Psychology, Department of


First Advisor

William D. Spaulding

Second Advisor

Rebecca Brock

Date of this Version

Winter 12-12-2016


A Research Proposal Presented to the Faculty of The Clinical Psychology Training Program In Fulfillment of the Masters-Equivalent Research Project (MERP) Proposal Requirement.

Copyright (c) 2017 Elaina Montague


Background: Schizotypy is a construct that captures quantitative dimensions of the psychosis continuum from clinical to non-clinical expressions. The purpose of this study was to determine the factor structure and criterion validity of a newly revised self-report measure, the Schizotypal Personality Questionnaire–Brief Revised Updated (SPQ-BRU; Davidson, Hoffman, & Spaulding, 2016) for predicting later cognitive-perceptual experiences in college undergraduates.

Method: The data analytic sample was comprised of 2,474 undergraduate students (female = 71.9%) attending a university in the Midwest. First, we aimed to identify a model of best fit by comparing latent measurement models of schizotypy using confirmatory factor analysis (CFA). Second, we estimated a latent cognitive-perceptual factor from multiple measures collected at a second time point in a subsample of participants (n = 357). Using structural equation modeling (SEM), we tested the impact of latent schizotypy on participants’ self-reported cognitive-perceptual experiences at time 2.

Results: Overall, CFA findings supported a 4-factor model of schizotypy described by Callaway and colleagues (2013), (χ2 (450) = 2814, p < 0.001, CFI = 0.931, TLI = 0.942, RMSEA = 0.046, CIRMSEA = 0.044—0.048, SRMR = 0.052). The 4-factor model replicated in the subsample for aim 2, (χ2 (48) = 111.073, p < 0.001, CFI = 0.961, TLI = 0.947, RMSEA = 0.061, CIRMSEA = 0.046— 0.075, SRMR = 0.041). Consistent with our hypothesis for aim 2, the latent cognitive-perceptual model had excellent fit of the data (χ2 (1) = 0.002, p = 0.963, CFI = 1.000, TLI = 1.024, RMSEA = 0.000, CIRMSEA = 0.000—0.000, SRMR = 0.000). Lastly, the SEM model for aim 3 obtained good fit of the data, (χ2 (13) = 33.636, p = 0.0014, CFI = 0.952, TLI = 0.920, RMSEA = 0.067, CIRMSEA = 0.039—0.095, SRMR = 0.041). This final path model explained 41.4% of variance in time 2 cognitive-perceptual experiences (p < 0.001).

Conclusion: This investigation bolsters a growing body of evidence for the dimensional approach to psychometrically-defined schizotypy. In addition, this study strengthens support for the predictive power of schizotypy. Psychometric and methodology issues in the context of the dimensional approach to schizotypy will be discussed.