On some psychometric properties of the questions about behavioral function (QABF) scale

This study examined certain psychometric properties of the questions about behavioral function (QABF) scale. The QABF was completed on 118 problem behaviours presented by 40 young people with severe intellectual disabilities, and measures of inter-rater reliability, internal consistency, and construct validity were calculated overall and for subgroups of behaviours defined by topography and reported frequency. Levels of inter-rater reliability for both individual items and subscales were generally modest, but inter-rater reliability of rankings of subscale scores exceeded that reported for other comparable assessments. Inter-rater agreement was higher for higher rate behaviours and lower for maladaptive (versus disruptive and destructive) behaviours. Internal consistency was high for all QABF subscales, but also for the scale as a whole. Factor analysis yielded five factors clearly corresponding to the five QABF subscales and a sixth relating to a single item concerning the repetitive nature of the behaviour. Methodological issues which may have reduced levels of reliability are discussed. We conclude that in terms of psychometric properties, the QABF presents some specific advances on previous brief scales assessing behavioural function, but also that it continues to be appropriate to use other observational or experimental methods together with such scales when conducting functional assessments.