Online Learning Units on Underused Adjectives for College EFL Students Online Learning Units on Underused Adjectives for College EFL Students Yi-Hsin Lee (李怡欣) Yu-Li Yeh (葉由俐) National Tsing Hua University [email protected] [email protected] This study aims to design and evaluate five online units for increasing students’ awareness of underused adjectives for EFL writing. Data-driven learning (DDL) was incorporated to engage learners in distinguishing synonymous adjectives from a bilingual collocation concordancer, TANGO. Each unit involves several synonyms of one of five overused adjectives, important, beautiful, hard, deep, and big, and presents two tasks for practice. The first task requires students to induce patterns from concordance lines of adjective-noun collocates from TANGO. The second task was for students to actively practice employing a stronger and more specific adjective in context. 19 English majors in a freshman writing class did the online exercises for 4 weeks. The findings indicated that, from the three controlled tests, students made significant improvement in the immediate posttest and retained their word knowledge for synonym use after a time lapse of two months. In writing, moreover, students avoided using general adjectives and tried to apply more specific items. As for the learning process, a majority of the students reported that inductive learning was beneficial though they found it time-consuming and sometimes difficult to verbalize the differences among semantically similar words. Finally, with mutual translation, TANGO was considered a useful tool for learning synonyms and their collocates. INTRODUCTION To communicate messages effectively in writing, writers are expected to avoid general terms or overused modifiers and use specific words. A case study comparing the corpora of French learners of English as a foreign language (EFL) and native speakers (NS) by Granger and Tribble (1998) explicitly points out the phenomenon of overused adjectives by non-native learners (NNS). The EFL learners are apt to be too dependent on superordinates such as real, important, and different throughout their text due to the lexical poverty of most learner output. Such phenomenon that learners overuse the above adjectives is what Ham and Rundell (1994:178, cited in Granger and Tribble, 1998) termed the default terms, a factor which makes a writing task a weak one. The finding contributes pedagogical implications for foreign language teaching and learning, and language teachers should encourage learners to employ words with a higher degree of specificity for successful communication. In order to solve the problem in overusing general adjectives, the use of concordancing with data-driven learning (DDL) could be an alternative to help learners. DDL presents abundant examples for learners to discover rules from contextual clues and examples in corpus evidence (Johns and King, 1991). Moreover, presenting concordance data to learners can help learners successfully discriminate among semantically similar items and attend to their collocational patterns and semantic features (Partington, 1998). The purpose of the present study therefore is to develop and evaluate online DDL learning units for helping learners apply synonymous alternatives of underused adjectives. It is hoped that students will be able to use more specific words and avoid the overused items in writing after they complete the instructional units. 157 Yi-Hsin Lee (李怡欣) & Yu-Li Yeh (葉由俐) LITERATURE REVIEW AND RESEARCH QUESTIONS Before bringing about our research issues, we will first review previous studies in the areas of vocabulary learning, comparison between learner corpus and native corpus, and implications for English synonym teaching and learning. Vocabulary Learning In this section, we focus mainly on the importance of word selection in writing as well as the causes for vocabulary errors and the teaching of synonyms. Vocabulary and Writing. The importance of word selection for writing has been recognized by Johnson (2000), who stresses that the compared with a reader, a writer is more obligated to be precise and should use the right word so as to express the intended messages. Studies have shown that vocabulary improvement and lexical selection in writing tasks were emphasized by evaluators of student writing. In one of the major studies in this area, Santo (1998) indicated that lexical errors were considered the most serious problems. Another study by Engber (1995) pointed out that the diversity of lexical choice and the correctness of lexical forms had a significant effect on reader judgment of the overall quality of essays written by L2 writers with intermediate to high-intermediate proficiency. Taking into consideration the frustration that learners experienced in spending much time searching for appropriate lexical items but still expressing themselves with imprecision, Santos (1988) suggested that lessons on vocabulary building and lexical selection be incorporated into an ESL writing course. The lessons should be designed with