Error Analysis on Medical Students' Writing Lee Eun-pyo (Eulji University, School of Medicine) This study was to identify and classify errors by analyzing medical students' writing, especially their formal and informal letters. Twenty five sophomore medical students in a class of 35 were the subjects of the study. These students were considered intermediate to advanced level based on their TOEIC score, because all of them scored higher than the national average of Korean college students (their average was 155 points higher than the national average 553). The study was also done to see if the subjects of high scores of TOEIC produced writing with less errors. The results were to be compared with the study done on the low level learners at other university. Many subjects referred to their Korean/English dictionary when they wrote. Some of them actually wrote a Korean version of composition and then translated word by word into English. Such phenomenon caused them to come up with awkward expressions in their writing. They took for granted that whatever, written in the dictionary, was absolutely appropriate and correct as well as direct translation. The study revealed that approximately one fourth of errors (26%) of these subjects resulted from L1 transfer. Other major errors involved in wrong words (16%), prepositions (15%) and articles (14%). Results and implications of the findings were further to be discussed. I. INTRODUCTION When learning a foreign language, it is indispensable to learn vocabulary of the language. And since a level of certain vocabulary is strongly needed for writing, many learners rely heavily on the use of Korean-English dictionary when they are faced with uncertainty of words and expressions. These learners, including of advanced level, are expected to make various errors. There are a number of studies done on vocabulary acquisition. Grabe and Stoller's (1997) case study of L2 Portuguese vocabulary learning suggests the effectiveness of using bilingual dictionaries when they are used in a consistent and appropriate manner. Frazer (1999) also emphasizes the importance of consulting a dictionary as an explicit L2 vocabulary learning strategy, especially when the learner wants to verify his/her inference of the meaning. The results of Park's (2001) study on Korean EFL Learners' Vocabulary Learning Strategies show that 84.5% of her subjects used bilingual dictionary for their vocabulary learning. If such dependence creates errors and awkward unnatural acquisition of the language due to transfer, then the transfer errors should be analyzed in depth to implement a desirable way of teaching them to acquire the language with minimum errors. Moreover, when teachers of L2 understand the types of learners' errors, they will be better able to assist them according to the learners' needs. This study was done to identify and classify errors of medical students' writing, both formal and informal letters. The purpose of it was to find out the most common errors and the frequency of errors the subjects made, to see if advanced students (the ones with high scores of the TOEIC) wrote longer with less errors, to compare the results with the previous studies on the low level subjects, and to see what implications could be made from the results. II. REVIEW OF THE LITERATURE Richards (1971) argues that many of the learners' errors come from the strategies that they use in language acquisition and the reciprocal interference of the target language items. Error Analysis (EA) would allow teachers to figure out what areas should be focused on and what kind of attention is needed in an L2 classroom. So the language teachers can be better able to develop curriculum and select materials that can facilitate L2 learning processes (Richards, 1995; Schachter & Celce-Murcia, 1977). It is important to understand what constitutes an error. An error refers to a systematic error of competence, both covert and overt, that deviates from the norms of the target language. Covert errors are grammatically correct but not interpretable within the context of communication whereas overt errors refer to the obviously ungrammatical utterances (Brown, 1994; Corder, 1967; Ellis, 1996). Errors need to be explained as to whether they are interlingual or intralingual. Generally interlingual errors can be identified as transfer errors which result from a learner's first language features (e.g., lexical, grammatical, or pragmatic, etc.). Intralingual errors are subdivided as overgeneralizations, ignorance of rule restrictions, incomplete application of rules, and false concepts hypothesized (i.e., learners fail to comprehend fully). Overgeneralization errors occur when learners yield deviant structures based on other structures of the target language. Ignorance of rule restrictions refers to the application of rules to inappropriate contexts. Incomplete application of rules arises when learners fail to develop a structure fully. False concepts hypothesized occur when learners do not completely understand a distinction in the target language. (Ellis, 1996; Richards, 1995). However, it is not always possible to distinguish transfer errors from intralingual errors. Classification of intralingual errors can also be problematic. Quite a number of researchers, such as Brown (1994), Littlewood (1995), Lococo (1976), and White (1977) have found that L2 learners at a beginning level produce a large number of interlingual errors. As these learners progress in acquiring the norms of the target language, more and more intralingual errors are manifested. Moreover, adult learners more commonly produce transfer errors than child