate previous research on the role of general rationality and irrationality in distress (e.g., Malouff et al., 1993), by using a specific measure of rational and irrational beliefs.

General rational and irrational beliefs were strongly associated with both functional and dysfunctional distress at both baseline and the exam period (except dysfunctional distress at the exam period). When entered simultaneously into a regression equation, only general rational beliefs had an independent contribution to functional distress at both baseline and the exam period (i.e., those reporting higher general rational beliefs reported less functional distress). Unfortunately, the rational belief items are constructed as the opposite of irrational beliefs in this measure, perhaps artificially creating a bipolar relation between rational and irrational items (i.e., high rational is low irrational), which would limit our ability to understand this result in the context of the binary model of distress. Indeed, the scores of general rational beliefs and general irrational beliefs were highly correlated in this sample (r= -.72), suggesting a bipolar relation. Future studies should further explore this relation (i.e., general rational beliefs and functional feelings) using rational and irrational beliefs scales that do not measure them as bipolar constructs.

Functional distress was lower at baseline than at the exam period, while a reverse pattern was found for dysfunctional distress. Specific rational beliefs were lower at baseline as compared to the exam period, while specific irrational belies were higher at baseline than at the exam period. This pattern suggests that an increase in specific rational beliefs about the exam and a decrease of specific irrational beliefs generate mainly functional distress, a pattern consistent with Ellis’ binary model of distress (see also David et al., 2005). This pattern also suggests that specific rational and irrational beliefs are less stable than general beliefs. It is possible that either they might become more available and accessible during stressful situations (e.g., the exam period) and thus, mediate the impact of the stressful event on distress, or that they are state dependent (e.g., exam distress makes them more accessible). Taking into account the moderate correlations between specific rational and irrational beliefs at baseline and the exam period, we tend to support the idea that the stressful event activates the specific rational and irrational beliefs which then mediate the impact of the stressful event on distress. However, future studies should investigate these possibilities.

Total general rationality measured at baseline seemed to be a better predictor of general distress at the exam period than the specific rationality measured at baseline, but only specific rationality measured at the exam period made an independent contribution to distress at the exam when both general rationality and specific rationality at the exam were used as predictors. In the case of irrationality, specific irrational beliefs at the exam period better predicted distress at the exam period than did general irrational beliefs; when both specific irrational beliefs at baseline and general irrational beliefs were used as predictors, neither had an independent contribution. Overall, these findings seem to suggest that specific measures of rational and irrational beliefs make a unique contribution to general distress, beyond the contribution of general rational and irrational beliefs. This is particularly notable when the specific beliefs are measured during the stressful event. Here, they seem to cancel out the contribution of general rational and irrational beliefs. Thus, specific rational and irrational beliefs may partially or totally mediate the impact of general rational and irrational beliefs on distress. This finding highlights the need to evaluate specific rational and irrational beliefs as related to a stressful event, particularly in the context of REBT/CBT (Ellis, 1994).

The relations between optimism, pessimism, and response expectancies and general distress measured at both baseline and the exam period were anticipated and are consistent with previous findings in the field (e.g., David, Montgomery, & Bovberg, 2006). Both optimism and response expectancy for relaxation were negatively associated with distress; both pessimism and response expectancy for anxiety were positively associated with distress. Analyses examining the associations of these variables with functional and dysfunctional distress reveal interesting relationships. High pessimism was associated with both high functional and dysfunctional distress at baseline and the exam period, while high optimism was associated with low functional and dysfunctional distress at both baseline and the exam period. In the case of response expectancies, we found that high response expectancy for anxiety was associated to high functional and dysfunctional distress at the exam period, while high response expectancy for relaxation was associated to low functional and dysfunctional distress during that period. Future studies should further explore these relations in the binary model of distress.

