The effects of task, text, and proficiency on second language reading

Recent research in L1 reading has suggested the facilitating effects of tasks during reading vary depending on variables such as text types and learners' factors (McDaniel & Einstein, 1989). The purpose of this study was therefore to examine issues of task, text, and proficiency in second language reading. In the Main Study, two groups of Japanese university EFL students (N = 103) varying in English language proficiency read a narrative passage and an expository passage in one of the following three task conditions: outlining, answering embedded questions, and reading only. All three groups then produced written recalls immediately after reading and one week later. Written recalls were propositionally analyzed. Prototypical recalls were also created for qualitative analyses.; English language proficiency emerged as the strongest factor affecting the results. Text type was also important with all students scoring higher in recalling main ideas from the narrative passage than from the expository passage. There were no significant differences related to task types. The same effects persisted over a week, except text type affected both main ideas and total ideas recalled in the delayed test. An analysis of the text presentation order showed a significant interaction between the presentation order and text type for total ideas recalled and suggested that knowledge about an upcoming task influenced the recall performance differently between expository and narrative texts. The qualitative analyses showed qualitative effects for the task types on the information recalled by the participants.; In a Follow-up Study, Think Out Loud (TOL) protocols were analyzed from 18 students similar to those in the Main Study. The data revealed that these L2 learners concentrated on lower level processes when no particular learning task was imposed on them, and that higher level processes were induced by the outlining task. Based on these findings, pedagogical implications and limitations of the study are discussed.