<p><b>Coherence has been claimed to be one among the difficult aspects of writing to teach and learn . Both teachers and learners find it hard to explain why one piece of writing is incoherent and another is not. There have been a few studies (Gramegna, 2007; Wikborg, 1990) investigating the types of coherence breaks in EFL/ESL writing and a few studies (Johns, 1986; Lee, 2002b) investigating the effects of teaching aspects of coherence on students’ writing performance. However, there have been no studies identifying the coherence problems in students’ writing and then using these to teach students how to avoid these coherence problems. In addition to teaching students to avoid coherence problems, an approach called topical structure analysis (TSA) (Liangprayoon, Chaya, & Thep-ackraphong, 2013; Schneider & Connor, 1990) has been considered as an effective technique to help improve students’ writing coherence. Some studies (Attelisi, 2012; Connor & Farmer, 1990) have examined the effect of teaching TSA in promoting EFL students’ awareness of coherence, but investigating the effect of the combination of teaching coherence problems and TSA on EFL students’ writing performance has not been done in any previous studies. My study fills both of these gaps.</b></p>
<p>There are two main phases in my study. In Phase 1, I developed a system of coherence problems based on the analysis of 69 essays written by Vietnamese EFL students. Then, this system of coherence problems was used to identify the types and frequency of coherence problems in the writing of Vietnamese EFL students. In Phase 2, an intervention was conducted in a university in Vietnam. Fifty-eight students participated in the intervention and were divided into a control group and a treatment group. While the combination of teaching the system of coherence problems developed in Phase 1 and teaching TSA was delivered to the treatment group in the intervention, the control group was taught other aspects of writing such as grammar, vocabulary, etc. The aim of Phase 2 of this study was to examine the effect of the intervention on students’ writing performance, especially on coherence and overall writing quality.</p>
<p>Data for Phase 1 was 69 essays on a single topic. Using this set of essays, the types of coherence problems and their frequency were identified. Data for Phase 2 included three sets of writing. The first set of writing was 58 essays written by students in both the treatment group and the control group before the intervention was conducted. The second set of writing involved the students revising their first draft right after the intervention finished. The third set of writing was 58 new essays on a different topic from the first and the second sets written by students in the control and treatment groups four weeks after the intervention. These three sets of writing were analysed and rated by the same raters as in Phase 1. In addition, students completed questionnaires and participated in focus group interviews in Phase 2.</p>
<p>Findings of Phase 1 show that there were five main types of coherence problems that Vietnamese EFL students had in their writing. These were macrostructure-related problems, topic unity-related problems, paragraph unity-related problems, cohesion-related problems, and metadiscourse-related problems. Of these five main types of coherence problems, paragraph unity-related problems were the most frequent ones, and “no/ little elaboration of the proposition made” was the most frequent subtype of paragraph-unity problems. </p>
<p>Findings from Phase 2 show that there was a significant difference in terms of coherence problems, types of topical progression, and writing quality between the two groups after the intervention finished. To be specific, there were significantly fewer instances of coherence problems in the writing of the treatment group than in the writing of the control group, and the scores for coherence and overall quality achieved by the treatment group were significantly higher than those achieved by the control group. Also, the focus group interviews suggest that most students found the intervention helpful in improving their writing performance. This means the combination of teaching coherence problems and TSA to students had a positive impact on students’ writing coherence and writing quality. Based on the findings, some suggestions are made for teaching writing to EFL students in general and for the teaching of coherence in the Vietnamese university context in particular.</p>