Key points are not available for this paper at this time.
Recent research (1) has shown that words which appear frequently in written English, as measured by Thorndike-Lorge ( 4 ) word counts, are more likely to be rated as good by Ss than are infrequent words. This finding raises a question as to why this goodness-frequency relation exists. A simple explanation might involve ease of emission of frequenr as opposed to infrequent words. Frequent words are also generally short words ( 5 ) , with relatively few syllables. The frequent, short, word is easier to say than an infrequent, long, word. The greater ease of emission of frequent words may account fo r their greater rated goodness. The purpose of this pape r is to determine whether relation between goodness and frequency persists when pronounciability is held constant. W e obtained from lip-reading manuals (2, 3) 108 pairs of homophones (e.g., teamteem; straight-strait) of known Thorndike-Lorge frequencies. These homophone pairs were presented to 29 Ss in random order with order within pairs also determined randomly. Ss were told to encircle most pleasantly toned word of each pair. In 68 of 108 pairs more frequent word in pair was chosen more often as the more pleasantly A sign test indicates this preference for frequent members of homophone pairs to be significant (P < .01). Of a total of 3,132 possible choices, Ss omitted 14, chose more frequent words as more pleasant 1,836 times, and chose less frequent as more pleasant 1,276 times. A normal curve approximation of binomial expansion shows this preference for frequenr words to be highly significant ( P < .00001). Even with ease of emission controlled, frequent words are judged to be more pleasantly toned. The previously obtained relation between goodness and frequency does not seem attributable to greater ease of emission of frequent words.
Frincke et al. (Thu,) studied this question.