We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
The Effects of Scaling on the Correlation Coefficient: A Test of Validity.
- Authors
Martin, Warren S.
- Abstract
Measurement is an important aspect of marketing research. However, the number of studies documenting the effect of measurement errors on commonly used statistics are few [4, 5, 6, 8, 10, 11]. Furthermore, it is not unusual to discover the application of sophisticated multivariate techniques to marketing problems with little regard for the measurement problem [7, p. 87]. The determination of the effect of varying the measurement scale on various statistics is important because this gives the researcher: (1) a guide for the creation of better models, (2) an a priori indication of the magnitude of the error term, and (3) an additional framework within which models can be evaluated. Generally the correlation coefficient decreases as the number of response categories grows smaller. The magnitude of the reduction increases with the absolute magnitude of the correlation coefficient. It is concluded that the amount of information lost by collapsing scales is greater when the original variables are highly correlated.
- Subjects
MARKETING research; MULTIDIMENSIONAL scaling; MARKET surveys -- Design &; construction; INDUSTRIAL research; STATISTICAL correlation; MARKETING models; QUANTITATIVE research; CONSUMER behavior research; PSYCHOMETRICS; SCALING (Social sciences)
- Publication
Journal of Marketing Research (JMR), 1973, Vol 10, Issue 3, p316
- ISSN
0022-2437
- Publication type
Article
- DOI
10.2307/3149702