Vocabulary Size of the Eiken Grade Pre-1 Test: Is There a Variation Within the Same Section?

Abstract

Corpus-based studies have revealed a disparity in vocabulary size between written and spoken texts. It has also been found that text types within the same register affect lexical size. Taken together, it is suggested that it may take analyses on separate corpora to yield accurate vocabulary size of English tests. The present study examined vocabulary size of listening and reading passages of the Eiken Grade Pre-1 test from this perspective. The purpose of the present study was to determine whether there is a variation in vocabulary size among Listening Parts and within the reading section. Samples in the present study were derived from 16 past tests. Three corpora were compiled from the Listening Part 1, 2, and 3, respectively. Another three corpora were developed from the 300-word, 400-word, and 500-word reading passages, respectively. Results on the listening corpora showed that the most frequent 3,000 word families were necessary to account for 98% of running words in dialog-based Part 1 and that it took the most frequent 4,000 word families to cover 98% of words in monolog-based Part 2 and 3. Lexical-load analyses on the reading corpora followed by an examination into words outside the criterion word lists indicated that the most frequent 5,000 word families sufficed to yield 98% coverage in all the three reading corpora. Pedagogical suggestions in light of these findings are discussed in detail.