Validity of the Test of Visual Perceptual Skills–Fourth Edition in People With Schizophrenia

Participants

People in the stable phase of schizophrenia were recruited from community rehabilitation facilities belonging to a psychiatric center in northern Taiwan from January 2021 to June 2022. The participants were selected according to the following inclusion criteria: (1) diagnosis of schizophrenia by psychiatrists according to the Diagnostic and Statistical Manual of Mental Disorders (5th ed.; American Psychiatric Association, 2013); (2) age older than 20 yr; and (3) an agreement to participate in the research and provide informed consent. The exclusion criteria were as follows: (1) diagnosis of severe brain injury or intellectual developmental disorder and (2) visual deficits. This study was approved by the institutional review board of the hospital. All participants provided informed consent.

Procedures

Before executing the assessments, the rater received training to become familiar with the measures. There were three criteria for determining whether the rater was familiar with the measures. First, the rater had 3 yr of clinical experience as a research assistant conducting assessments in the psychiatric center. Second, at least 3 hr of training was provided by a senior occupational therapist with 20 yr of clinical experience in the rehabilitation of people with schizophrenia. During training, the rater became familiar with conducting the assessments and with the instructions, test items, and procedures of administration. Third, the rater assessed two healthy adults under the supervision of the senior occupational therapist to ensure familiarity with the measures.

People with schizophrenia who met the inclusion and exclusion criteria of this study were referred by occupational therapists who worked in the psychiatric center. Participants completed the TVPS–4, Activities of Daily Living Rating Scale–III (ADLRS–III; Chu, 1997), Mini-Mental State Examination (MMSE; Folstein et al., 1975), Behavioural Assessment of the Dysexecutive Syndrome (BADS; Wilson et al., 1996), and Clinical Global Impressions–Severity Scale (CGI–S; Guy, 1976). The assessments were administered in a quiet place to prevent any interference that may have influenced the participants’ administration. The demographic data of the participants were collected from their medical records.

Measures

The TVPS–4 assesses visual perception using seven subscales. On the Visual Discrimination subscale, participants are asked to find one of five images that match the target image. On the Visual Memory subscale, participants are asked to remember a target image and then find the image on the following page. On the Spatial Relationships subscale, participants are asked to identify an image that differs from other images on the same page. On the Form Constancy subscale, participants are asked to find one of five images that are similar to the target image, regardless of the change in the size or position of the target image. On the Sequential Memory subscale, participants are asked to remember images and their order and then identify the same images in the same order on the following page. On the Visual Figure–Ground subscale, participants are asked to find a target image embedded in one of the fields of four complex designs. On the Visual Closure subscale, participants are asked to choose one of four incomplete images that matches the target image with a complete figure. Every subscale begins with two unscored practice items, followed by 18 test items. Each test item was rated 0 (incorrect) or 1 (right). Participants obtain five out of the seven test items inaccurately on a subscale and, consequently, discontinue the subscale. Figure 1 shows the practice items of the seven subscales. The sum score of the test items for each subscale ranged from 0 to 18 to display the particular subscale function of visual perception. The score on the overall scale of the seven subscales ranged from 0 to 126 to represent global visual perception. A higher score demonstrates a superior visual– perception function (Martin, 2017). Reliability is a prerequisite for the validation of psychometric properties. The internal consistency reliability of the TVPS–4 was estimated in this study. For the overall scale, Cronbach’s α was .95. For the subscales (i.e., Visual Discrimination, Visual Memory, Spatial Relationships, Form Constancy, Sequential Memory, Visual Figure–Ground, and Visual Closures), Cronbach’s αs were .67, .60, .87, .82, .70, .82, and .79, respectively.

OPEN IN VIEWER

Figure 1.

Practice items for each of the seven subscales of the Test of Visual Perceptual Skills–Fourth Edition.

The ADLRS–III is a self-reported questionnaire to assess comprehensive ADLs (including basic and instrumental ADLs) for people with mental illness (Chiu, Lee, Lai, et al., 2015). The ADLRS–III contains 98 items related to ADL function that are administered in real life. The ADLRS–III scores range from 0 to 100, with higher scores indicating better ADL function. The construct validity of the ADLRS–III has previously been examined in people with schizophrenia (Chiu, Lee, Lai, et al., 2015).

The MMSE assesses general cognitive functions (e.g., attention and memory). Its scores range from 0 to 30, with higher scores indicating superior general cognitive function. The MMSE has sufficient test–retest reliability in people with schizophrenia (de Leon et al., 1993).

The BADS assesses executive functions with six subtests: Rule Shift Cards, Action Program, Key Search, Temporal Judgement, Zoo Map, and Modified Six Elements. The total BADS score ranges from 0 to 24. The BADS has been examined discriminative validity, convergent validity, and internal consistency in people with schizophrenia (Vargas et al., 2009).

The CGI–S assesses the severity of psychiatric illness ratings by clinicians according to the observation of symptoms, function, and behaviors in the past 7 days. The results are ranked on a 7-point scale on which 1 = normal, not at all ill; 2 = borderline, mentally ill; 3 = mildly ill; 4 = moderately ill; 5 = markedly ill; 6 = severely ill; and 7 = the most extremely ill (Busner & Targum, 2007). The CGI–S showed satisfactory convergent validity in people with schizophrenia (Turkoz et al., 2013).

Data Analysis

We investigated ecological validity by calculating the extent of correlation (Pearson’s r) between the TVPS–4 scores (i.e., overall scale and seven subscales) and ADLRS–III scores. The standard of ecological validity was as follows: r < .30, low correlation (poor ecological validity); .30 ≤ r < .60, moderate correlation (adequate ecological validity); and r ≥ .60, strong correlation (good ecological validity; Salter et al., 2005). We hypothesized that TVPS–4 scores would be moderately correlated with ADLRS-III scores.

We evaluated convergent validity using Pearson’s r to estimate the correlation between the TVPS–4 scores and the two cognitive measures (i.e., the MMSE and BADS). The criteria of convergent validity were as follows: .40 ≤ r < .70, moderate correlation; and r ≥ .70, strong correlation (Hobart et al., 2004). We hypothesized that TVPS–4 scores would be moderately correlated with scores on the two cognitive measures.

The participants were divided into three groups based on different severities of psychiatric illnesses (CGI–S scores = 2, 3, and 4, respectively). We examined discriminative validity using an analysis of variance (ANOVA; α = .05) to compare the three groups and determine the existence of statistically significant differences among the TVPS–4 scores. We hypothesized that TVPS–4 scores could be used to distinguish different severities of psychiatric illnesses. The floor and ceiling effects were calculated as the percentage of participants who graded the minimum and maximum scores, respectively. A percentage displaying ≥15% demonstrated a noticeable floor or ceiling effect (McHorney & Tarlov, 1995).