Options
Effects of planning and revising on Chinese ESL learners' text quality
Author
Ong, Justina
Supervisor
Zhang, Lawrence Jun
Abstract
This study examined the effects of planning, Pre-task (PT), Extended Pre-task 0F,ree-writ ing (FW) and Control (CG), sub-planning, Topic given (TG), Topicideas (TI) given and Topic-Ideas-Macro-structure (TM) given and revising, Draft @) available and No Draft (ND) available conditions, on the overall text quality, fluency, lexical complexity and Flesch reading ease of Chinese ESL learners' first and revised drafts of argumentative essays. This study was propelled by several reasons such as scarcity of research studies examining the effects of cognitive processes on text quality, contradictory theoretical perspectives on planning and free-writing strategies, and limited research studies coupled with inconsistent results on how manipulation of task variables affects the task performance. Furthermore, a review of eight prominent writing models pointed to the importance of planning, transcribing and revising processes in the production of written texts. However, to date, the direct impact of cognitive processes on text quality has not yet been established. As this study was inspired by the works of Kellogg (1998, 1990), Ellis and Yuan (2004), Glynn et al. (1982) and Galbraith and Torrance (2004), their studies were magnified and microscopically reviewed and evaluated, before presenting the findings of the current study.
Data collection was implemented through a self-designed experiment and post-writing questionnaires. A two-way Analysis of Variance (ANOVA) factorial design, 4 X 3, was conducted on the first drafts and a three-way Analysis of Variance (ANOVA) factorial design, 4 X 3 X 2, was conducted on the revised drafts. Nine sets of post-writing questionnaires were originally created. These helped the researcher to elicit information about writers' actions, cognitions and perceptions during the planning, writing and revising stages. The first part of the analysis involved 108 general questionnaires, 313 pre-writing tests and 313 pre-writing tests questionnaires of three genres; and the second part involved an examination of four dependent variables on 107 first drafts and 108 revised drafts and finally, an analysis of 108 post-writing experimental questionnaires.
Results show that the main effect of planning on the text quality of Chinese ESL learners' first drafts of argumentative essays was not statistically significant, F(3, 95) = 2.55, MSE = 96.84, p = .061, v' = .074. Even though there was no significant difference in PT, EPT, FW and CG on the analytical text quality of the first drafts of argumentative texts, FW had the highest analytical text quality rating, followed by CG, PT and EFT. The main effects of sub-planning were robust on the first drafts, F(2, 95) = 8.50, p < .001, 7" = ,152 and the revised drafts, F(2, 84) = 10.19, p < ,001, v* = .195, with TIM scoring significantly higher than TG, and TI scoring significantly higher than TG. Furthermore, the interaction effect of Planning X Revising of the revised drafts was significant, F(3, M) = 5 . 9 7 ,=~ ,001, i12 = ,176. After applying split file on the main planning conditions, independent t-tests were conducted on the D and ND conditions of FW, CG and PT. Writers in FW scored significantly higher in the ND than the D conditions, t(25) = -4.04, p = .001. A reverse direction was, however, seen in CG, with the D condition scoring nearly significantly higher than the ND condition, t(26) = 2.10, p = .045.
Results also reveal that the main effect of planning on the fluency 11, F(3,95) = 56.34, MSE = 5.72, p < ,001, qz = ,064, and the lexical complexity I, F(3, 95) = 26.93, MSE = 302.91, p < ,001, v2 = ,460, and the lexical complexity II, F(3, 95) = 30.97, MSE = .44, p c .001, v" = .494, and the main effect of sub-planning on the Flesch reading ease, F(2, 95) = 7.59, p = ,001, qz = .138, of Chinese ESL learners' first drafts of argumentative essays were all significant. For fluency II and lexical complexities, writers in the FW condition scored significantly higher than writers in both the PT and EPT conditions. For Flesch reading ease, writers in the TG condition had a significantly higher readability index than writers in both the TI and TIM conditions. Finally, results show that the main effect of sub-planning on the lexical complexity I, F(2, 84) = 5.05, p = .009, q2 = .l07 and lexical complexity 11, F(2,84) = 5.17, p = ,008, q2 = .110, of Chinese ESL learners' revised drafts of argumentative essays were significant, with TI having significantly higher lexical complexities than TG, and TIM having significantly higher lexical complexities than TG. The results were interpreted in the light of the previous research studies, theoretical frameworks and writing models. Finally, Cognitive Processes in a Task Environment ESL Writing Models are proposed.
