A Test of the Form-Resistant Correlation Hypothesis: Ratings, Rankings, and the Measurement of Values

In: Public opinion quarterly: journal of the American Association for Public Opinion Research, Band 52, Heft 4, S. 526-538

Abstract

A number of researchers have argued that ranking techniques are more appropriate than rating methods for the measurement of values in surveys. The form-resistant correlation hypothesis proposes that observed associations among values & between values & other variables should remain invariant across measurement methods. However, recent research on parental values for child qualities suggests that ratings & rankings produce different correlational results. Here, data from a sample of 466 US adults, collected as part of the 1980 General Social Survey, are used to test the hypothesis that discrepancies between rating & ranking results are due to the fact that, when responding to rating questions, some respondents avoid making difficult choices between valued qualities by rating all the qualities as highly & equally desirable. Consistent with this hypothesis, when nondifferentiating respondents are removed from the analyzed sample, the substantive results of analyses of rating data resemble the results typically obtained using ranking data, which suggests that ranking may be the superior method of measuring values. 3 Tables, 28 References. AA