emphasis on the importance of lexical choice and elicitation or presentation of synonymous expressions Vocabulary Errors and Synonym Teaching. Martin (1984) examined vocabulary errors among advanced learners and suggested that the teaching of vocabulary via glosses or synonyms in the target language could possibly lead to improper lexical choices. When it is implied that an old word would act as the equivalency for a new item, learners might take the two words as exactly interchangeable alternatives. Further, if an exercise simply requires the learner to replace one item in a sentence with a new one in a list, the differences are likely left behind. It is cautioned that learners must be guided to notice whether synonyms behave identically in all contexts and to appreciate the subtle distinctions among semantically related words. What a teacher should do is to compare and contrast new words by pointing out the nearest neighbors and different situations in which each word occurs. Comparison Between Learner Corpus and Native Corpus The use of adjectives in writing was under investigation with a comparison between the Louvain corpus (French learners of English, 227,964 words) and a core subset of BNC (1,080,072 words) by Granger and Tribble (1998). The study showed that advanced French students used such adjectives as real, different, important, longer and true more frequently than proficient NS writers. It is revealed that learners tend to be overreliant on superordinate adjectives such as important in their academic writing, while excluding words with higher degree of specificity such as critical/crucial /major/serious/significant/vital, and so forth. Another contrastive study was launched by Gui and Yang (2002), who developed Chinese Learner English Corpus (CLEC) which comprises 1,185,977 words from writing of intermediate to advanced learners. They compared it with other corpora of English speakers. The comparison of big, great, and large used between CLEC and FLOB (Freiburg Lancaster-Oslo/Bergen) Corpus revealed that Chinese learners used great more frequently. 158 Online Learning Units on Underused Adjectives for College EFL Students Gui and Yang (2002) pointed out that Chinese learners regarded great as a general intensifier to apply to any cases. Besides, the misuse of big and large indicated that learners were still not familiar with their collocates. As a result, researchers advocate applying the results to pedagogy. Flowerdew (2001) has suggested that the findings from these comparative studies be utilized in design of materials to address students’ needs and deficiencies, as has been the case in the compilation of dictionaries for NNS. Instead of giving form-focused instruction based on language teachers’ intuition, Granger and Tribble (1998) proposed the idea of utilizing NNS learner data for a more systematic account of learner difficulties. Tschichold (2003) has also explicitly recommended that computer assisted language learning (CALL) activities can be adapted to help learners actively practice alternative words or expressions for the overused items. Such vocabulary enhancement activity could serve to strengthen learners’ knowledge of the target adjectives after the presentation of the teaching materials in the activity. Implications for English Synonym Teaching and Learning This section particularly emphasizes the pedagogical implications of studying corpus data recommended by researchers. In addition, a collocation concordancer aid, TANGO, is introduced for its application to synonym learning. The Importance of Studying Corpus Data. The study by Harvey and Yuill (1997) gave a detailed account of what role the Collins COBUILD English Language Dictionary (CCELD) (1987) played in the completion of written tasks by EFL learners. The learners were required to identify and distinguish various types of information about the word they could look for. Of the reasons for the look-ups, synonym searching was the fourth most frequent one. However, 36.1% of synonym searches were reported by the informants to be unsuccessful. In almost all these cases, the users indicated that the entry did not give them the information they need for using the alternatives. Although CCELD offers an extra column for main source synonyms, it fails to explicitly present with the synonyms their register, connotation, difference of nuance, or collocation, clues that will make it evident for learners to choose an appropriate one for use from the list. Learners, therefore, may encounter difficulty in selecting an appropriate synonym especially when they have little access to how the given synonyms occur in particular contexts. The suggestion made by Harvey and Yuill was to provide synonyms in conjunction with their collocational patterns, semantic features, and stylistic guidance rather than an implied equivalent alone. Corpus data, therefore, is a powerful alternative to serve such purpose in vocabulary learning and teaching. A Bilingual Collocation Aid–TANGO. As mentioned, encouraging learners to study collocational patterns of