Factor structure, reliability, and validity of the Penn State Worry Questionnaire: differences between African-American and White-American college students

https://doi.org/10.1016/j.janxdis.2004.11.001Get rights and content

Abstract

This study examined differences in the factor structure of the Penn State Worry Questionnaire (PSWQ) between African-American (n = 181) and White-American (n = 180) college students. Results from a confirmatory factor analysis indicated that the traditional single-factor solution did not provide the best fit for the data from either ethnic group. A multiple-group factor analysis indicated that underlying structure of Factor 1 was roughly equivalent between ethnic groups. Structure of Factor 2, however, differed between groups. Specifically, item 10 loaded on different factors for each group. In support of these analyses, an exploratory factor analyses (EFA) among White-American participants indicated the presence of a two-factor model while an EFA among African-Americans indicated the presence of three factors. Despite some overlap in the overall factor structure between ethnic groups, African-Americans scored significantly lower on the PSWQ than the White-American group. Furthermore, among African-Americans level of ethnic identity was negatively related to state and trait measures of anxiety, but unrelated to measures of depression and worry.

Introduction

Despite suggestions that the concepts of race and ethnicity are important in the examination of psychopathology (Bernal & Castro, 1994; Carter, Sbrocco, & Carter, 1996; Okazaki, 1997), few empirical investigations have attempted to explore the possibility that ethnic identity may moderate the expression of pathology. This investigation represents one such attempt. The purpose of this study was three-fold. First, to investigate the factor structure of the Penn State Worry Questionnaire (PSWQ—a commonly used measure of trait worry; Meyer, Miller, Metzger, & Borkovec, 1990) among African-Americans compared to White-Americans. Second, to seek evidence of convergent and discriminant validity for the PSWQ with associated measures of anxiety and a measure of depression among African-Americans. And third, to examine the relationship between ethnic identity and the expression of anxiety and depression. As noted by Carter, Miller, Sbrocco, Suchday, and Lewis (1999), such investigations are important in that they highlight basic similarities and differences between ethnic groups and have the potential to further understanding of the measurement of psychological constructs and the applicability of these constructs across varying groups of individuals. While the PSWQ has been shown to be a good measure of the concept of worry among White-Americans (e.g., Brown, Antony, & Barlow, 1992), there have been few investigations of the PSWQ among African-American adults.

Worry has been conceptualized as a cognitive component of anxiety that involves attentional vigilance and distorted information processing (Barlow, 2002, Craske, 1999). It is commonly experienced at non-clinical levels and has been theorized to serve the adaptive function of facilitating cognitive planing and problem solving, possibly by helping to control future unwanted threats from the environment (Borkovec, Robinson, Pruzinsky, & DePree, 1983; Borkovec, Shadick, & Hopkins, 1991). Paradoxically, those with elevated trait anxiety or worry evidence an attentional bias toward the threatening information, while those who are low in trait anxiety or worry are thought to divert their attention away from the threatening information when under state anxious conditions (Craske, 1999).

One of the more widely used measures of worry, developed by Meyer et al. (1990), is the PSWQ. The factor structure of the PSWQ has been examined in three investigations. In the development of the PSWQ, Meyer et al. (1990) conducted a series of studies to assess its validity. It was reported that items of the PSWQ loaded onto a single factor. Furthermore, the PSWQ was moderately correlated with measures of anxiety and, to a lesser degree, depression and exhibited high internal consistency. The phrasing of the items on the questionnaire and the high test-retest reliability coefficient (r = .93) suggests that the PSWQ measures worry as a trait rather than as a state emotional reaction.

Similar results were found by Brown et al. (1992) in their examination of the validity of the PSWQ in a sample of 436 anxiety disorder patients. The authors reported high internal consistency and that the items of the PSWQ loaded onto a single factor. In addition, it was found that the PSWQ exhibited the capability to distinguish generalized anxiety disorder (GAD) from other anxiety disorders and non-anxious participants. Brown et al. (1992) also reported that the PSWQ was significantly associated with measures of anxiety and depression.

In an examination of the factor structure of the PSWQ in older adults, however, Beck, Stanley, and Zebb (1995) reported somewhat different results. Their factor analyses of the PSWQ indicated the presence of two factors (general worry and worry absence) for both GAD and non-disordered older participants. Similar to previous studies, the PSWQ for older adults was significantly correlated with measures of anxiety and depression and exhibited strong internal consistency. Although the factors were different, the authors concluded that further use of the PSWQ among geriatric populations is supported by its ability to distinguish between disordered and non-disordered participants.

In general, then, the PSWQ is a psychometrically sound instrument that has garnered evidence that it is unidimensional among adults and children as measured with a revised version of the PSWQ (Chorpita, Tracey, Brown, Collica, & Barlow, 1997). The PSWQ also exhibits psychometric specificity in that elevated scores are more typically associated with GAD than other anxiety disorder groups and non-anxious controls (although Starcevic, 1995, was not able to discriminate between clinically depressed and GAD patients on the basis of scores on the PSWQ), and is often used as an indicator of treatment change (Stober, & Bittencourt, 1998). The data from the extant literature, however, is based on predominantly White-American samples. Consequently, one cannot assume that similarly strong psychometric evidence supporting the use of the PSWQ will be found among ethnic minority populations in general, and African-Americans in particular.

In general, there is a paucity of research on the construct of worry among African-Americans. To date, two studies have examined the expression of worry among African-American samples. Brantley et al. (1999) examined the relationship between minor stressors and generalized anxiety disorder (GAD) among low-income patients. In this report, it was found that those diagnosed with GAD reported more minor stressors than non-anxious controls. While 75% of the patient sample was comprised of low-income African-Americans, the data were not examined for racial differences. Therefore, it is impossible to determine how much of the noted association between GAD and minor stressors is attributable to African-American patients.

