Psychometric evaluation of a Swedish version of the Reiss Screen for Maladaptive Behavior.

Researchers translated the Reiss Screen for Maladaptive Behavior into Swedish. They gave the 36-item checklist to the adults with intellectual disability, plus two staff who knew each adult well.

Each pair of staff rated the same person. The team then checked how well the scores hung together and how closely the two raters agreed.

The total score was reliable; Cronbach’s alpha was 0.82, above the usual “okay” cutoff of 0.80. That means the items measure one general theme—overall mental-health risk.

Item-by-item agreement between two staff was only fair. Kappa averaged 0.46, so the same person could be rated “yes” by one staff and “no” by another on many single questions.

English et al. (1995) ran an almost identical study with a Dutch translation. Both papers found the same good total-score reliability and the same so-so single-item agreement. The two studies act as direct replications even though the languages differ.

Carretti et al. (2013) tested a different tool, the Vienna Frailty Questionnaire, in the same ID population. Their inter-rater kappas were much higher (around 0.75). The contrast shows that modest agreement is not a fixed feature of ID checklists; it is specific to the Reiss Screen wording.

Fujiura (2012) argues we should let people with ID speak for themselves when possible. The current study used only proxy report, so it may miss internal states the person could reveal in a simplified interview.

If you screen adults with ID in Sweden, the Swedish RSMB can flag who needs a full mental-health evaluation. Treat the total score, not single items, as your red flag. When two staff disagree on a question, gather extra examples before you decide. Consider adding brief self-report questions when the client’s verbal skills allow it; the tool’s modest item-level reliability leaves room for richer data.