THE EFFECTS OF TYPE OF FEEDBACK, AMOUNT OF FEEDBACK AND TASK-ESSENTIALNESS IN A L2 COMPUTER-ASSISTED STUDY

Abstract

For decades SLA research has looked at learner-external factors as influencers of learning outcomes, one of them being computerized types of feedback, with non-conclusive results. Moreover, a potentially significant characteristic of feedback, amount of feedback, has often been overlooked in previous investigations (Li, 2010; Loewen, 2012). To date, only one study (Ellis, Sheen, Murakami, & Takashima, 2008) has addressed the effects of providing different amounts of feedback on the same error with results indicating no advantage for single or multiple corrections. In addition, while task-essential (TE) practice has been identified as a desirable and optimal feature when designing a task (Loschky & Bley-Vroman, 1993), and has been included in several studies, its isolated contribution has not yet been identified.

Within CALL, the present study seeks to fill these gaps by investigating the effects of type of feedback provided (i.e., “right/wrong” vs. rule explanation) during input-based practice (task-essential vs. non task-essential). Participants were randomly distributed in three groups: Rules [task-essential practice with rule explanation in the feedback], Right/Wrong [task-essential practice with right/wrong feedback], and NF [task-essential practice without feedback]. In addition, the amount of feedback was tracked in order to investigate whether there was a threshold from which participants started to require fewer feedback episodes to achieve accuracy. Finally, to determine the isolated contribution of TE practice, a fourth group, NTE (Non Task-essential practice without feedback), was included. Results indicated that “right/wrong” feedback was more beneficial than rule explanation when interpreting temporal sentences at immediate post- and delayed posttests. The production of critical items at immediate post- and delayed posttests was similar for both feedback groups. As for amount of feedback, different types of feedback did not require different amounts, and the same amount of feedback led to L2 development regardless of feedback types. Interestingly, feedback was negatively correlated to development. Regarding the trajectory of the feedback during treatment, the learning process had the shape of an inverted U that showed a gradual decline in the need for feedback. Finally, with regard to the role of task-essentialness, NF and NTE groups performed similarly indicating no clear benefit of task-essentialness.