Consistent with previous research and theoretical models (Chang & Bridewell, 1998; David et al., 2006), pessimism, optimism, and response expectancies were associated with both general and specific measures of rational and irrational beliefs. Overall, in this study, the personality-type predictors (i.e., general rational and irrational beliefs, optimism, and pessimism) and the more specific predictors (i.e., specific rational and irrational beliefs, response expectancy) were highly correlated. However, post hoc analyses revealed some unique contributions. When all theoretically relevant predictors were put simultaneously into a regression equation, general distress at baseline was predicted only by optimism, while general distress at the exam was predicted only by specific irrational beliefs at the exam. Optimism made an independent contribution to functional distress at baseline, but no predictor made an independent contribution to dysfunctional distress at baseline. Functional distress at the exam period was predicted by specific rational and irrational beliefs at the exam period, and no predictor made an independent contribution to dysfunctional distress at the exam. Taken together, these results highlight the importance of specific irrational beliefs in distress during stressful events, beyond the contribution of other predictors. Specific irrational beliefs seem to mediate the impact of the other predictors on distress. This possibility should be further explored in future research.

Our investigation of the relative utility of the unitary and binary models of distress yielded complex results. Overall, our data replicated and extended the previous findings in the field, as related to the unitary model of distress (e.g., David et al., 2005; David et al., 2006), and offered new support for the binary model of distress. As predicted by the unitary model of distress, general distress was associated with both personality and specific predictors. When general distress was decomposed according to the binary model into more specific functional and dysfunctional distress, the specific measures (i.e., not the general measures) of rational and irrational beliefs predicted the pattern of functional and dysfunctional distress consistent with the binary model (Ellis, 1994). These results suggest that specific rational and irrational beliefs might be better predictors of functional and dysfunctional distress than are general rational and irrational beliefs or other investigated factors. Indeed, as predicted by the binary model of distress, pessimism, optimism, and response expectancies seem to make interesting contributions to functional and dysfunctional distress, but they are less differential. Future analyses should elaborate a more integrative theory of the binary model of distress, based on these complex results.

Our results supporting the binary model of distress may have important clinical implications. Clinicians could distinguish between functional and dysfunctional distress responses and intervene accordingly (David et al., 2005). For example, it might be functional for someone to feel sad about an upcoming surgery; however, it might be dysfunctional to feel depressed or indifferent about it. That is, depressed mood may contribute not only to unnecessary negative subjective experiences, but also to inappropriate responses (e.g., self-blame), while feeling indifferent is likely to reduce the motivational resources to deal with the stressor (Anderson, 1994).  On the other hand, although accompanied by a negative subjective experience, sadness might be associated with adaptive coping mechanisms to deal with the stressor (Bonnano, 2004). As such, the binary model suggests focusing clinical interventions on reducing dysfunctional distress while recognizing the appropriateness of functional distress. Interventions based on a unitary model of distress would not discriminate between functional and dysfunctional distress (David et al., 2005). Future studies should investigate the implications of the unitary and binary models of distress in both theoretical and clinical contexts.

The study is not without limitations. First, the EBS was developed for this study, and therefore, a thorough investigation of its psychometric properties has not been done. However, the indicators for its reliability and construct validity in this sample were acceptable. Next, our sample was relatively small, perhaps limiting statistical power to reveal other relations for which we can see only trends in this sample. Third, our investigation focused only on distress related to an academic exam. The patterns obtained in this study in relation to this stressful condition should be replicated in other stressful conditions to explore the generalizability and robustness of these findings. Finally, having only women in this study limits the generalizability of the results. Future research should replicate these findings in a more heterogeneous sample.

 

REFERENCES

Alden, L., & Safran, J. D. (1978). Irrational beliefs and non-assertive behavior. CognitiveTherapy and Research, 2, 357-364.

Anderson, K. J. (1994). Impulsivity, caffeine, and task difficulty: A within-subjects test of the Yerkes-Dodson law. Personality and Individual Differences, 6, 813-829.

Beck, A. T., & Beamesderfer, A. (1974). Assessment of depression: The depression inventory. In P. Pichot (Ed.), Modern problems in pharmacopsychiatry: Psychological measurements in psychopharmacology (pp. 151-169). New York: Karger, Basel.

Bernard, M. E. (1998). Validations of general attitude and beliefs scale. Journal of Rational-Emotive and Cognitive-Behavior Therapy, 16, 183-196.

