Sentence Combing Plus (SC+): its effects on students' writing quality and strategy use. (On-going topics).

Abstract

The purpose of the present study was to check, on the one hand, whether or not Sentence Combining Plus (SC+) can affect L2 learners' writing performance, and on the other hand, in what type of strategies are learners involved when doing SC+ exercises. Two homogenous groups of sophomore students were randomly assigned to experimental and control groups. The study consisted of two parts. In the first part, the experimental group was exposed to SC+ exercises as treatment and the control group practiced SC (Sentence Combining) only. The results of t-tests revealed that the practice of SC+ caused a significant improvement on the students' writing performance. This improvement was, in particular, evident in the areas of content, organization, and language use. In the second part of the study, ten students--five from the experimental and five from the control group--were randomly selected and studied for the type of strategies they used through introspective interviews. Results of the quantitative and qualitative analyses of the introspective interviews showed that students who had practiced SC+ used more effective strategies. Moreover, they paid more attention to aspects of cohesion and evaluation than did the students from the control group. Based on the findings of the study, we can safely state that the practice of SC+ in intermediate writing courses will improve students' writing skill. An evidence for this, besides students' well performance on a posttest writing task, has been an enhancement of effective strategy use by them.

**********

Sentence Combining (SC) and Writing Quality

Different techniques have been innovated to help L2 learners to improve their overall writing ability and quality. Sentence combining (SC) is one of these activities. By definition, SC is a technique in which the learner combines basic sentences to produce longer and more complex sentences.

Several pieces of research have been done into the effects of SC on L2 learners' writing ability and quality. The results obtained from these bodies of research show that this technique can enhance the quality of learners' writing (see, for example, Miller and Ney, 1967; Combs, 1975; Cooper, 1977, 1981; Klassen, 1977; Daiker et al., 1978; Swan, 1979; Kerek et al., 1980). While these pieces of research show the usefulness of SC technique, other investigations have criticized SC as a legitimate practice in writing classes (see, for example, Zamel, 1980; Crowhurst and Klein, 1983; Holzman, 1983; Enginarlar, 1994). Enginarlar (1994: 217) states the weaknesses of SC as following:

   Firstly, although they involve some meaning negotiation and composing at
   sentence level, they are in fact non-rhetorical exercises, removed from the
   realities of real composing processes. Secondly, in the absence of a
   context beyond sentence level, as is often the case, the class discussion
   of various solutions created by students is doomed to be ineffectual.

As such, Enginarlar (1994) presents a new form of the older technique. This is Sentence Combining Plus (SC+), which involves not only sentence combining but also ordering jumbled sentences to make meaningful texts. With regard to the type of strategies second language writers use at the time of writing, Johnson (1992) conducted a study on the learning strategies L2 writers used when doing SC tasks. The strategies reported to include: questioning, planning, restating content, constructing meaning, constructing cohesion, and evaluation.

The present study intended firstly to investigate whether or not SC+ exercises have any effects on the overall writing ability and quality of L2 learners by testing the following null hypothesis: H0: There is no significant difference between students' quality of writing whether taught by SC+ or SC.

The study also attempted to investigate the type of learning strategies the participants used during the completion of SC+ and SC exercises. The specific research question addressed in this part of the study was: RQ: What learning strategies do the participants most frequently use while performing SC+ and SC exercises?

Method

Participants

Thirty-three sophomore students from two intact classes majoring in Teaching English at the Department of Foreign Languages and Linguistics of Shiraz University participated in this study. These students had studied English academically as their major for two years (four semesters). They took part in this study in their fifth semester while they were attending Advanced Writing 1 as a two-credit course. They attended the course two sessions a week. One session of the classes was used for the present study and in the other session, the teachers followed their regular syllabus. In their previous writing courses, the students were taught the structure of English sentences, the principles of paragraph writing, and different types of paragraphs. In this semester, however, they were going to use and practice their previously learned language knowledge to write compositions.

