Date of Award


Publication Type

Doctoral Thesis

Degree Name





Life Orientation Test, optimism, personality psychology, pessimism, positive psychology


Kathryn D Lafreniere




Optimism has been a favourite topic of research in positive psychology. Optimism, defined as a generalized positive expectancy for the future, is generally regarded as a positive trait. However, despite positive findings for optimism, some researchers have suggested that optimism is not beneficial in all contexts. Alternatives to optimism have been proposed, including flexible optimism (Seligman, 1991; Forgeard & Seligman, 2012) and cautious optimism (Wallston, 1994). While such criticism of optimism lacks substantial empirical support, there are a few studies that appear to support these contentions. Previous research suggests that optimism is associated with maladaptive persistence in gambling (Gibson & Sanbonmatsu, 2004) and poorer health (de Ridder, Schreurs, & Bensing, 2000). Furthermore, research on defensive pessimism and unrealistic optimism supports the notion of a “dark side” of optimism. A new construct is proposed to reconcile these divergent findings: expectancy flexibility. Expectancy flexibility is defined as the ability to change one’s expectations of the future in response to contextual cues. It was hypothesized that expectancy flexibility would moderate or mediate the associations between optimism and various outcomes. Four studies were conducted to validate the Expectancy Flexibility Scale (EFS), an instrument developed to measure expectancy flexibility. The first two studies were used to develop a scale with good internal consistency reliability, a low correlation with optimism (to provide discriminant validity), and a moderate correlation with theoretically related constructs (to provide convergent validity). The purpose of the third study was to test whether shifts in expectations actually occur in response to negative feedback, and whether these shifts were predicted by scores on the EFS. The fourth study tested whether the EFS was associated with constructs believed to be outcomes, including preventive health behaviours, academic success, and problem gambling. In all four studies, participants were undergraduate students who were recruited through a participant pool at a Canadian university. The EFS and several other self-report questionnaires were completed by participants via an online platform. The findings of Study 1 and Study 2 supported the reliability and validity of the EFS. Internal consistency reliability was in the acceptable range (α > 0.70). Supporting the scale’s convergent validity, expectancy flexibility was associated with related measures like defensive pessimism and cognitive flexibility. Weak and non-significant correlations were found between expectancy flexibility and optimism, locus of control, and coping flexibility, supporting the scale’s discriminant validity. The findings of Study 3 partially supported the hypothesis that expectancy flexibility is associated with shifts in expectations. In the gambling scenario, losses were generally associated with reduced expectations, while gains were associated with no change or slight increases in gambling expectations. This pattern of findings was not evident in the academic scenario, where disappointing exam results did not produce a negative shift in expectations. In Study 4, expectancy flexibility was positively associated with academic approach coping, social health, general academic skills, and confidence; it was negatively related to substance use and problem gambling. Analysis of the qualitative questions generally supported the hypothesis that expectancy flexibility is associated with shifts in expectations. However, the moderational and mediational models were not supported. Overall, the results provide support for the validity of the flexible optimism construct.