The Qualitative and Quantitative Effects of Coherence-focused Writing Instruction : The Case of Japanese Junior and Senior High Schools

2003年在初中开展了一项研究项目。首先，对三个班的134名二年级学生进行了水平测试、预研作文和作文问卷调查。之后，所有参与者都接受了连贯写作的指导。从接下来的一周开始，每个班都被要求根据三种不同的教学方法写一篇作文：(a)方法1，主要关注连贯性和组织，而不关注语法错误；(b)方法2，主要关注在给定时间内尽可能多地写作；(b)方法3，默读相关的英语段落，然后和同伴讨论写什么。出于教学考虑，未安排对照组。方法1、2课时15分钟，方法3课时25分钟。每个班由他们所在高中的三位老师授课。研究人员查看了更多收到的作文，并在接下来的一周内将其连同评论一并退回。这样重复了六次。研究人员对参与者进行了与研究前相同的调查，如问卷调查。同样的研究方法，除了有一个对照班，在2004年在高中被采用，有139名参与者。分析测量是(a)连贯性（Mochizuki， Kubota和Iwasaki通过修改Aoki（1991）的方法开发的测量方法），(b)整体评估，(c)语法正确性（12个指标，如无错误T单位的数量），(d)数量（9个指标，如总单词），(e)语法复杂性（4个指标，如T单位的数量），(f)词汇丰富度（7个指标，如复杂单词的数量）。初中阶段的结果显示，假设1“以连贯性为中心的方法1在连贯性方面优于方法2和方法3”得到部分支持。假设2“方法1在语法正确性方面优于方法2和方法3”得到支持。假设3“以数量为中心的方法2在写作更多方面优于方法1和方法3”不被支持。假设4“方法1、方法2和方法3帮助参与者生成更多语法复杂的句子”得到支持。假设5“方法1、方法2和方法3有助于参与者使用更丰富的词汇”得到部分支持。对于Research Question 1“三种方法中哪一种在连贯性、语法正确性、数量和词汇丰富度方面分别更优？”结果表明，方法1在连贯性和语法正确性方面表现最好，在数量上没有差异，而方法3在词汇丰富度方面表现最好。对于研究问题2“写作能力和技巧与英语熟练程度有什么关系？”研究表明，在方法3的课堂上，整体评估与英语熟练程度高度相关。高中阶段的结果显示假设1和2都被支持，假设3部分被支持，假设4和5不被支持。对于Research Question 1，结果表明Method 1在连贯性、语法正确性和数量上都是最好的，在词汇丰富度上没有差异。对于研究问题2，熟练程度与连贯性、数量和语法复杂性的相关性较弱。少

A research project in junior high school was conducted in 2003. First, a proficiency test, a pre-research composition and a questionnaire on composition were given to 134 second-year students in three classes. Thereafter all the participants received instruction on coherent writing. From the following week, each class was instructed to write a composition based on three different teaching methods : (a)Method 1, composition with primary focus on coherence and organization without any attention on grammatical errors, (b)Method 2, composition with primary focus on writing as much as possible in a given time, and (b)Method 3, composition after reading a related English paragraph silently and then discussing what to write with a partner. No control group was arranged due to pedagogic consideration. Method 1 and 2 classes were given 15 minutes, and Method 3,25 minutes for the whole process. Instruction was given to each class by a total of three teachers of their high school. Researchers rec … More eived compositions and returned them with comments the following week. This was repeated six times. The same investigations as the pre-research ones such as a questionnaire, were done to the participants. The same research method except that there was a control class, was employed in senior high school in 2004 with 139 participants.The analysis measurements were (a)coherence (Mochizuki, Kubota and Iwasaki's measurement developed by revising Aoki's (1991) method), (b)holistic assessment, (c)grammatical correctness using twelve indices such as the number of error-free T units, (d)quantity using nine indices such as total words, (e)grammatical complexity with four indices such the number of T-units, and (f)vocabulary richness with seven indices such as the number of sophisticated words.The results of junior high school showed that Hypothesis 1 "Coherence-focused Method 1 is superior to Method 2 and 3 in terms of coherence" was partially supported. Hypothesis 2 "Method 1 is superior to Methods 2 and 3 in terms of grammatical correctness" was supported. Hypothesis 3 "Quantity-focused Method 2 is superior to Methods 1 and 3 in terms of writing more" was not supported. Hypothesis 4 "Method 1,2 and 3 help participants to produce more grammatically complex sentences" was supported. Hypothesis 5 "Methods 1,2 and 3 help participants to use richer vocabulary" was partially supported. For Research Question 1 "Which of the three methods is superior in terms of coherence, grammatical correctness, quantity, and vocabulary richness, respectively?" it was shown that Method 1 was best in terms of coherence and grammatical correctness, that there was no difference in terms of quantity, and that Method 3 was best in terms of vocabulary richness. For Research Question 2 "How do composition abilities and skills relate to English proficiency?" it was shown that holistic assessment correlated highly with English proficiency in Method 3 classes.The results of senior high school showed that Hypotheses 1 and 2 were both supported, that Hypothesis 3 was partially supported, and that Hypotheses 4 and 5 were not supported. For Research Question 1, it was shown that Method 1 was best in terms of coherence, grammatical correctness and quantity, and that there was no difference in terms of vocabulary richness. For Research Question 2, proficiency correlated weakly with coherence, quantity, and grammatical complexity. Less