Dimensionality analysis of the German version of the Screen for Cognitive Impairment in Psychiatry (SCIP-G)

Abstract

Background

Psychiatric disorders, especially schizophrenia, are characterised by cognitive impairment. The rapid detection of cognitive dysfunction - also in the course of the disease - is of great importance. The Screen for Cognitive Impairment in Psychiatry (SCIP) was developed to provide screening of psychiatric patients in clinical practice and is available in several languages. Prior psychometric investigations into the dimensionality of the SCIP have produced two different models: a one-factor model assumes that the five subscales of the SCIP load together, whereas an alternative model suggests that the subscales load on two factors, namely verbal memory and processing speed. We carried out a confirmatory factor analysis of the German version of the SCIP (SCIP-G).

Methods

323 patients with psychotic, bipolar affective, and depressive disorders were studied.

Results

The one-factor approach did not yield an acceptable model fit (chi-squared test: χ² = 109.5, df = 5, p < 0.001, χ²/df = 21.9). A two-factor solution, with the subtests Verbal Learning Test-Immediate Recall, Delayed Recall Test of the VLT, and Working Memory Test loading on the first factor, whereas the subtests Verbal Fluency Test and Psychomotor Speed Test loading on the second factor, obtained a good model fit (χ² = 6.7, df = 3, p = 0.08, χ²/df = 2.2).

Conclusions

These data show that a good model fit can be achieved with a two-factor solution for the SCIP. This study is the first to conduct a confirmatory factor analysis using the German SCIP version and to test its dimensional structure using a hypothesis-testing approach.