Author:
Streiner David L.,Norman Geoffrey R.,Cairney John
Abstract
Abstract
Not all of the items that have been generated will prove to be useful. This chapter discusses the criteria of a good item, such as its reading level, lack of ambiguity, asking only a single question, freedom from jargon, being worded in a positive direction, and length. Another consideration is the distribution of responses, whether all response alternatives have been used by at least some of the respondents, and when items with very low endorsement frequencies may actually be useful. It also discusses the homogeneity of the scale and the various ways of measuring it (e.g. KR-20, split-halves, and Cronbach’s alpha). Finally, it discusses when homogeneity is and is not important, and which statistical tests should not be used when homogeneity is not required.
Publisher
Oxford University PressOxford
Reference124 articles.
1. Jackson, D.N. (1970). A sequential system for personality scale development. In Current topics in clinical and community psychology (ed. C.D. Spielberger), Vol. 2, pp. 61–96. Academic Press, New York.
2. Kornhauser, A. and Sheatsley, P.B. (1959). Questionnaire construction and interview procedure. In Research methods in social relations (Revised edn) (eds. C. Selltiz, M. Jahoda, M. Deutsch, and S.W. Cook), pp. 546–87. Holt, Rinehart and Winston, New York.
3. A comprehensive strategy for developing closed-ended survey items for use in studies of older adults.;Journal of Gerontology: Psychological and Social Sciences,2002
4. Does item homogeneity indicate internal consistency or item redundancy in psychometric scales?;Personality and Individual Differences,1991