An item response theory analysis of the Forensic Restrictiveness Questionnaire (FRQ)

Forensic psychiatric care settings are intended to be more therapeutic than penal settings. They attempt to be more homely, recovery-oriented, person-centered, and less overtly punitive. However, forensic inpatient settings are still highly secure, risk averse, and diminutive of patient autonomy. Accordingly, a body of research is investigating how patients experience their care and how these experiences are associated with treatment outcomes. The self-report Forensic Restrictiveness Questionnaire (FRQ) is a 15-item questionnaire of patients’ perceptions of the restrictions upon their autonomy. There has been interest in validating the FRQ in several countries. Despite the promising preliminary empirical support for the FRQ, its psychometric properties are not well understood. In this paper we draw on Item-Response Theory (IRT) to investigate the properties of individual FRQ items to identify candidate items for alteration, removal or retention to assist researchers validating the FRQ in new contexts. The results suggest the FRQ is more sensitive to measuring the perceptions of patients that have above average amounts of restrictiveness. Measurement error rises sharply for the approximately 5% highest scoring respondents but is low for the majority of individuals. Users are likely to respond in a dichotomised manner and not use the ‘Not Sure’ option. The response category ‘Not Applicable’ should be removed from a revised FRQ