Bonanno, G. A. (2004). Loss, Trauma, and Human Resilience – How We Underestimated the Human Capacity to Thrive After Extremely Aversive Events? American Psychologist, 59, 20-28.

Carver, C. S., Pozo, C., Harris, S. D., Noriega, V. Scheier, M. F., & Robinson, D. S. (1993). How coping mediates the effect of optimism on distress: A study of women with early stage breast cancer. Journal of Personality and Social Psychology, 65, 375-390.

Carver, S. C., & Scheier, F.S. (1994).  Situational Coping and Coping Dispositions in Stressful Transaction. Journal of Personality and Social Psychology, 66, 184-195.

Chang, E. C. (1997). Irrational beliefs and negative life stress: Testing a diathesis-stress model of depressive symptoms. Personality and Individual Differences, 22, 115-117.

Chang, E. C., & Bridewell, W. B. (1998). Irrational beliefs, optimism, pessimism, and psychological distress: A preliminary examination of differential effects in a college population. Journal of Clinical Psychology, 54, 137-142.

David, D., Schnur, J., & Belloiu, A. (2002). Another search for the “hot” cognitions: Appraisal, irrational beliefs, attributions, and their relation to emotion. Journal of Rational-Emotive and Cognitive-Behavior Therapy, 15, 93-131.

David, D., Montgomery, M., Macavei, B., & Bovbjerg, D. (2005). An empirical investigation of Albert Ellis’ binary model of distress. Journal of Clinical Psychology, 61, 499-516.

David, D., Montgomery, G. H., & Bovbjerg, D. H. (2006). Relations between coping responses and optimism-pessimism in predicting anticipatory psychological distress in surgical breast cancer patients Personality and Individual Differences, 40, 203-213.

David, D., Szentagotai, A., Lupu, V., & Cosman, D. (2008). Rational emotive behavior therapy, cognitive therapy, and medication in the treatment of major depressive disorder: A randomized clinical trial, posttreatment outcomes, and six-month follow-up. Journal of Clinical Psychology, 64, 728-746.

David, D., Lynn, S., & Ellis, A. (2009). Rational and irrational beliefs. Research, theory, and clinical practice. Oxford University Press, London/New-York.

DiGiuseppe, R., Leaf, R., Exner, T., & Robin, M. (1988). The development of a measure of irrational/rational thinking. Paper presented at the World Congress of Behavior Therapy, Edinburg, Scotland.

DiLorenzo, T. A., Bovbjerg, D. H., Montgomery, G. H., Valdimarsdottir, H., & Jacobsen, P. B.  (1999). The application of a shorted version of the profile of mood states in a sample of breast cancer chemotherapy patients. British Journal of Health Psychology, 4, 315-325.

Elkin, I., Shea, M. T., Watkins, J. T., Imber, S. D., Sotski, S. M., Collins, J. F. et al.  (1989). National Institute of Mental Health Treatment of Depression Collaborative Research Program: General effectiveness of treatments. Archives of General Psychiatry, 46, 971-982.

Ellis, A. (1962). Reason and emotion in psychotherapy. New York: Stuart.

Ellis, A. (1994). Reason and emotion in psychotherapy (Rev. Ed.). Secaucus, NJ: Birscj Lane.

Ellis, A., & DiGiuseppe, R. (1993). Are inappropriate or dysfunctional feelings in rational-emotive therapy qualitative or quantitative? Cognitive Therapy and Research, 5, 471-477.

Goldfried, M. R., & Sobocinski, D. (1975). Effect of irrational beliefs of emotional arousal. Journal of Consulting and Clinical Psychology, 43, 504‑510.

Hollon, S. D., DeRubeis, R . J., Evan, M. D., Wiemer, M. J., Garvey, M. J., Grove, W. M., & Tuason, V. B., (1992). Cognitive therapy and pharmacotherapy for depression: Singly and in combination. Archives of General Psychiatry, 49, 774-781.

Kirsch, I. (1999).  How expectancies shape experience.  (1st ed.).  Washington, DC:  American Psychological Association.

