Some Determinants of Word-Frequency Estimates

Abstract

Subjects saw Turkish nonsense words and estimated their frequency of occurrence (72 subjects, 12 words, Exp. I; 33 subjects, 16 words, Exp. II). Results indicated that: (a) low-frequency stimuli were overestimated while high-frequency stimuli were underestimated; (b) stimuli were judged more frequent when they were positively evaluated than when they were negatively evaluated; (c) stimuli were judged more frequent in a distributed presentation than in a massed presentation; (d) stimuli were judged more frequent when they were rated after a 2-wk. delay than when they were rated immediately; (e) a 2-wk. delay enhanced the interaction between true frequency and judged frequency; (f) stimuli were judged more frequent when they appeared at the beginning or end of the presentation period rather than in the middle.