Monday, 31 March 2014

Reporting instrument development and testing

Roger Watson, Editor-in-Chief

David L Streiner
We commissioned two leading psychometricians, David L Streiner and Jan Kottner, to write a special paper for us; one that would be helpful to authors in submitting manuscripts to JAN on instrument development and testing. I am pleased to present their superb paper: 'Recommendations for reporting the results of studies of instrument and scale development and testing'.

Jan Kottner
The paper is a readable, authoritative and contemporary 'take' on the subject. The sources cited are the key ones and the ideas expressed should resonate beyond the pages of JAN.

Old notions of reliability and validity are challenged and, of supreme importance, the need to develop a new scale where an suitable alternative exists, is challenged. The recommendations cover every aspect of an instrument development paper - as currently required by JAN - from the title to the conclusions.

In particular, authors should be cautious in the claims they make about their instruments: no longer should we refer to an instrument as being 'reliable'; instead we should refer to the ways reliability was tested. Similarly, we should no longer refer to instruments as being valid or that some particular aspect of validity has been established; rather, we should realise that all tests of validity are providing insight into the construct validity which is, in fact, unobtainable. We can only claim any level of validity for an instrument for the samples or populations with which it has been tested and  only then if we have really been able to establish that the instrument really does measure what is was designed to measure.

I urge all authors to read these recommendation, especially if you intend to submit a manuscript on instrument development.





Reference

Streiner DL, Kottner J (2014) Recommendations for reporting the results of studies of instrument and scale development and testing Journal of Advanced Nursing



No comments:

Post a Comment