Psychometric evaluation of the Nurses Observation Scale for Inpatient Evaluation in The Netherlands.

The interrater reliability, temporal stability and factorial, convergent, discriminant and predictive validity of the Nurses Observation Scale for Inpatient Evaluation (NOSIE-30) were investigated in a heterogeneous group of psychiatric inpatients in the Netherlands (n = 179). Data in support of the scale's dimensional structure (recoverability of Anglo-American subscale structure in this Dutch sample and distinctiveness of subscales), discriminatory power and convergent validity are presented. Interrater reliability was satisfactory at global scale level. However, 3 subscales (irritability, psychosis and depression) were found to lack interrater reliability. Although temporal stability coefficients were high, large score changes are presupposed to show that pre- versus posttherapy differences are attributable to real change rather than error. NOSIE-30 had limited predictive value.