magnitude estimation Technique
arnold on Sep 16, 2008
Bass et al (1974) answered this same question.
Specifically they pondered:
Is it possible to obtain reliably established geometric means and standard deviations of magnitude estimations of expressions of frequency and amount used to modify attitudinal ratings, and if so, does the importance of the issue to the rater affect their rating?
Findings: They did a study to show that it is possible to fix quantitative meanings that are associated with verbal judgments of frequency and amount. They further hypothesized that judges generate one set of ratios of extensively when rating issues of consequence to them and a different set when judging issues of little importance to them. This hypothesis was rejected. They were however able to come up with tables of 4-point through 9-point scales of expressions of frequency and amount as well as the percentage overlap in distribution between adjacent scale points. These were statistically significant and had small variances. Alas, they did show that it is possible to fix quantitative meanings that are associated with verbal judgments of frequency and amount. The 4-point through the 6-point appear to be the best to use as they have the least amount of overlap.
BASS, CASCIO AND O’CONNOR (1974), “MAGNITUDE ESTIMATIONS OF EXPRESSIONS OF FREQUENCY AND AMOUNT,” Journal of Applied Psychology, Vol. 59, No. 3, 313-320