Data collection was implemented through a self-designed experiment and post-writing questionnaires. A two-way Analysis of Variance (ANOVA) factorial design, 4 X 3, was conducted on the first drafts and a three-way Analysis of Variance (ANOVA) factorial design, 4 X 3 X 2, was conducted on the revised drafts. Nine sets of post-writing questionnaires were originally created. These helped the researcher to elicit information about writers' actions, cognitions and perceptions during the planning, writing and revising stages. The first part of the analysis involved 108 general questionnaires, 313 pre-writing tests and 313 pre-writing tests questionnaires of three genres; and the second part involved an examination of four dependent variables on 107 first drafts and 108 revised drafts and finally, an analysis of 108 post-writing experimental questionnaires.
Results show that the main effect of planning on the text quality of Chinese ESL learners' first drafts of argumentative essays was not statistically significant, F(3, 95) = 2.55, MSE = 96.84, p = .061, v' = .074. Even though there was no significant difference in PT, EPT, FW and CG on the analytical text quality of the first drafts of argumentative texts, FW had the highest analytical text quality rating, followed by CG, PT and EFT. The main effects of sub-planning were robust on the first drafts, F(2, 95) = 8.50, p < .001, 7" = ,152 and the revised drafts, F(2, 84) = 10.19, p < ,001, v* = .195, with TIM scoring significantly higher than TG, and TI scoring significantly higher than TG. Furthermore, the interaction effect of Planning X Revising of the revised drafts was significant, F(3, M) = 5 . 9 7 ,=~ ,001, i12 = ,176. After applying split file on the main planning conditions, independent t-tests were conducted on the D and ND conditions of FW, CG and PT. Writers in FW scored significantly higher in the ND than the D conditions, t(25) = -4.04, p = .001. A reverse direction was, however, seen in CG, with the D condition scoring nearly significantly higher than the ND condition, t(26) = 2.10, p = .045.
Results also reveal that the main effect of planning on the fluency 11, F(3,95) = 56.34, MSE = 5.72, p < ,001, qz = ,064, and the lexical complexity I, F(3, 95) = 26.93, MSE = 302.91, p < ,001, v2 = ,460, and the lexical complexity II, F(3, 95) = 30.97, MSE = .44, p c .001, v" = .494, and the main effect of sub-planning on the Flesch reading ease, F(2, 95) = 7.59, p = ,001, qz = .138, of Chinese ESL learners' first drafts of argumentative essays were all significant. For fluency II and lexical complexities, writers in the FW condition scored significantly higher than writers in both the PT and EPT conditions. For Flesch reading ease, writers in the TG condition had a significantly higher readability index than writers in both the TI and TIM conditions. Finally, results show that the main effect of sub-planning on the lexical complexity I, F(2, 84) = 5.05, p = .009, q2 = .l07 and lexical complexity 11, F(2,84) = 5.17, p = ,008, q2 = .110, of Chinese ESL learners' revised drafts of argumentative essays were significant, with TI having significantly higher lexical complexities than TG, and TIM having significantly higher lexical complexities than TG. The results were interpreted in the light of the previous research studies, theoretical frameworks and writing models. Finally, Cognitive Processes in a Task Environment ESL Writing Models are proposed.
Date Issued
2010
Call Number
PE1128 Ong
Date Submitted
2010