The dimensional structure of short forms of the Wisconsin Schizotypy Scales

https://doi.org/10.1016/j.schres.2015.05.016Get rights and content

Abstract

The Wisconsin Schizotypy Scales (WSS) are widely used for assessing schizotypy. Confirmatory factor analysis (CFA) indicates that a two-factor structure, positive and negative schizotypy, underlies these scales. Recently developed 15-item short forms of the WSS demonstrated good reliability and validity. This study examined the factor structure underlying the short-form WSS. Consistent with the original scales, CFA on three large samples (n = 6137, 2171, and 2292, respectively) indicated that a two-factor model with positive and negative dimensions provided better fit than a generic schizotypy model for the short-form WSS. The short-form dimensions correlated highly with the original scale dimensions and displayed good stability across 10 weeks. Preliminary construct validity was demonstrated through associations with interview and questionnaire measures of psychopathology, functioning, and personality comparable to those found with the original WSS. This is the first study examining the dimensional structure of the short WSS and the validity of these dimensions. The findings support the multidimensional nature of schizotypy and the appropriateness of dimensions derived from the short-form WSS.

Introduction

Current research conceptualizes schizotypy as a multidimensional phenotype that encompasses clinical and subclinical manifestations of the schizophrenia spectrum (Lenzenweger, 2010, Kwapil and Barrantes-Vidal, 2014). Ample evidence supports significant overlap between schizotypy and schizophrenia across behavioral and neurobiological domains, suggesting that the identification of schizotypic individuals should facilitate the detection of etiological risk and protective factors for schizophrenia-spectrum disorders (for a review see Ettinger et al., 2014). It also allows for the examination of etiological factors relatively untainted by confounds accompanying full-blown schizophrenia, such as hospitalization, medication, and stigma. Since their development by the Chapmans and colleagues, the Wisconsin Schizotypy Scales (WSS, also known as the Chapman Scales of Psychosis Proneness)—including the Perceptual Aberration (Chapman et al., 1978), Magical Ideation (Eckblad and Chapman, 1983), Physical Anhedonia (Chapman et al., 1976), and Revised Social Anhedonia (Eckblad et al., 1982) Scales—have been widely employed in the study of schizotypy. Cross-sectional and longitudinal investigations provided evidence for the reliability and validity of the WSS (e.g., Gooding et al., 2005, Kwapil et al., 2008, Kwapil et al., 2013).

Schizotypy, and by extension schizophrenia, are conceptualized as multidimensional constructs (Raine et al., 1994, Vollema and van den Bosch, 1995, Stefanis et al., 2004), with positive and negative symptom dimensions the most consistently replicated factors. Using confirmatory factor analysis (CFA) with 6137 young adults, Kwapil et al. (2008) found evidence for a two-factor structure with positive and negative factors underlying the original WSS. In addition, they reported that, as hypothesized, the schizotypy dimensions were differentially associated with symptoms and impairment. Positive schizotypy was associated with psychotic-like experiences, substance abuse, mood disorders, and mental health treatment; negative schizotypy was associated with negative and schizoid symptoms and decreased likelihood of intimate relationships. Both dimensions were related to schizotypal and paranoid symptoms and poorer functioning. Kwapil et al. (2013) reported that both dimensions predicted schizophrenia-spectrum disorders using data from the 10-year follow-up study conducted by Chapman et al. (1994).

Despite the demonstrated validity of the WSS, the combined length of the scales (166 items) can be problematic; therefore, Winterstein et al. (2011) created 15-item short forms for each of the four WSS. They chose items based upon content analysis and psychometric properties using classical test theory, item response theory, and differential item functioning. Winterstein et al. (2011) reported good internal consistency for the short-form scales as well as preliminary evidence for validity. Gross et al. (2012) investigated the reliability and validity of the short-form WSS using interview ratings of psychotic-like and schizophrenia-spectrum symptoms and questionnaire measures of personality and social impairment. Despite the drastic reduction in items, the short scales demonstrated good reliability, correlated highly with the original scales, and exhibited hypothesized associations with measures of psychopathology, personality, and impairment. Fonseca-Pedrero et al. (2014) reported that the brief versions of the Perceptual Aberration and Magical Ideation Scales had good psychometric properties and loaded on a single underlying factor. However, they did not examine the properties of the anhedonia scales.