Kirsch, I. (1990).  Changing expectations: A key to effective psychotherapy.  Pacific Grove, CA:  Brooks/Cole.

Kovacs, M., Rush, A. J., Beck, A. T., & Hollon, S.D., (1981). Depressed outpatients treated with cognitive therapy or pharmacotherapy. Archives of General Psychiatry, 38, 33-39.

Jones, R. A.  (1968). A facored measure of Ellis’ irrational beliefs system with personality and maladjustment correlates. Unpublished doctoral dissertation. Texas Technical College.

Linder, H., Kirkby, R., Wertheim, E., & Birch, P. (1999). A brief assessment of irrational thinking: The shortened General Attitude and Beliefs Scale. Cognitive Therapy and Research, 23, 651-663.

Malouff, J. M., Schutte, N. S., & McClelland, T. (1992). Examination of the relationship between irrational beliefs and state anxiety.  Personality and Individual Differences, 13, 451-456.

McDermut, J. F., Haaga, A. A. F., & Bilek, L. A. 91997). Cognitive bias and irrational beliefs in major depression and dysphoria. Cognitive Therapy and Research, 21, 459-476.

Montgomery, G. H, David, D., DiLorenzo, T. A., & Schnur, J. B. (2007). Response expectancies and irrational beliefs predict exam-related distress. Journal of Rational-Emotive and Cognitive-Behavior Therapy, 25, 17-34.

Montgomery, G. H., Weltz, C. R., Seltz, M., & Bovbjerg, D. H. (2002). Brief presurgery hypnosis reduces distress and pain in excisional breast biopsy patients. International Journal of Clinical and Experimental Hypnosis, 50, 17-32.

Montgomery, G. H., & Kirsch, I. (1997).  Classical conditioning and the placebo effect.  Pain, 72, 107-113.

Pennebaker, J. W. (1997). Writing about Emotional Experiences as a Therapeutic Process. Washington, D.C.: American Psychological Society.

Scheier, M. F, & Carver, C. S. (1985).  Optimism, coping, and health:Assesssment and
implications of outcome expectancies.  Health Psychology, 4, 219-247.

Shacham, N. (1983). A shorted version of the profile of mood states. Journal of Personality Assessment, 47, 305-306

Smith, T. W. & Brehm, S. (1981). Person perception and the Type A coronary-prone behavior pattern. Journal of Personality and Social Psychology 40, 1137-1149.

Smith, T. W. (1982). Irrational beliefs in the cause and treatment of emotional distress: A critical review of the rational-emotive model. Clinical Psychology Review, 2, 505-522.

Smith, T. W. (1989). Assessment in rational-emotive therapy: Empirical access to the ABCD model. In M. E. Bernard & R. DiGiuseppe (Eds.), Inside rational-emotive therapy: A critical appraisal of the theory of Albert Ellis, (pp. 135-153). San Diego: Academic Press.

Snyder, C. R., Harris, C., Anderson, J. R., Holleran, S. A., Irving, L. M., Sigmon, S. T., Yoshinobu, L., Gibb, J., Langelle, C., & Harney, P. (1991). The will and the ways: Development and validation of an individual-differences measure of hope. Journal of Personality and Social Psychology, 60, 570-585.

Snyder, C. R., Hoza, B., Pelham, W. E., Rapoff, M., Ware, L., Danovsky, M., Highberger, L., Rubinstein, H., & Stahl, K. (1997).  The development and validation of the Children’s Hope Scale. Journal of Pediatric Psychology, 22, 399-421

Sohl, S., Schnur, J. B., & Montgomery, G. H. (2009). A meta-analysis of the relationship between response expectancies and cancer-treatment-related side effects. Journal of Pain and Symptom Management. 38, 775-784.

Solomon, A., Arnow, B. A., Gotlib, I. H., & Wind, B. (2003). Individualized measurement of irrational beliefs in remitted depressives. Journal of Clinical Psychology, 4, 439-55.

Wallen, S., DiGiuseppe, R., & Dryden, W. (1992). A practitioner’s guide to rational-emotive therapy. New York: Oxford Press.