semantically similar words would be effective for synonym teaching. Therefore, a collocation aid, TANGO, is selected as a helpful tool to retrieve adjective-noun (AN) collocation from Sinoroma parallel corpus (Wu, et al., 2003). When the user types in a word, TANGO can display relevant citations in the bilingual corpus with (a) highlighted words and collocates, (b) occurrence counts, (c) clustered citations according to collocates, and (d) sorted citations according to frequency of collocation. Therefore, when the user submits a query of an adjective, possible AN collocates will be displayed on the return page. The distinguishing advantage of the bilingual collocation aid is that the highlighted collocates are shown with translation equivalents in context. With the availability of Chinese counterparts, users can easily examine the relevant instances that they need. It is claimed (Wu, et al., 2003) that the collocation concordancer is beneficial for inductive learning to intermediate or advanced learners, that is, the target learners in our study. 159 Yi-Hsin Lee (李怡欣) & Yu-Li Yeh (葉由俐) Research Questions Research on concordancing learning has provided strong evidence in that synonym learning through analyzing corpus data is particularly recommended for clarifying differences in meaning. Furthermore, words with apparent similarity in meaning still should be taught with their typical collocates (Harvey & Yull, 1997; Partington, 1998). Therefore, it is necessary to design concordance-based materials to increase learners’ awareness of collocations of near-synonyms for appropriate word use. Following the principles for designing concordance-based exercise (Hunston, 2002), our study uses NNS learner data in identifying learning difficulties and focusing on five overused adjectives by Taiwanese EFL learners. The study thus will address three research questions: 1. 2. 3. Are the online learning units effective in students’ learning of synonymous adjectives in a controlled context? Will the online material improve students’ use of synonymous adjectives in writing? What is students’ feedback on the concordance-based exercises? RESEARCH METHOD In this section, we will include description of the background of the participants, the instruments to be used, the development of materials, data collection procedures, and data analysis. Participants The study involved an intact class of 19 freshman students from the Department of Foreign Languages and Literature in National Tsing Hua University. The participants took freshman writing as a required course which contained 2 periods per week and each period last for 50 minutes. Most of the students had received formal instruction on English for six years during their junior and senior high school years. Instruments Two types of instruments, tests and questionnaires, were used in the study. First, three tests, a pretest and two posttests, immediate and delayed, were designed to assess learners’ synonym learning. There were two testing types, translation and blank-filling items, equally distributed to the five sets of synonyms--important, beautiful, hard, deep, and big. Two sets of questionnaires were designed to first obtain information about students’ background and their preference for learning at the outset of the study, and later students’ perception of the online practices after the experiment period. Material Development Since insights derived from learner corpus can provide a basis for improving learners’ overuse of adjectives (Granger and Tribble, 1998), the present study initiated a comparison between non-native speaker (NNS) corpus of Taiwanese learners and native speaker (NS) corpus before designing the online units. The NNS learner corpus contains, with a total of 114,045 words, descriptive and argumentative essays by freshman English-major students of National Tsing Hua University. The NS data for contrastive analysis is LOCNESS corpus, consisting of 66,598 words of argumentative writing by British students. The analysis was carried out by comparing word frequencies in the two corpora so as to identify overused adjectives by EFL students. Based on the result yielded from the comparison, important, 160 Online Learning Units on Underused Adjectives for College EFL Students beautiful, big, hard, and deep were the five words chosen to help improve the phenomenon of overuse. The comparison of high-frequency vocabulary in the Taiwanese learner corpus reveals that Taiwanese learners tend to use these relatively general words with a significantly higher frequency than native speakers do. To select synonymous words for our target learners, WordNet were chosen because detailed information for distinguishing semantically similar words is provided in it. However, since Tschichold (2003) stressed that learners needed to be offered comprehensible alternative words or expressions for practice, we then select only those words with higher frequency in British National Corpus (BNC) to serve as the target words for learning. Finally, to facilitate students learning with induction, only