The present study extends the work of Winterstein et al. (2011), Fonseca-Pedrero et al. (2014), and Gross et al. (2012) by examining the factor structure underlying the short WSS in three large samples. This is the first study examining the dimensional structure of the short WSS and the validity of these dimensions. It was hypothesized that the two-factor structure reported by Kwapil et al. (2008) for the original scales will be replicated in the shortened scales. Assuming that the structure is supported, we hypothesized that the short-form dimensions would exhibit good temporal stability and comparable associations with measures of schizotypic symptoms, impairment, personality, and social functioning as reported in Kwapil et al. (2008).

Section snippets

Participants

WSS data were obtained from three large, independent samples of undergraduates. The first two samples completed the original WSS. We then derived the short WSS scores and dimensions from the original scales, allowing us to compare performance, factor structure, and the correlates of these factors between the original and short scales. The first sample included 6137 students (76% female) with a mean age of 19.4 (SD = 3.7). This sample was used by Kwapil et al. (2008) to examine the factor

Results

Descriptive statistics for the short WSS in the first two samples are reported in Gross et al. (2012) and descriptives for the third sample are reported in Table 1. These values were comparable across the three samples, and the short scales demonstrated good reliability and high correlations with their original scale counterparts.

Discussion

The construct of schizotypy affords a unique opportunity to investigate neurodevelopmental and psychosocial factors underlying schizophrenia. Further, it provides a rich model for conceptualizing psychosis as a dynamic continuum ranging from subclinical manifestations, to schizophrenia-spectrum personality disorders, to schizophrenia. Phenomelogical, genetic, cognitive, and neurobiological evidence for putative overlap between schizotypy and schizophrenia supports the dimensional

Role of funding source

Neus Barrantes-Vidal and Thomas R. Kwapil are supported by the Spanish Ministerio de Ciencia e Innovación (PSI2008-04178), Fundació La Marató de TV3 (091110), and the Generalitat de Catalunya (Suport als Grups de Recerca - 2009SGR672).

Contributors

Georgina M. Gross, MS, contributed to the design and analyses, and was lead author of the manuscript. Paul J. Silvia, PhD contributed to the analyses and writing of the manuscript. Neus Barrantes-Vidal, PhD, contributed to the writing of the manuscript and study design. Thomas R. Kwapil, PhD, designed the study and contributed to the data analyses and writing of the manuscript.

Conflict of interest

None of the authors had a conflict of interest.

Acknowledgments

We thank Martha Diaz and Leigh Dickerson for assistance with data collection, and Charlotte Chun and Sarah Sperry for the comments on drafts of this manuscript.

References (42)

  • N.C. Andreasen et al.

    Positive and negative symptoms in schizophrenia

    Arch. Gen. Psychiatry

    (1990)
  • P.M. Bentler et al.

    Practical issues in structural modeling

    Sociol. Methods Res.

    (1987)
  • L.H. Brown et al.

    The relationship of social anxiety and social anhedonia to psychometrically identified schizotypy

    J. Soc. Clin. Psychol.

    (2008)
  • L.J. Chapman et al.

    Scales for rating psychotic and psychotic-like experiences as continua

    Schizophr. Bull.

    (1980)
  • Chapman, L.J., Chapman, J.P., 1983. Infrequency scale for personality measures. Unpublished scale available from T.R....
  • L.J. Chapman et al.

    Scales for physical and social anhedonia

    J. Abnorm. Psychol.

    (1976)
  • L.J. Chapman et al.

    Body image aberration in schizophrenia

    J. Abnorm. Psychol.

    (1978)
  • G. Claridge et al.

    The factor structure of “schizotypal” traits: a large replication study

    Br. J. Clin. Psychol.

    (1996)
  • J. Cohen

    A power primer

    Psychol. Bull.

    (1992)
  • P.T. Costa et al.

    NEO-PI-R Professional Manual

    (1992)
  • M.L. Eckblad et al.

