Abstract
Abstract Previous research studies have shown that presentation of words in semantic clusters have harmful effects on learning. This study was conducted to investigate whether these harmful effects could be found in typical English classroom settings and what effects would yield if semantic clusters were studied under deep processing activities. Subjects in this study were eight male and 11 female students from a business-oriented program in a senior high school where the researcher taught. They studied twenty-eight English words grouped in four lists of seven in their regularly scheduled English classes. Two lists were semantically-related (SR) words grouped under the categories of physical complaints and professions, and the other two lists were semantically-unrelated (SU) words. One SR and one SU list were studied through shallow-processing learning activities, and the other SR and SU lists were studied through deep-processing activities. Results of this study showed that acquisition of SR words was not different from acquisition of SU words on the immediate recognition posttests under both shallow and deep processing conditions. Acquisition of SU words tended to be better than acquisition of SR words on the immediate production posttests under shallow learning, while acquisition of SR words tended to be better than acquisition of SU words on the immediate production posttests under deep learning. Besides, there was a tendency for SR words to be retained better than SU words under both shallow and deep processing conditions on the delayed recognition posttests. However, no difference was found between the retention of SR and SU words on the delayed production posttests under both shallow and deep processing conditions.