synonyms with sufficient instances provided by Sinoroma parallel corpora were selected (Wu, et al., 2003). Following Hunston’s (2002) suggestion, a list of synonyms was presented first in each learning unit. Learners would read the example sentences of each target synonym and make notes on a notepad online in the system. After reading the examples, learners would induce patterns and make a summary of how the synonyms can be appropriately used. Finally, exercises were provided to reinforce the synonymous adjective learning. There were in total five learning units developed in the environment, one for each of the five overused words identified. Procedures In the first week, background questionnaire was administered and students took the pretest. Students wrote on the topic, “Why I choose to study in Tsing Hua University” as the first writing task in the second week. They were also instructed on how to use the online learning units in a demonstration provided by the researcher. During the following four weeks, students were required to do various tasks in each of the five units for 20 minutes in class and complete the rest after class. In the sixth week, students took the immediate posttest and filled out the evaluation questionnaire. The researcher also interviewed students for general perception about the units or further clarification needed for their responses to the evaluation questionnaire. Later in the semester, students wrote the second composition about persuading an international student to study in Tsing Hua University. Students took the delayed posttest at around the end of the semester, 8 weeks after the immediate posttest. Data Analysis Both quantitative and qualitative data collected were analyzed to answer the research questions. The quantitative data from the pretest, the posttest and the delayed posttest were used to evaluate the effectiveness of the online learning units. The data from the background questionnaire, the evaluation questionnaire and interviews were coded and analyzed to present students’ feedback of the online units. RESEARCH RESULTS In the following section, the results of the tests, students’ use of adjectives in writing, and students’ feedback on the online units will be discussed. The Results of the Three Tests Because of the small sample size (N=19), Nonparametric Method was employed to analyze the test results. Wilcoxon Singed Ranks test was run to see if there were significant differences between (1) total scores of the pretest and the posttest (2) total scores of the posttest and the delayed posttest (Table 4.1). The total score of the pretest was significantly higher 161 Yi-Hsin Lee (李怡欣) & Yu-Li Yeh (葉由俐) than the posttest (p=.000< .05) with no significant difference found in the comparison of the posttest and the delayed posttest. Hence, the positive results indicated that students’ knowledge of synonyms had increased significantly after completion of the learning units. Additionally, students generally did not show much regression in the delayed posttest. The answer to research question 1,therefore, is that the online units did enhance students’ learning of synonymous words. Table 4.1 Result of Comparison Among Three Tests by Wilcoxon Signed Rank Sum Test Asymp. Sig. N Mean Rank Sum of Ranks Z score (1-tailed) a .00 .00 Immediate Negative Ranks 0 Post-Pre Positive Ranks 19b 10.00 190.00 -.318a .000* Ties 0c Total 19 d 9.00 81.00 Delayed Negative Ranks 9 Post-Pre Positive Ranks 8e 9.00 72.00 -.214a .426* Ties 2f Total 19 Note1: *p< .05 a. immediate post< pre d. delayed post< pre Note2: b. immediate post> pre e. delayed post> pre c. immediate post = pre f. delayed post = pre Since the test was composed of questions equally distributed to the five adjectives, further analysis was done into how students performed in the five sets of adjective and which synonymous adjectives were more effectively learned by the participants, given the online design. Figure 4.1 showed the mean scores of the five sets of synonyms. The result revealed that the mean score of hard was the lowest in the pretest. In the immediate posttest, important had the highest mean while the mean score of hard was the lowest. Students were found to obtain lowest scores for big in the delayed posttest. Thus, we could infer that students learned more for the synonyms of important, but they acquire less knowledge for synonyms of hard, after the online learning. Compared to the synonyms of the other four adjectives, the learning unit might be less effective for students to learn the synonyms of big. 20 15 important beautiful 10 hard big 5 deep 0 pretest immediate posttest delayed posttest important 7.05 16.37 16.26 beautiful 7.58 15.42 14.47 hard 5.32 12.53 13.37 big 8.21 13.32 12.42 deep 7.32 14.05 15.11 Figure 4.1 The Mean Scores for the Five Sets of Synonyms Students’ Adjective Use in Writing How students use the adjectives in their actual writing is now analyzed to answer research question 2. Both the total words and the observed adjectives (the five overused adjectives, 162 Online Learning Units on Underused