The cloze test has been regarded as a valid and reliable second language proficiency test. However, the cloze test has not been fully exploited as a diagnostic instrument. But the responses made by the learners will provide us with a great deal of information about the learners’s linguistic competence.
The present research attempts to investigate some aspects of the linguistic competence of some Japanese who are learning English as a foreign language, by analyzing the incorrect as well as correct responses in a cloze test. It also attemts to determine the comparative reliability of two scoring methods (i.e. the exact-word method and the acceptable-word method). The Ss were 10 highschoolers (HS), 10 college freshmen (CF), 10 graduate students (GS), and 10 English native speakers (NS). A cloze test which used a passage of approximately 350 words which had every sixth word systematically deleted was constructed and given to the Ss. The total time for doing the test, the total number correct scores, and the scores for content words and function words were computed.
Concerning the time to finish the test, which may be considered as a measure of the Ss' language fluency, it was found that Japanese Ss took significant more time than the English Ss. However, no significant difference between CF and GS was found. The language fluency of the Japanese Ss which has reached a plateau at the advanced stage was discussed. As for the total number correct scores, there was a significant effect for the four groups. The Newman-Keuls test showed that the scores of the NS group was significantly the highest followed in decreasing order by the GS group, the CF group and the HS group. The scores of function words and content words were by and large the same as those of the total scores. But the scores for function words were much higher than those for content words especially in the HS group although there was no significant difference between these two kinds of words in the GS group and the NS group. The results are discussed from the viewpoint of the teaching method for elementary and intermediate level learners.
A comparison of the scores of the exact-word method and the acceptable-word method shows that the latter is more reliable. And finally an analysis of the idiosyncratic errors made by the Japanese Ss showed that the number of errors caused by so-called mother tongue interference was relatively small.