Premium
Graphic rating scales — How many categories?
Author(s) -
McKelvie Stuart J.
Publication year - 1978
Publication title -
british journal of psychology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.536
H-Index - 92
eISSN - 2044-8295
pISSN - 0007-1269
DOI - 10.1111/j.2044-8295.1978.tb01647.x
Subject(s) - psychology , judgement , rating scale , reliability (semiconductor) , discriminative model , scale (ratio) , task (project management) , validity , test validity , psychometrics , cognitive psychology , social psychology , developmental psychology , power (physics) , artificial intelligence , computer science , physics , management , quantum mechanics , political science , law , economics
Two experiments investigated the reliability and validity of a continuous and various categorized (five, seven, 11) rating scales on two different tasks (involving attitude judgement and psychophysical judgement). Although subjects appeared to prefer the continuous scale, it did not offer any advantages in terms of reliability or validity. It is proposed that a relatively small number of categories (five or six) should generally be used because ( a ) the five‐category scale was most reliable (at least on the attitude judgement task), ( b ) confidence judgements made by subjects using the continuous scale indicated that they were operating essentially with five or six categories, and ( c ) other evidence suggests that whereas there is no psychometric advantage in a large number of scale categories (greater than nine to 12), there may be a loss of discriminative power and validity with fewer than five.