“Reliability: A Review of Psychometric Basics and Recent Marketing Practices”, 1979-02-01 (; backlinks; similar):
The basic theories and measurement procedures for reliability and the closely related concept of generalizability are reviewed, illustrated, and evaluated for use in marketing research.
A critique is given of a subset of previous marketing research studies in which reliability estimates were used and recommendations are made for future research.
…Given the state-of-the-art of measurement in marketing, 4 recommendations are offered.
First, marketing researchers need to develop multi-item scales to measure constructs in the area. Most constructs by definition are too complex to be measured effectively with a single item, and multi-item scales are necessary for appropriate reliability and validity assessment. Despite encouraging signs of multi-item scale development found in this review, more and better multi-item scales need to be developed.
Second, the development of reliable scales presents a useful starting point for improving the quality of marketing research. If multi-item scales are developed which initially demonstrate low reliability, reliability often can be increased to acceptable levels by improving the clarity of the instructions, reducing ambiguity in the items, or by simply adding similar items to the scale.
Third, in reporting reliability coefficients, researchers should fully explain (1) relevant scale characteristics, (2) the procedure used to assess reliability and the source(s) of error which is treated, (3) appropriate references and previous reliability estimates (if any) for the scale, and (4) the interpreted meaning of the reliability coefficient. This information may help to overcome the problems of ambiguity in the area. For example, on the basis of the use of an internal consistency estimate and the researcher’s background, internal consistency estimates have been referred to as measures of reliability, validity, homogeneity, and generalizability.
Last, coefficient alpha offers a useful and usable approach to assessing the reliability of measurement scales in marketing research. Though the development of alternative forms and multi-facet generalizability studies will be needed for situations in which time and other facets of measurement need investigation, alpha can be fruitfully employed for scales containing a minimum of 3 items. Clearly, the development of reliable scales is a necessary condition for improving the quality of marketing research and theory.