Date of Award


Publication Type

Doctoral Thesis

Degree Name




First Advisor

Alan Scoboria


Belief Flexibility, Bias Against Disconfirmatory Evidence, Delusions, Jumping-to-Conclusions, Reasoning, Schizotypy



Creative Commons License

Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.


This series of five studies examined the nature of belief flexibility for delusion-like beliefs in non-clinical samples. Findings in the literature show an overlap in psychotic symptoms between clinical and non-clinical populations. While studies about the flexibility of clinical delusions abound in the literature, there is a dearth of research pertaining to the flexibility of unusual delusion-like beliefs (i.e. beliefs that do not meet criteria for clinical significance). The aims of this dissertation were to: (1) develop and validate a new self-report measure, the Windsor Belief Flexibility Scale (WBFS), for measuring the flexibility of delusion-like beliefs in non-clinical samples, (2) to establish relationships between belief flexibility and reasoning processes already demonstrated to be related to delusional thought such as belief conviction, Jumping-to-Conclusions (JTC), and Bias Against Disconfirmatory Evidence (BADE), (3) to investigate the relationships between belief flexibility and these reasoning processes with affect, and (4) to compare the nature of belief flexibility for delusion-like beliefs with other belief domains (such as scientific, religious and political). The development of the measure is described in Chapter 2. Clark and Watson’s (1995) recommendations for scale development were followed to generate a two part measure: (1) a pool of delusion-like beliefs, which were reduced using exploratory factor analytic techniques (EFA; Study 1), and replicated using confirmatory factor analysis (CFA; Study 2) and (2) a theoretically driven item pool for rating aspects of belief flexibility for selected delusion-like beliefs in Studies 3 (EFA) and 4 (CFA). The resulting model included two second-order factors of belief flexibility and belief conviction, with the former composed of first-order factors of willingness to consider contradictory evidence and willingness to consider contradictory social feedback and the latter composed of unwillingness to doubt belief and intuitive reasoning. Convergent and divergent validity was explored by measuring relationships between the WBFS, its subscales, and related constructs (e.g., schizotypy, insight, cognitive flexibility, dichotomous thinking, and reasoning biases). In Study 5, the WBFS was used to investigate the generalization of a model of reasoning biases derived in clinical populations (So et al., 2012) to a general population sample. Results from this study replicated the distinction between conviction, belief flexibility and Jumping-to-Conclusions. The BADE was added to the model, and correlations with paranoia, worry, anxiety, and depression were explored. The BADE bias was positively associated with persecutory thoughts and anxiety and negatively associated with worry. No other significant or sizable associations were noted. Finally, using a partial invariance factor analytic method, belief flexibility and reliance on intuitive reasoning for different types of beliefs (e.g. political, religious and scientific beliefs) were examined. Implications for construct validity of the WBFS are discussed.