부산대학교 도서관

This study sought to explore the construct validity of independent and integrated writing tasks included in the TOEFL iBT with regard to four aspects: 1) the linguistic features of Korean EFL students’ performance on independent and integrated writing tasks across language proficiencies 2) differences of written vocabulary usage in the two writing tasks 3) potential connections between the skills of reading, listening, and speaking, and 4) students' perceptions on the impact of topics in writing prompts. For the purposes of this study, a total of 101 secondary school students undertook TOEFL online practice tests administered by the Educational Testing Service (ETS) a total of three times each. The participants were divided into three proficiency levels (low, intermediate, and high) based on the average of their total scores. In addition, 303 independent and 303 integrated writing responses were analyzed for text variables. Through analyses of the independent and integrated writing tasks, several conclusions were made. First, discrepancies were observed among the variables that accounted for the independent and integrated tasks. Essay length, nouns, verbs, complex nominals, and pronouns were significant factors for the independent writing task while types (the unique form of words) and the number of academic words were significant factors for the integrated writing tasks. In addition, when writing scores were analyzed separately according to language proficiency levels, different results were shown for linguistic features. Second, the coverage rate of general and academic words for students’ independent essays differed significantly according to proficiency levels while the coverage rate of academic words did not differ for integrated tasks. Third, both the independent and integrated task scores correlated significantly with TOEFL reading and listening scores. However, the reading and listening scores explained a larger variance for integrated tasks than for independent tasks. Fourth, when the independent and integrated tasks for speaking and writing were examined, all tasks were shown to have a significant relationship with each other. Fifth, scores for each of the reading, listening, and speaking sections were examined for their accountability in the writing tasks. With regard to the independent writing task, speaking scores were the only significant predictor while reading and listening performance were found to be significant predictors for the integrated writing task scores. Finally, the influence of topics and perceived task difficulty were identified as significant factors for students’ integrated writing performance. These results are likely to provide meaningful insights into the teaching and assessing of writing in EFL contexts.