Adjectives for College EFL Students important, beautiful, hard, deep, and big, and their synonyms) in the two essays were calculated respectively. The observed adjectives, then, were marked for their being general or specific. The first essay had a total of 3,336 words with 17 general adjectives while the second essay 4,578 words in total with 30 general and 21 specific adjectives. Our first step was to normalize the length of writing, that is, total words were divided by 100 to obtain the unit for each composition. Since not every student used the observed adjectives in both writing assignments, we only included students who employed the target items in both their first and the second essays for our description here. Table 4.2 showed the general and specific adjectives used by these students. Table 4.2 Students’ Adjective Use in the First and Second Compositions The first essay The second essay Trend 1 2 3 4 General/Unit Specific/Unit General/Unit S1 1/1.44 3/2.61 Specific/Unit S14 1/1.6 3/2.21 S9 1/1.66 2/2.06 1/2.06 S16 3/1.68 6/2.13 3/2.13 S18 1/1.99 1/1.57 1/1.57 S6 2/1.83 1/1.37 S8 1/1.09 2/2.38 S12 1/2.03 1/2.66 1/2.66 S15 1/1.55 1/2.53 2/2.53 S17 2/1.49 1/2.59 Four trends of students’ word use of specific or general adjectives in the two essays surfaced. The first was that students S1 and S14 still used the general and overused items without the awareness of employing specific alternatives. The second was that students S9, S16 and S18 did not improve in using general words but had tried to apply specific adjectives in their second writing task. S6 and S8, in the third category, had reduced the number of overused items. Finally, there were students who not only avoided the general adjectives but also learned to use words with a higher degree of specificity, as S12, S15 and S17. Generally speaking, except S1 and S14 in category 1, the students sampled for observing their word use in writing made improvement in reducing using general words and/or using more specific alternatives quantitatively. Next, how students actually used the target items in writing was compared. In the first essay, it was found that no specific adjective items had been used by students. Three general words, important, beautiful and big, occurred frequently in student’ writing. Among the three general words, beautiful, which was used to describe the campus or the view in the university, had the most frequent occurrences. Similar to the first essay, the three general adjectives, important, beautiful and big occurred in students’ second essay and beautiful was still the most frequently used. However, some students had tried to apply alternatives such as lovely, instead of beautiful, to “campus” and “scene”. Another noteworthy instance was that students learned to use crucial as in the sentence, “it is crucial for a school to have the quality of humanity”. The comparison evidenced that students used more specific adjectives such as crucial, significant, lovely, and pretty after learning through online units. It could also be inferred that the online units raised students’ awareness of avoiding using overused and general items. For instance, with the concept of beautiful, we found that 163 Yi-Hsin Lee (李怡欣) & Yu-Li Yeh (葉由俐) students were able to employ more specific adjectives which were not included in the online unit. The examples were splendid/enchanting/ gorgeous scenery, picturesque environment/view and wonderful campus. After the completion of the learning units, students themselves were able to look for an appropriate substitute to express their ideas. Students’ Feedback on the Online Units An evaluation questionnaire was administered, after the completion of the units, to obtain information about the participants’ perception of the online learning units. For students’ general attitude, around half of the students (52.6%) reported that they liked the synonym learning, 36.8 % held neutral responses and 10.5% of students responded with a negative attitude. Half of the students (53.1%) found it difficult to make distinctions among semantically similar words, and a majority of students (72.7%) indicated that it took much time to analyze corpus data. Despite such difficulties, the participants mostly (73.7%) agreed that mutual translations in the bilingual concordances helped them learn English synonyms effectively and a large portion of students (84.2%) reported that reading concordance lines could help them consider the subtle differences among the collocates of synonymous words. An interview with the participants was also conducted after the immediate posttest. For the improvement of the online units, students indicated that the system was sometimes unstable and there should be a clear leave-taking message after they complete the exercises. As for TANGO, a few students mentioned that the some Chinese equivalents were misleading in the bilingual corpora. Furthermore, when students wanted to consult the corpora of VOA (Voice of America) or BNC, the connection speed was oftentimes quite slow. In sum, the results of the third research question showed that a great majority of students perceived the synonymous learning with the concordancer positively. CONCLUSION This study targets at investigating whether online learning units could increase EFL learners’ awareness of synonymous adjectives and application of them in writing. 