Placement in Extensive Reading

Abstract

This thesis involves developing measures for placement in extensive reading. There are two phases in the research: (l) looking closely at some graded reader schemes and (2) constructing and trialling tests of vocabulary and reading. In the first phase, three studies were carried out; two studies examined the wordlists and reading schemes of graded readers and examined the GSL using the Windows-based Range program, and the other looked at text-based self-assessment. The results showed that none of the lists could represent the others as a word source for a vocabulary test. Data from observation, interview, and subjective and objective assessments indicated that text-based self-assessment, reflecting reading across various series of graded readers, can be a promising way of determining learners’ reading level. In the second phase, a 50-item-multiple-choice vocabulary-size test and a 75-item graded reading-based cloze test were constructed and piloted with 30 Thai learners of English. The statistical evidence showed that the tests had acceptable test characteristics, namely, item facility, item discrimination and internal consistency. It also indicated that the bilingual version of the vocabulary-size test was easier than the English version and the vocabulary-size and graded reading-based tests measured what they were intended to. The Thai version of the vocabulary-size test and the graded reading-based cloze test were then administered to 400 Thai learners of English and the scores were analyzed by SPSS 12.0 for Windows. The results confirmed the hypothesis that vocabulary should be a good predictor of comprehension of graded reading materials (r=.787, p<.01) and showed that the two tests can be used as placement tests to direct learners to the appropriate level of graded reading with a reasonable degree of accuracy.