خوش بینی، بدبینی و "روح مبارزه": یک رویکرد جدید به ارزیابی امید و انطباق

Abstract The present study describes the construction and validation of a new scale for measuring optimism, pessimism and “fighting spirit”, entitled the Positive And Negative Expectancy Questionnaire (PANEQ). Principal axis factor analysis of an initial pool of items generated by a scenario study yielded three factors, labelled negative affect/pessimism, fighting spirit and positive affect/optimism, all of which showed high internal (coefficient alpha) and re-test reliability. The original factor analysis was subsequently replicated in a separate sample using confirmatory factor analysis, and a structural equation model for the relationship between PANEQ and three measures of health and well-being provided evidence for discriminate validity of the three scales.

Introduction The Life Orientation Test (LOT; Scheier & Carver, 1985) is one of the most widely used measures of optimism and pessimism. These authors' definition of dispositional optimism stems from their control theory of self regulation and adjustment, where optimists have a favourable outlook on life, believe that good rather than bad things will happen to them, and consequently are more likely than pessimists to face adversity with continued effort (Carver and Scheier, 1982 and Scheier and Carver, 1985). Dispositional optimism (as measured by the LOT) has generated a considerable body of research, which has linked optimism to better physical and psychological well being (Carver et al., 1993, Scheier and Carver, 1985, Scheier and Carver, 1992, Scheier et al., 1986 and Scheier et al. 1989). For example, in a study on health change among cardiac patients, dispositional optimism predicted success in achieving lowered levels of saturated fat, body fat and global coronary risk, and was also positively related to increased aerobic capacity at the end of the program (Shepperd, Marato & Pbert, 1996). In an earlier study by Carver and Gaines (1987), dispositional optimism was found to be negatively related to post-partum depression in a group of 75 pregnant women. However, serious concerns have been raised about the validity of the LOT, including the view that the optimism–pessimism model used by Scheier and Carver masks the broader and more established construct of neuroticism. The optimism–neuroticism hypothesis ( Chang, 1998a) argues that previous findings linking optimism to both better physical and psychological health results from its overlap with neuroticism. For example, Smith, Pope, Rhodewalt and Poulton (1989) found that optimism was related to lower concurrent and future symptom reporting and to higher problem-solving coping across two separate studies, but when the effects of neuroticism (as measured by trait anxiety) were controlled for, these relationships disappeared. By contrast, when the effects of optimism were partialled out the relationship between neuroticism and those measures remained significant. Similarly, in a study by Boland and Cappeliez (1997) optimism (as measured by the LOT) failed to predict a range of stress-related outcomes in a prospective design, but both neuroticism and the baseline values of these outcome variables did do so. A recent meta-analysis assessing the benefits of optimism in relation to various indices of coping, symptom reporting and negative affect showed that the most reliable association was between optimism and negative affect ( Andersson, 1996). These studies indicate that the reported health benefits of optimism can be explained by its covariation with neuroticism. However, it has been argued that such findings may be confounded by using the LOT as a unidimensional construct of optimism–pessimism (Chang, 1998a). A growing number of studies have challenged Scheier and Carver's (1985) unidimensional model of optimism and pessimism, and have shown that when used separately, pessimism is usually related to negative affect or neuroticism and optimism to positive affect or extraversion (for example, Chang et al., 1994, Chang et al., 1997, Marshall et al., 1992, Mook et al., 1992 and Mroczek et al., 1993). This indicates that optimism, extraversion and positive affect on the one hand, and pessimism, neuroticism and negative affect on the other, form part of the broader personality constructs of E and N (extraversion and neuroticism), respectively. However, differentiation into separate measures of optimism and pessimism would seriously compromise the LOT — the distinct measures of optimism and pessimism in the scale contain only four items each, which would not provide psychometrically adequate samples of behaviour (Kline, 1993). Chang et al., 1997 and Chou and Bentler, 1995 have recently presented evidence for the independent predictive power of optimism and pessimism, based on a new bi-dimensional questionnaire comprising a 6-item optimism scale and a 9-item pessimism scale. These scales also comprise relatively few items, but Chang et al. (1997) found that they were nonetheless differentially related to criterion variables: optimism was related to life satisfaction but not depression, and the reverse was true for pessimism. These findings remained when the effects of positive and negative affect were controlled in a series of hierarchical regressions, and even though the effect sizes were comparatively small, the results support the view that optimism and pessimism are independent predictors of psychological well-being. Recently, Chang (1998a) set out to test more directly the optimism–neuroticism hypothesis using the same proxy for neuroticism (trait anxiety) that Smith et al. (1989) had used. His findings failed to support the hypothesis, since optimism still correlated with life satisfaction, constructive problem solving and problem and emotional engagement coping strategies, even after controlling for any confounding influences of trait anxiety. Pessimism correlated inversely with life satisfaction, constructive problem solving, and positively with problem and emotional disengagement coping and physical symptom reports, although the magnitude of these correlations was modest, with only one of them exceeding 0.3. Again, when Chang (1998a) constructed a unidimensional measure of optimism–pessimism and repeated the analysis, almost all of these relationships disappeared after controlling for the contribution of trait anxiety. In summary, these findings support the view that optimism–pessimism is more appropriately defined as two separate but related constructs. These in turn are systematically related to indices of psychological and physical health and to coping styles, and the relationships exist independently of the effects of neuroticism. However, the unique variance explained by optimism and pessimism separately is relatively small, and the problem is confounded by the lack of adequate measures — for example, distinguishing optimism and pessimism using the LOT yields only four items in each separate scale. There is clearly considerable scope for refining the measurement of optimism and pessimism, and the present study was aimed primarily at generating psychometrically improved measures that would distinguish unambiguously between the two constructs. In addition, the study set out to explore more closely their relationship to positive and negative affect.