learners. There are some previous studies on Error Analysis based on learners' written work. One of them, done many decades ago, was by Duskova (1969) who identified a total of 1007 errors based on the writings of 50 Czech learners of English and analyzed them in terms of 9 categories. She discovered that errors in articles (260) were most frequent, followed by errors in lexis (233) while there were 54 errors in syntax and only 31 in word order. Kim (1987) identified a total of 2455 errors in the English compositions of 12th grade Korean EFL learners. The findings showed that errors in BE and auxiliaries were most common (419), followed by errors in prepositions (287) and that intralingual errors arose more than transfer errors. In the following year, Kim (1988) investigated errors in English verbs with reference to tense, mood, and voice. The 120 subjects were the 11th grade Korean EFL learners who were asked to translate 42 Korean sentences into English. Results revealed that errors in mood were most frequent (903), followed by errors in voice (885) and tense (720), among the total of 2508 errors. With regard to the sources of the errors, overgeneralization (65%) occurred the most while L1 transfer occurred at 22% and simplification at 13%. Kim (1989) conducted EA with 200 10th grade Korean EFL learners, using their English translations of 30 Korean sentences. She identified 1122 errors in which transfer errors resulting from L1 structures were higher (24%) than overgeneralization errors (23%). In the essay writings of 200 10th grade Korean EFL learners, Kim (1998) identified 2122 errors and classified them in terms of 6 domains and subdivided them into 22 linguistic categories. Her findings revealed that errors in articles were most common (354) and that there were only 8 errors in word order and 2 in voice. There are some more studies done on spoken errors by Lennon (1991), Kim (1997), and Chin (2001) to name a few. According to Peyton (1988), Reid (1993), Staton (1988), it is shown that while routinely conveying ideas and/or messages closely related to themselves, students involved in writing tasks are able to become real individuals actually participating in social, communicative acts with meaningful, authentic purposes as much as possible. Error Analysis research has limitations of providing only a partial picture of learner language and having a substantive nature that it does not take account avoidance strategy in L2 acquisition since EA only investigates what learners do. Learners who avoid the exact structures they find difficult due to the differences between their native language and the target language may be viewed to have no difficulty with them as Brown (1994) and Ellis (1996) pointed out. III. METHOD The primary purpose of the study was to analyze what errors intermediate to advanced level learners at a medical college make in their writing by reviewing their formal and informal letters. Since these learners were considered relatively advanced level according their scores of the TOEIC, the results were also compared with those of basic level learners from the previous study. The number of errors and length of their writing were analyzed to see if they were correlated with their scores of an official test. The subjects' writing was evaluated and the sentences with errors were recorded to identify the types and frequency of errors. 1. Subjects The subjects in the present study were second year premedical students who took English Writing in the third semester of their two-year English curriculum. There were 35 students registered for the class. However, the students who did not take the TOEIC were not included in this study. Also a particular student, who had studied in England for over eleven years and scored 940 from the TOEIC, was excluded from the study because his writing was almost flawless. The majority of the subjects were sophomore students but 3 students graduated from another college and took the exam again to study medicine. There were 13 female and 12 male subjects involved in the study. All of them were hard working students at Eulji University located in Daejeon, central region of Korea. They were enrolled in the English Writing in the spring semester of 2000 and took the official exam at the end of that year. 2. Procedures Learners in the English Writing were given 6 topics to write about throughout the semester. However, two compositions were reviewed and analyzed to see the types of errors made. One was an informal letter to the person close to them and the other was a formal letter to a person in charge for information on a research program they might be interested in U.S. hospital. Their written words and sentences were counted to see if the length of the writing was correlated with the scores of the TOEIC. Each of their writing was reviewed by checking errors in different category and their frequency. The results were then compared with those of the studies done on the learners of basic level. The number of errors the subjects made was then examined if it was correlated with the scores of the TOEIC. Every sentence in the subjects' letters was reviewed. The total of 214 sentences that contained errors were then recorded. Of them, 78 sentences had multiple errors (36%). Then all the recorded errors were individually reexamined in order to determine the categories of errors. Furthermore, it was determined whether these errors could lead to communication breakdown to be deserved most careful attention as Ellis (1996) suggested. Teachers of L2 should be concerned what causes them to produce errors and how to help them with it. All 214 sentences were marked