The Reliability of Type-Token Ratios for the Oral Language of School Age Children

Abstract

This study investigated the alternate forms reliability of four type-token ratios (TTRs) of oral language samples obtained from 52 elementary school children (9 through 12 years of age). The four TTRs included the basic type-token ratio, the corrected type-token ratio, the root type-token ratio, and the bilogarithmic type-token ratio. Language samples of 600 words were segmented into 50-word, 100-word, and 200-word samples. Within each TTR measure, there were no significant differences among the means for samples of the same size, but all means for a given sample size differed significantly from the means of all other sample sizes. Further, for samples of the same size the reliability coefficients calculated for each TTR measure were neither consistent nor significant. These findings indicate that under the conditions of the present study TTRs are not comparable when calculated for different sample sizes ranging from 50 to 600 words, and further, that they are not reliable measures of the language performance of individual elementary school children from regular classrooms for language samples of 50 to 200 words.