# Reliability and validity of the German version of the Forensic Restrictiveness Questionnaire

## TL;DR

This study validates the German version of a questionnaire measuring how mentally ill forensic patients feel about being restricted.

## Contribution

The paper provides the first validation of the German version of the Forensic Restrictiveness Questionnaire.

## Key findings

- The German version showed very good internal consistency.
- Convergent validity was confirmed with negative correlations to related scales.
- The questionnaire demonstrated discriminant validity but not full criterion validity.

## Abstract

The Forensic Restrictiveness Questionnaire (FRQ) is a self-administered questionnaire for forensic mental health inpatients about their subjective experiences of restrictiveness. The present paper describes the validation of the German version of the FRQ.

Patients were recruited from eight forensic psychiatric hospitals in Germany. Internal consistency was explored using Cronbach’s α. The German version of the EssenCES (assessing ward climate), the MQPLa (assessing quality of life) and the Mental Health Component subscale of the SF-12 were used to explore convergent validity. The Physical Health Component subscale of the SF-12 was used to explore divergent validity. Patient’s levels of leave (Lockerungen), substance use behavior and occurrence of incidents were used to describe criterion validity.

The analysis indicated very good internal consistency according to Cronbach’s α. Convergent validity could be confirmed as the FRQ-G mean score was significantly negatively correlated with the EssenCES mean score and the MQPLa mean scores. No sufficient correlation could be shown for the Mental Health Component of the SF-12. A low correlation was found with the physical component of the SF-12, indicating discriminant validity. Very few significant correlations were found to establish criterion validity.

The data indicate the FRQ-G to have good construct validity (structural, convergent, divergent) but failed to fully demonstrate criterion validity. Possible reasons include an underpowered sample size and possible measurement errors. Implications for future research are discussed.