either I (incomprehensible) or C (comprehensible) to determine how many of them were seriously erred in conveying what they meant. For instance, the following two sentences show the difference in interpretation: * I want to know how expenses is needed, having a dormitory and scholarship. (I) * I need the material about scholarship. (C) The first sentence has several errors and the meaning is not clear due to the phrase "having a dormitory and scholarship." On the other hand, the second sentence is understandable even though it contains a couple of errors. It can be interpreted as "I need some information on getting a scholarship." Fifty eight sentences (27%) were incomprehensible. And of them, forty (69%) sentences were caused by transfer errors due to word by word translation. Among the 58 sentences, 9 (16%) were due to selection of the wrong word. IV. RESULTS AND DISCUSSIONS The number of words which students used for informal letter ranged from 87 to 777. And the number of sentences they wrote ranged from 10 to 71. The average numbers of words and sentences written in an informal letter were 202 and 19 respectively. However, they must have not felt comfortable writing a formal letter which was a request to obtain information regarding their imaginative research in the future. They used average of 141 words, ranging from 49 to 266, and 11 sentences, ranging from 5 to 23, for a formal letter. To determine any correlation between the TOEIC score and their writing, the results of the TOEIC were reviewed. It showed that their mean score was 708 which was 155 points higher than the national average of college students of 553 according to Newsletter released by the TOEIC. The second highest scored subject with 855 was considered the highest for the study because the student with 940, whose writing was flawless, was excluded. There were 2 students scored below 600, 10 students between 600 to 695, 11 students between 700 to 795, and 3 students between 800 and 855. The lowest was 595, scored 260 points lower than the highest. Figure 1 The Results of the TOEIC 2 (595, 595) 10 (670,670, 645, 690, 665, 605, 605, 665, 630, 660) 11 (770, 775, 770, 705, 730, 775, 710, 790, 715, 770) 3 (825, 855, 820) TABLE 1 The Number of Words, Sentences in the letters, Errors, and the TOEIC Score _______________________________________________________________________________ No IN WI SI WF SF ER TOEIC_ __ 1 JA 190 16 173 21 21 770 2 YW 186 14 186 16 16 645 3 YL 158 16 128 12 17 665 4 JS 139 16 143 12 23 660 5 JO 235 18 177 14 23 775 6 JW 136 8 6 770 7 JC 246 25 159 10 18 855 8 SY 176 18 106 6 15 705 9 JK 124 13 145 7 13 595 10 JW 104 17 111 10 20 595 11 SN 777 71 266 20 33 690 12 TK 139 16 125 7 12 630 13 HR 122 10 147 10 16 825 14 ST 265 23 126 13 12 730 15 JY 137 17 122 13 12 775 16 JY 339 33 239 11 16 820 17 CH 152 14 122 11 24 710 18 SK 87 11 49 5 10 665 19 JS 135 16 112 8 11 790 20 WK 121 14 126 11 10 605 21 YH 119 16 106 7 10 605 22 SY 166 13 117 9 19 670 23 MY 185 22 223 23 17 715 24 SY 206 18 106 9 15 670 25 NY 181 18 88 7 12 770 ________________________________________________________________________________ AVERAGE 202 19 141 11 16 708_ _ IN: WI: SI: WF: SF: ER: Initials of subjects' names Number of words written in informal letter Number of sentences written in informal letter Number of words written in formal letter Number of sentences written in formal letter Errors each subject made The results show that there is a correlation between the words the subjects used in informal letters and the TOEIC scores. The higher the scores, the more words were used. TABLE 2 Correlation between WI and the TOEIC score WI TOEIC Pearson Co Corr N Pearson Co Corr N WI 1.000 23 .556* .006* 23 TOEIC .556* .006 23 1.000 24 However, there was no correlation between the number of errors and the TOEIC scores or the number of words in formal letters and the TOEIC scores. In other words, high scored from the TOEIC does not mean less errors in the writing. There were total of 401 errors in the two letters. The most common errors were transfer errors from L1 (26%) including wrong word order. The majority of the subjects used Korean/English dictionary to refer certain words and expressions they wished to write. Moreover, as they had a tendency to translate from Korean to English, they produced some awkward expressions which made almost impossible to interpret the meaning. Then the selection of wrong words was the next noticeable error accounting for 16% of the total errors. For example, when they wanted a reply, they wrote "Please answer my letter." Or when inquiring what can be acquired from the research, it was written as "I want to know what I learn in there concretely." Followed were errors of prepositions and articles which accounted for 15% and 13% respectively. The results of the study show some different outcome compared with the studies done on foreign subjects' writing. However, compared with the results from the ones done on Korean subjects, they are somewhat similar in those most common errors are from transfer, prepositions, and articles though some studies show errors were common in Be, auxiliaries, and overgeneralization.. However, one noticeable difference revealed in this study was that using the wrong words (16%) was the second most common errors in the writing. Even the students of good command of English sometimes do not distinguish inappropriate words from commonly used words. The next table shows 14 sub-categories of which the errors were divided. TABLE 3 Sub-Categorized Errors Into Fourteen Groups: ____________________________________________________________________________________ Description Frequency ____________________________________________________________________________________ 1) incorrect use of word order caused by transfer or awkward expressions and words (106 errors, 26%) 2) incorrect selection of word (64 errors, 16%) 3) incorrect use or deletion of preposition (60 errors, 15%) 4) omission or incorrect use of article (54 errors, 14%) 5) incorrect plural (25 errors, 6%) 6) incorrect use of tense (24 errors, 6%) 7) incorrect use or omission of pronoun (18 errors, 4%) 8) subject verb agreement (16 errors, 4%) 9) incorrect use of noun (12 errors, 3%) 10) redundant use of words (9 errors, 2%) 11) incorrect use of verb BE (4 errors, 1%) 12) incorrect use of gerund (4 errors, 1%) 13) incorrect use of auxiliary (3 errors, 0.7%) 14) incorrect use of voice (1 error, 0.2%) ____________________________________________________________________________________ Other than errors from transfer, errors of wrong word, preposition, and article were the next commonly made ones. The rest sub-categories were rather minor that they were not analyzed any further. Also for the present study, spelling and punctuation errors were not taken into consideration. The next table shows the examples of transfer errors. TABLE 4 Examples of Transfer Errors ___________________________________________________________________________ ___________________________________________________________________________ Description Example ___________________________________________________________________________ Translation of "gomapgetseumnida," or would appreciate, to "will thank." I will thank you for introducing another expert. Then I will be great if I meet you. I will greatly appreciate to you if you send me the materials of your school. I will thank you if you inform me. Translation of "dapjang," or reply, to "answer." I long for your answer. Please answer my letter. / answer about my inquiry. I will wait from your answer. Translation of "seonbae," senior, to "superior." I waited school superior. / (Also wait was used without the preposition "for" because in Korean, it is not really necessary.) Direct translation of Korean version to English. How about your life? Even very late, I congratulate your entrance new school. I don't know what I prepare to admit entrance of university. I halt a study in my school. I will have a rest from my college for next one year. I want to know how expenses is needed. Sending you letter like this is not to bury with you in memory of the past. I will live very hard. You should marry your girlfriend after passing the exam. If there are scholarship for foreign intern, please give me data. I want to know what I learn in there concretely. I want to research about preventive medicine. I want to ask leave of absence for 1 year and I will go to your university. Whenever I look the ceiling to sleep, I can't image that I did drive. I am writing with hand. It's natural that our unripe feeling of love is ends up with parting. Why don't you go downtown and engage in animated conversation together? For coming true my dream, I decide to study under more advanced surroundings. I hear your news. ____________________________________________________________________________________ As shown above, the direct translation of Korean words and sentences into English caused their writing so awkward that clear interpretation was not possible for some. Using the wrong words, accounted for 16% of the total errors, was also related to the transfer errors because the word "accepting" in "I would like very much to know the of accepting some scholarship" was a direct translation of "bat-da" or receive in Korean. Therefore, if this category is considered as part of transfer error, then the study shows fairly high frequency rates of transfer (42%) compared to any other studies done in the past. The following table lists the examples of preposition errors. TABLE 5 Examples of Preposition Errors. _____________________________________________________________________________ _____________________________________________________________________________ Description Examples ______________________________________________________________________________ Omission of on I congratulate sincerely your marriage. From now, I will study hard. Omission of to I look forward your answer. Omission of for Addition of to you Addition of about Addition of in Addition of of Incorrect Use of Preposition by -> until at -> in about -> on on -> in to -> into approval -> approve of) For -> In order to I would like to apply the research course. I will greatly appreciate to you if you send me the materials of your school. Answer about my inquiry. I promised to meet her in last weekend. Despite of living the same city, we don't meet for six months. You are care of yourself by the time we will meet. I want to pursue the study at foreign country. I want to research about preventive medicine. I am spending great effort on my life. Science is divided to many ways. My parents didn't approval this. For coming true my dream, have my dream come true __________________________________________________________________________ Some of these preposition errors also had to do with translating from Korean. The next table shows the examples of article errors. TABLE 6 Examples of Article Errors __________________________________________________________________________ __________________________________________________________________________ Description Examples __________________________________________________________________________ Omission of Article "a" We have to dedicate to anyone who are in inferior situation. I want to pursue the study at foreign