The students belonged to the same proficiency level, however, they were given a pretest in the form of a topic to be developed into a composition for their writing ability. The topic (Describe a wedding party you attended recently and explain whether or not it was a success) lent itself to narration, description, and explanation. An independent t-test indicated that there was no significant difference between the two groups' mean scores at the outset of the study at 0.05 level, implying that the two groups belonged to the same population. Thus, one class was randomly assigned to the experimental and the other to the control group. In the second part of the study, ten students--five from the experimental and five from the control group--took part in introspective interviews as they were composing.

Design

The study used a quasi-experimental research design, which followed a pretest/posttest control group design. Due to the regulations of the university, two separate teachers taught the courses. However, the two teachers followed the same syllabus and willingly included the materials of the present research in their class schedule. They were oriented and briefed toward the procedures of the study. Moreover, the teachers welcomed the investigators in all class sessions and used their help to regulate the procedures and avoid any idiosyncrasies.

At the end of the term, the two groups were asked to take part in a posttest in the form of writing an expository composition on another topic. This topic (Describe the best film or TV serial you have seen and explain how and why it affected you) also lent itself to narrative, descriptive, and explanatory details. The compositions, both pre- and posttest, were scored using a rating scale suggested by Jacobs et al. (1981). The scale attends to five components: content, organization, vocabulary, language use, and mechanics. A 0.98 inter-rater reliability was reached between the two raters of the compositions.

Materials

Ten passages--half descriptive and half narrative--were selected for ten sessions for each group. The compound and complex sentences of these passages were extracted and parsed into simple sentences according to the outline suggested by Cooper (1973). The text readability of the selected passages were estimated through Fry's Extended Readability Graph found in Farhadi et al.(1995). The average text readability of each group of passages was six showing an intermediate level, and, therefore, appropriate for the students.

Procedures

The students of the experimental group were put into groups of two or three and followed three stages. In the first stage, they were given sets of simple sentences, extracted from the passages, on slips of papers and were asked to combine the simple sentences of each set to make meaningful compound or complex sentences. If there were some original simple sentences they were advised to keep them for the second stage. In the second stage, they were asked to arrange all the sentences (original simple sentences as well as the compound and complex sentences they had written) to make a meaningful and coherent text. In the third stage, they received feedback on their tasks and revised their texts. The students of the control group practiced only the first stage.

The data needed for the second part of the study were collected over a three-week period, in individual sessions with each participant. Before the main introspective interviews, each participant was instructed how to describe his or her thoughts as s/he would begin, work through, and complete the exercises. Think-aloud protocols and introspective interviews were audio-taped by the permission of the participants. Each protocol was then analyzed according to the coding manual of cognitive strategies proposed by Johnson (1992). An intra-coder reliability of 0.87 and an inter-coder reliability of 0.85 were achieved for the codings.

Findings and Results

Results of The First Part of The Study

Table 1 shows the means of the students' scores on the posttest. As indicated in the table, the mean of the experimental group was found to be much higher than that of the control group (82.5 vs. 58.8). The difference between the performance of the two groups on the posttest was significant (t= 5.49 at 0.05 level). Therefore, the null hypothesis was rejected indicating that SC+ had a remarkable effect on students' writing practice. See <http://rapidintellect.com/AEQweb/spri02.htm> Also, the two groups' pre- and posttest mean scores were compared using paired t-tests. The t-value was found to be significant for the experimental group (t=6.25 at the 0.05 level), but not for the control group (t=1.15 and p=0.275).

Since a five-component scorning scheme was used to score the pre- and posttest compositions and since the experimental group's improvement was statistically significant, a one-way ANOVA was applied to find out whether the differences among the five components were also significant. It was found that the differences between the five components were significant (F=21.94, P<0.05). A post hoc test (Scheffe test) indicated that students in the experimental group outperformed the students in the control group on the four components namely, content, organization, vocabulary, and language use.