Scott, Enig, and Heimberg (2002) investigated differences in worry symptoms among African-American, Caucasians, and Asian-American college students. Participants in this study were administered the PSWQ, the Worry Domains Questionnaire (WDQ), and the Generalized Anxiety Disorder Questionnaire. The authors noted that all three ethnic groups reported comparable levels of worry on the PSWQ. On the WDQ, however, African-Americans scored significantly lower than either the Asian-American or Caucasian groups. There was also a non-significant trend for Caucasians only to meet DSM-IV criteria on the self-report measure of GAD. While the results are informative, the authors did not attempt to examine the factor structure of the PSWQ. It is possible that the structure of worry (as measured by the PSWQ) is different for African-Americans. More importantly, the authors did not include a measure of ethnic identity. It is quite possible that the sample may have been less ethnically identified than the general African-American population, thereby limiting the generalizability of the findings from this study. The present study extended the findings of Scott et al. (2002) by examining the factor structure and convergent and discriminant validity of the PSWQ and by the inclusion of a measure of ethnic identity.

This study first compared the underlying structure of the PSWQ for African-Americans to the structure that is commonly reported in the literature. In addition, the relationships between the PSWQ and theoretically similar and dissimilar measures were examined. While the factor analysis was exploratory in nature, based on previous evidence (Carter, Sbrocco, Lewis, & Freedman, 2001) it was predicted that higher levels of ethnic identity would be associated with lower levels of anxiety and depression among African-Americans.

Section snippets

Participants

Participants were 181 African-American and 180 White-American students enrolled in undergraduate courses at American University (see Table 1 for demographic information). There were no inclusion or exclusion criteria used for participation.

Penn State Worry Questionnaire (PSWQ)

The PSWQ (Meyer et al., 1990) is a 16-item questionnaire that assesses the general trait of worry. Each of the items is rated from 1 (not at all typical of me) to 5 (very typical of me). Total scores are computed by averaging participants’ scores across

Demographic information

Demographic information for both ethnic groups is presented in Table 1. As indicated, both samples were predominantly female and scored within the normal range for a college sample on all measurement devices. To determine if the groups were comparable, a between groups MANOVA was conducted on all continuous variables (an ANOVA on age), and chi-squares were conducted on dichotomous variables. Results indicated no significant differences between groups in age, education level, STAI-S, STAI-T, or

Discussion

The data from this study suggest that the concept of worry, as measured by the Penn State Worry Questionnaire, is roughly equivalent between African-American and White-American college students. Specifically, results of a CFA indicated that the data from neither group fit the single factor solution. Follow-up EFA indicated that the underlying structure of the PSWQ for White-Americans contained two factors. The first factor (general worry) can likely be considered similar to the single factor

Acknowledgment

We would like to acknowledge the contribution of Dr. Scott Parker as statistical consultant on this project.

References (36)

  • J. Stober et al.

    Weekly assessment of worry: An adaptation of the Penn State Worry Questionnaire for monitoring changes during treatment

    Behaviour Research and Therapy

    (1998)
  • American Psychological Association

    Ethical principles of psychologists and code of conduct

    American Psychologist

    (2002)
  • J.L. Arbuckle

    AMOS user's guide

    (1997)
  • I. Atwell et al.

    Diagnosing personality disorders in Africans (Blacks) using the Azibo Nosology: Two case studies

    The Journal of Black Psychology

    (1991)
  • D.H. Barlow

    Anxiety and its disorders: The nature and treatment of anxiety and panic

    (2002)
  • A.T. Beck et al.

    Beck Depression Inventory manual

    (1987)
  • J.G. Beck et al.

    Psychometric properties of the Penn State Worry Questionnaire in older adults

    Journal of Clinical Geropsychology

    (1995)
  • P.M. Bentler

    Comparative fit indexes in structural models

    Psychological Bulletin

    (1990)
  • Cited by (47)

    • Worry, anger rumination, ego-strength and adjustment to heart disease

      2023, Annales Medico-Psychologiques
      Citation Excerpt :

      The PSWQ has good internal consistencies with alphas from .88 to .95 (Besharat, 2007; Meyer et al., 1990) and test-retest reliability from .74 to .93 [4,36]. Several studies have confirmed factor structure, convergent and discriminant validity of the PSWQ [4,17,36]. Cronbach's alpha coefficient was .83 in the present study.

    • Cognitive functions in young adults with generalized anxiety disorder

      2019, European Psychiatry
      Citation Excerpt :

      The participants are asked to indicate how much the statement generally applies to themselves by rating the items on a Likert-type scale ranging between 1 (not at all typical of me) and 5 (very typical of me). The PSWQ demonstrated good reliability in clinical samples (Cronbach’s α = .86 [31]) and in college students (α = .80 [32];). In the present study the PSWQ demonstrated excellent reliability (α = .98).

    • A review of sociocultural factors that may underlie differences in African American and European American anxiety

      2017, Journal of Anxiety Disorders
      Citation Excerpt :

      For EAs, ethnic identity and trait anxiety were not significantly associated. A similar study was conducted assessing the relation between the Penn State Worry Questionnaire (PSWQ; Meyer, Miller, Metzger, & Borkovec, 1990) and the STAI-Trait subscale (Carter et al., 2005). As in the previous study, the STAI-Trait was negatively associated with ethnic identity in AAs and unrelated to ethnic identity in EAs.

    View all citing articles on Scopus
    1

    The opinions and assertions contained herein are the private ones of the authors and are not to be construed as official or reflecting the views of the Department of Defense or the Uniformed Services University of the Health Sciences.

    View full text