Getting closure on cloze: A validation study of the "rational deletion" method

Abstract

The present study was designed to (a) test the notion that the cloze procedure is related to the idea of closure in Gestalt psychology and (b) investigate the degree to which specific language skills (grammar knowledge, reading ability, and vocabulary knowledge) contribute to cloze test performance for second language learners and (c) investigate the degree to which cloze tests may require nonverbal reasoning skills. Forty-eight second language students from an Intensive English Language program sat for a battery of nine tests: a cloze test, a reading test, a vocabulary test, a grammar test, two tests of Closure 1, two tests of Closure 2, and the Raven’s Progressive Matrices (Plus version). Test scores were submitted to a factor analysis and multiple regression analysis. The factor analysis showed two clear factors, one verbal and one nonverbal. The multiple regression analysis indicated that grammar and reading are good indicators of cloze test performance but nonverbal abilities such as closure and higher order reasoning are not. Implications for teaching and research are discussed.