The Effects of Choice and Control in Computer-Assisted Language Learning in Teaching Supplementary Grammar to Intermediate Students of ESL and to Remedial English Students at the College Entry Level

Abstract

0 Several studies in the past fifteen years have measured the effectiveness of computer-assisted instruction (CAI), and many of these studies have been reviewed (see Vinsonhaler and Bass [1972]; Jamison, Suppes, and Wells [1974]; Edwards, Norton, Taylor, Weiss, and Dusseldorp [1975]; Kulik, Kulik, and Cohen [1980]; and Kearsley and Seidel [1983]). These surveys generally conclude that CAI is equal to or better than instruction with traditional media, and that this parity is accompanied by reductions in the time needed for instruction and by moderately favorable student attitudes toward computers. Nevertheless, there are many omissions in the studies surveyed above. For one thing, few were done at the college level. Kulik et al. (1980) were able to find only 59 studies at the college level meeting certain research criteria. Another omission is lack of attention to cognitive and affective variables in almost all of the studies in CAI which have been done to date. Boettcher, Alderson, and Saccucci (1981) find only two studies besides their own specifically addressing cognition with regard to CAI-a crucial omission if one contends that CAI is an appropriate vehicle for cognitively based approaches to teaching. There are also omissions in the research concerning certain experimental settings. Kulik et al. (1980) found that no studies whatsoever had been done at the college level using CAI lessons that were stand-alone units (i.e., not part of a larger course of study) and which were meant to supplement work done in class. Only 8 of the 59 studies surveyed were stand-alone units meant to replace a teacher, and only 3 of these were tutorials. None of these 8 studies dealt with language. A final omission in the research is an examination of variables within the CAI lessons themselves. Jamison et al. (1974) cite only two studies, one in programmed instruction (PI) and the other in CAI, comparing variables