Results of The Second Part of The Study

The analysis of the ten protocols yielded twelve types of learning strategies. This analysis showed that while completing SC+ and SC tasks, these participants used the same ten learning strategies identified by Johnson (1992). However, in addition to these ten strategies participants of the present study employed two more learning strategies. These two strategies were Language Awareness, a cognitive strategy, and Task Reaction, a social/affective one. By definition, Language Awareness (LA) focuses on the structure of sentences and applying appropriate grammatical rules, and Task Reaction (TR) shows students' feelings and emotions towards the task itself.

Table 2 presents the Chi-square values and percentages for each category of strategies employed by the participants during completing SC+ and SC exercises. See issue's website <http://rapidintellect.com/AEQweb/spri02.htm> Table 2 shows that in SC+ exercises the eight most frequently used strategies were: RC2 (Restating Content 2), RC1 (Restating Content 1), CM1 (Constructing Meaning 1), CM2 (Constructing Meaning 2), CC (Constructing Cohesion), E (Evaluation), LLP (Lower Level Planning), and HLP (Higher Level Planning). However, in SC, the five most frequently used strategies were: RC1, RC2, LLP, CM1, and E. The participants completing SC exercises used LLP more than those who engaged in SC+ ones. This shows that although both groups of participants tried to plan their sentence constructions according to local discourse constraints (i.e., at the sentence level), the participants of SC group did use this strategy more frequently. In contrast, the participants of SC+ group not only employed LLP, but also attempted to construct their sentences and join them based on the global discourse constraints (i.e., the whole, coherent text they were to produce).

Discussions

Based on the results of the present study, Sentence Combining Plus (SC+) proves to be more effective than the traditional SC practice in terms of fostering students' writing skill. Although both SC and SC+ exercises provided students with a context for writing grammatical sentences, SC+ exercises helped the students in the experimental group to go into a meaning negotiation process in order to produce coherent and meaningful texts. Such exercises also required students to attend to and use appropriate cohesive devices--a point which is not practiced in the ordinary SC exercises. We see a significant and meaningful difference between the two groups' mean scores on their posttest, which is an indication of the impact of SC+ exercises on the experimental group students, given the fact that the two groups did not show any significant difference in their pretest. Moreover, the students in the SC+ group showed a significant improvement within themselves based on the comparison between their pre- and posttest. Thus, we can safely conclude that the practice of SC+ in intermediate writing courses resulted in the improvement of both within and between group writing skills.

In terms of strategies, thinking of a coherent text led the students in the SC+ group to activate and use strategies more effectively. For example, two of the strategies causing significant differences between the two groups were CM1 and CM2. Since completing SC+ exercises demanded putting the sentences together to make a coherent and meaningful text, the participants had to form new ideas, make relations between these newly-formed ideas, and sometimes support the meaning, and relate content to their world background knowledge. Consequently, they employed CM2 more frequently.

Another example is the use of CC strategy which occurred more frequently in SC+ exercises. This indicates that SC+ exercises required the participants to attend to cohesion more than the participants of SC exercises. Furthermore, although both exercises, to some extent, required these participants to focus on the appropriateness of their constructions, "evaluation" occurred less frequently in SC exercises. An explanation for this difference can be that in SC+ the participants were asked to put their combined sentences into a text. This, in turn, made them evaluate their combined sentences against their fitness into the text they were to produce. So, with regard to the intended meaning and structure, they had to re-examine and revise the combined sentences and the whole text.

Conclusion

The results obtained from the first part of the study revealed that SC+ was highly effective. It helped the students of the experimental group to improve the quality of their writing. The t-tests and Scheffe' test showed that practice in SC+ made the students write compositions which provided the readers with much better content quality, well-arranged organization, and more appropriate and better grammatical sentences. Moreover, the findings of the second part of the study revealed that SC+ was superior to SC because of evoking learning strategies which were necessary for writing a good composition. The results of the Chi-square test indicated that the students of the SC+ group used those learning strategies which were really helpful to them in writing. Consequently, the second part of study substantiates the notion expressed by Enginarlar (1994: 219): SC+ can give more beneficial practice in revising and editing at the stage of pre-writing by making L2 learners aware of the variety of options as they compose, and by providing opportunities in a more realistic context to evaluate the appropriateness of their constructions in terms of effectiveness and naturalness, of expression.