19 English majors in a freshman writing class did the online exercises for 4 weeks. The major findings indicated that students made significant improvement of synonym use in controlled context. Students’ word knowledge still retained in a time lapse after they finished the online learning. In writing, moreover, students avoided using general adjectives and/or tried to apply more specific items. Students’ feedback showed that they did benefit from inductive concordancing learning though it was time-consuming and somehow difficult to verbalize the differences among synonymous words. Two pedagogical implications can be drawn for EFL teachers and researchers from the findings. First, to decrease students’ difficulty when engaged in studying corpus data, their induction skill needs more training at the onset of the study. As the traditional teaching method emphasized deductive teaching, one that most students were more familiar with, they lacked the experience of discovering patterns or rules from authentic language data. Consequently, more guidance should be offered by teachers if concordancing is to be incorporated into the curriculum. Second, the bilingual collocation concordancer, TANGO, could be used for facilitating student vocabulary learning since it offers appropriate AN collocations and provides alternative words for writing. When teachers present semantically related words, the information from TANGO would benefit students in comparing and contrasting the synonymous items. In addition, TANGO could serve as a consulting tool for learners to search, through possible collocates, for a proper alternative adjective for use in writing. 164 Online Learning Units on Underused Adjectives for College EFL Students In the current study, the small number of participants is one major limitation. For future research, more participants in different major areas could be invited so that the result can be generalized to other population of EFL learners with different background. Also, more writing should be collected in order to increase the size of learner corpus for investigation and elicit additional data of students’ adjective use in free production. Further, a longitudinal study could be conducted to observe students’ word use over a longer duration since students might need more time to acquire the semantically similar words before being able to apply them in actual writing. REFERENCES Engber, C. A. (1995). The relationship of lexical proficiency to the quality of ESL compositions. Journal of Second Language Writing, 4,139-155. Flowerdew, L. (2001). The exploitation of small learner corpora in EAP materials design. In M. Ghadessy, A. Henry & R.L. Roseberry (Eds.), Small Corpus Studies and ELT: Theory and Practice. Amsterdam: John Benjamins. Granger, S. & Tribble, C. (1998). Learner corpus data in the foreign language classroom: form-focused instruction and data-driven learning. In S. Granger (Ed.), Learner English on Computer. London & New York: Addison Wesley Longman. Gui, S. & Yang, H. (2002). Chinese learner English corpus. Shanghai: Shanghai Foreign Language Education Publishing. Ham, N. & Rundell, M. (1994). A new conceptual map of English. In W. Martin et al. (Eds.) EURALEX 94 Proceedings. Amsterdam 172-180. Harvey, K. & Yull, D. (1997). A study of the use of a monolingual pedagogical dictionary by learners of English engaged in writing. Applied Linguistics, 18(3), 253-278. Hayakawa, S. I. & the Funk and Wagnalls Dictionary Staff (Eds.). (1969). Modern guide to synonyms and related words. New York: Funk & Wagnalls. Hunston, S. (2002). Corpora in applied linguistics. Cambridge : Cambridge University Press. Johns, T. & King, P. (eds) (1991). Classroom Concordancing. Special Issue of ELR Journal 4, University of Birmingham: Centre for English Language Studies. Johnson, D. D. (2000). Just the right word: Vocabulary and writing. In R. Indrisano & J.R. Squire (Eds.), Perspectives on writing: research, theory, and practice. Newark, Dela.: International Reading Association. Martin, M. (1984). Advanced vocabulary teaching: The problem of synonyms. The Modern Language Journal, 68(2), 130-137. Partington, A. (1998). Patterns and meaning: Using corpora for English language research and teaching. Philadelphia: John Benjamins. Santos, T. (1988). Professors’ reaction to the academic writing of nonnative speaking students. TESOL Quarterly, 22(1), 69-90. Tschichold, C. (2003). Lexically driven error detection and correction. CALICO Journal, 20(3), 549-559. Wu, J. C., Yeh, K. C., Chuang, T. C., Shei, W. C., & Chang, J. C. (2003). The Role of Natural Language Processing in Computer Assisted Language. Proceedings of Int’l conference on ELT and e-learning in an electronic age. Tamkang University, May 28-29. 165
© Copyright 2026 Paperzz