    Magical ideation as an indicator of schizotypy

    J. Consult. Clin. Psychol.

    (1983)
  • Cited by (22)

    • The factor structure of the short form of the Wisconsin schizotypy scales

      2018, Psychiatry Research
      Citation Excerpt :

      Poorer reliability in the short-form WSS compared to the original WSS was reported in two large samples of undergraduate students (Gross et al., 2012). The PhAS, in particular, had Cronbach's alpha = 0.65 in a sample of 2292 undergraduate students (Gross et al., 2015), although reliability resulted as acceptable and above the conventional threshold of 0.70 for all subscales of the short-form WSS (Gross et al., 2012) when tested with an algorithm that is appropriate for binary data, as is the case with this study. Results of the CFA are difficult to compare to past investigation, since Gross et al. (2015) used a different estimator (MLR: maximum likelihood estimation with robust standard errors) than the estimator used in this study (DWLS), and did not report goodness-of-fit indicators.

    • Schizotypy is associated with reduced mnemonic precision in visual working memory

      2018, Schizophrenia Research
      Citation Excerpt :

      These items are chosen from the original full scales (see details in Winterstein et al., 2011a, 2011b) and have been validated with large sample sizes (Gross et al., 2012). Moreover, given that a model with two factors (one positive schizotypy factor and one negative schizotypy factor) could generally provide good fits to the data (Gross et al., 2015), the four sub-scale scores were further combined into positive and negative schizotypy scores based on the scoring recommendation in Gross et al. (2015), in which scoring coefficients were derived from a practical norm with 6137 participants. In addition, a total score was also calculated by summing all item scores together (ranging from 0 to 60).

    • Interaction between FKBP5 gene and childhood trauma on psychosis, depression and anxiety symptoms in a non-clinical sample

      2017, Psychoneuroendocrinology
      Citation Excerpt :

      Exploratory and confirmatory factor analyses of the four scales reliably produce two factors, positive and negative schizotypy, accounting for 80% of the variance (Kwapil et al., 2008). Although the technical school participants completed the short version of the scales (Winterstein et al., 2011), which raw scores are not comparable to those obtained with the original scales, the factor structure underlying both the short and the original scales is comparable (Gross et al., 2015). Participants were assigned positive and negative schizotypy factor scores based upon factor loadings derived from a sample of 6137 college students (Kwapil et al., 2008).

    • Brain bioenergetics and redox state measured by <sup>31</sup>P magnetic resonance spectroscopy in unaffected siblings of patients with psychotic disorders

      2017, Schizophrenia Research
      Citation Excerpt :

      The Structured Clinical Interview for DSM-IV-TR (First et al., 1995) was used for diagnosis ascertainment of all subjects. The following rating scales were also administered as appropriate to patients, siblings and/or controls: Wisconsin Schizotypy Scales-short form (Gross et al., 2015), Symptom Checklist-90-Revised (SCL-90-R) (Derogatis et al., 1973), Positive and Negative Syndrome Scale (PANSS) (Kay et al., 1987), Beck Depression Inventory-II (BDI-II) (Beck et al., 1996), State-Trait Anxiety Inventory (STAI) (Spielberger et al., 1983), Edinburgh Handedness inventory (Oldfield, 1971), and MATRICS Consensus Cognitive Battery (MCCB) (Marder and Fenton, 2004). 31 P MRS was performed on a 4T whole-body MR scanner interfaced with Varian INOVA console.

    • Typology of schizotypy in non-clinical young adults: Psychopathological and personality disorder traits correlates

      2016, Psychiatry Research
      Citation Excerpt :

      Symptoms of schizophrenia and schizotypy could be grouped in three dimensions: positive (unusual experiences, odd beliefs, and magical thinking), negative (anhedonia, apathy, social anxiety and social withdrawal) and disorganized (eccentricity, odd behavior and speech). Positive and negative dimensions were the most consistently replicated factors (Gross et al., 2015). Kocsis-Bogár et al. (2016) reported continuity between the positive and negative dimensions of schizotypy and positive and negative symptoms of schizophrenia, while Kwapil et al. (2013) reported that positive and negative schizotypy dimensions predicted schizophrenia-spectrum disorders.

    View all citing articles on Scopus
    View full text