Citation

Aish AM, Ramberg IL, Wasserman D. Arch. Suicide Res. 2002; 6(4): 309-323.

Affiliation

National Swedish Centre for Suicide Research and Prevention; Department of Public Health Sciences, Karolinska Institute; WHO Collaborating Centre on Suicide Research and Prevention

Copyright

(Copyright © 2002, International Academy of Suicide Research, Publisher Informa - Taylor and Francis Group)

DOI

10.1080/13811110214526

PMID

unavailable

Abstract

Measurement, because of its role in the scientific process, is one of the principal concerns in human sciences. This study investigates three measurement instruments for attitudes of mental health care staff toward suicidal patients in terms of their dimensionality and the validity and reliability of the individual indicators: the instruments designed by Suokas and Lonnqvist (1989), Samuelsson, Asberg and Gustavsson (1997) and by Ramberg and Wasserman. The empirical analyses are done with LISREL, a structural equation modeling approach that is particularly useful to study measurement instruments for abstract concepts. This approach provides an explicit test of the dimensionality of a construct and of the validity of each the indicators. It also gives explicit information on measurement error in the observed variables. The results show that each measurement instrument contains several components as hypothesised, but only some of them, tested as confirmatory factor analysis models, produced a good fit. Moreover the validity and the reliability of their indicators varied considerably. Empathy was the most consistently measured component in the three measurement instruments.