The second part of the study was; therefore, another way of validating the impacts of SC+ exercises on the participants' overall writing ability. Therefore, because of the dual nature of SC+--practice in syntax and rhetorical aspects--the findings of the present study firmly approves the fact that SC+ engages the learners in syntactic manipulation and meaning negotiation. The role of the SC was subordinate to that of the SC+ at both sentential and inter-sentential levels. Enginarlar (1994: 222) elaborates these functions of SC+ as following: The rationale behind the design of SC+ was to add a rhetorical dimension to the traditional SC practice, in which syntax is in the foreground.

Implications and Applications

Based on the above findings, the following implications and applications can be given:

1. This study supports the fact that to write more efficiently and coherently, language learners need not only to practice syntactic manipulation, but also to be provided with situations in which they can practice the rhetorical aspects of writing such as cohesion, constructing meaning, and evaluation.

2. SC+ exercises increase student involvement in a non-threatening atmosphere and thus are very appropriate and enjoyable practice in student-centered classes.

3. SC+ exercises can be considered as an integrative activity, involving the use of the four major language skills. In addition to writing, students are engaged in reading and also in speaking and listening while discussing their outputs in the groups.

4. Because of the positive effects of SC+ on composition activities, we suggest that material developers and writing practitioners include this technique in writing textbooks and classroom materials.

References

Combs, W.E. 1975. Further effects of sentence combining practice on writing ability. Studies in Language Education 16. Athens, Georgia: University of Georgia Press.

Cooper, T.C. 1981. `Sentence combining: An experiment in teaching writing.' Modern Language Journal 65: 158-165.

Cooper, T.C. (1977). `A strategy for teaching writing.' Modern Language Journal 61: 251-256.

Cooper, T.C. (1973). `An outline for writing sentence-combining problems.' English Journal 62: 96-102.

Crowhurst, M. 1983. `Sentence combining: maintaining realistic expectations.' College Composition and Communication 34/1: 62-72.

Enginarlar, H. 1994. `Sentence combining plus: a new use for an old technique.' ELT Journal 48/3: 214-224.

Farhadi, H., Jafarpur, A., and Birjandi, P. 1995. Testing Language Skills from Theory to Practice. Tehran: Center for Studying and Compiling University Books in Humanities, SAMT.

Holzman, M. 1983. `Scientism and sentence-combining.' College Composition and Communication 34/1: 73-79.

Jacobs, H.L., Zinkgraf, S.A., Wormuth, D.R., Hartfiel, V.F., and Hughey, J.B. 1981. Testing ESL Compositions: A Practical Approach. Massachusetts: Newbury House Publishers.

ohnson, K.E. (1992). `Cognitive strategies and second language writers: A reevaluation of sentence combining.' Journal of second language writing 1/1: 61-75.

Kerek, A., Daiker, D.A., and Morenberg, M. 1980. `The effects of intensive sentence combining on the writing ability of college freshmen.' Language and Style 13/3: 81-90.

Klassen, B.R. 1977. Sentence combining exercises as an aid to expediting syntactic fluency in learning English as a second language. Published Ph.D. dissertation, University of Minnesota.

Miller, B.D. and Ney, J.W. 1967. `Oral drills and writing improvement in the fourth grade.' Journal of Experimental Education 36: 93-99.

Swan, M.B. 1979, `Sentence combining in college composition: Interim measures and patterns.' Research in the Teaching of English 13/3: 217-224.

Zamel, V. 1980. `Re-evaluating sentence-combining practice.' TESOL Quarterly 14/4: 81-90.

Abdolmehdi is associate professor of TESL and Applied Linguistics at the Department of Foreign Languages and Linguistics. He received his Ph.D. from OISE/University of Toronto in 1995. He has taught ESL and content courses at BA, MA, and Ph.D. levels. Farzaneh completed her MA in TEFL in the department of Foreign Languages and Linguistics of Shiraz University. She is currently an instructor in the Islamic Azad University of Abadan, Iran teaching ESL courses.