country. UCLA is very nice place. I believe he is wonderful boy. I can feel he is kind boy. I hope you meet handsome guy. Did you good job? I am student of E. University. It was very great issue. I am 2nd year student. I hope you introduce me to professor who help me for my research study. It was interesting class. I hope we will have good time together after a long time. It's fearful experience for me. It's very special experience the 4500 students of 36 medical school gathered. Blind date doesn't only intend to make boy friend, but also make various friends. Incorrect Use of "the" I need the material about scholarship. It is important through the all my life. I would like very much the of accepting some scholarship. You are most important friend in my life. End of the all, I want to research study in field of psychiatry. Incorrect Use of "a" E-mail is so faster than a hand writing letter. I am concerned about financial problem and would like to know about a scholarship for foreigners. Getting a high grades of all subjects, I want to go abroad to get a whole schooling system and regulations. _________________________________________________________________________ In this category of errors, omission of indefinite article occurred in 16 cases (62%) out of 26 article errors. Incorrect use of definite article appeared in 6 cases (23%) and whereas incorrect indefinite article occurred in four sentences (15%). Other errors included incorrect tense, subject-verb disagreement, redundancy, plural, BE, auxiliary verbs, and voice errors. However, these errors accounted for less than 4% each, they were not listed one by one or analyzed any further. Like it was confirmed in other studies, subjects relied heavily on word by word translation and using bilingual dictionary to learn new vocabulary when they were asked to write (See Park (2001)). An interesting fact was revealed in a study done by Park (2001) that even the elementary students actively utilize the available contextual clues to determine the meaning of a word. V. CONCLUSIONS AND IMPLICATIONS The strong motivation for this study was to examine what kind of errors learners of intermediate to advanced level at a medical college make in writing by reviewing their formal and informal letters and their frequency rates in order to determine what teachers of English should do to help learners minimize those errors. The results reveal that these errors were multiple types. It was learned from the study that these learners tended to translate from Korean to English word by word when they wrote and used bilingual, that is Korean/English dictionary for their reference. About a quarter of the learners (26%) made errors that were resulted from transfer, which is similar to the study done, showing 32% errors from transfer, by Chin (2001) and higher than Kim's (1997) showing 15%. Some of these errors were considered crucial because the intention of writers was not clearly delivered. If the selection of wrong words is viewed in the same category as transfer errors due to some of those errors might have caused from transfer as in "Could you send me a letter about this problem?" meaning "Could you send me a catalog or brochure on this matter?" then this study reveals that the errors these subjects made account for 42% which is a lot higher than any studies done previously. In Korean, a catalog can be referred to a letter and an issue or matter to a problem. Selection of a certain word may be interfered if direct translation prevails. Therefore, such errors can be reduced if teachers discourage such practice and let them be aware of the awkward expressions. Learners should be exposed to such examples with or without having similar situations. As the results of the study turned out similar to the other recent studies including the most recent one done by Chin (2001), the types of Korean students' errors are similar other than the rates of incomprehensible sentences regardless of their levels achieved in an official test. If teachers of English can produce a guide-book on what errors most of them make in their writing and let them study along with their English study, their errors are expected to reduce because of their error patterns. REFERENCES Brown, H. D. (1994). Principles of language learning and teaching (3rd ed.). Englewood Cliffs, NJ: Prentice Hall. Chin, C. (2001). Error Analysis: An Investigation of Spoken Errors of Korean EFL Learners. Duskova, L. (1969). On sources of errors in foreign language learning. International Review of Applied Linguistics, 7, 11-36. Ellis, R. (1996). The study of second language acquisition. Oxford: Oxford University Press. Frazer, C. A. (1999). Lexical processing strategy use and vocabulary learning through reading. Studies in Second Language Acquisition, 21(2), 225-241. Grabe, W., & Stoller, F. L. (1997). Reading and vocabulary development in a second language: A case study. In J. Coady & T. Huckin (Eds.), Second language vocabulary acquisition: A rationale for pedagogy (pp. 98-122). Cambridge: Cambridge University Press. Kim, H. (1987). An analysis of learners' errors made in their English composition especially in the high school level. Unpublished master's thesis. Chungbuk National University, Chungju, Korea. Kim, I (1988). A study of the errors in the use of the English verbs with special reference to tense, mood, and voice. Unpublished master's thesis. Busan National University, Busan, Korea. Kim, M. (2001). The Use of Written Dialogue Entries in a College EFL Composition Classroom. English Teaching, 56(4), 31-53. Kim, Y. (1997). A study of errors and intelligibility of Korean college students' utterances. Unpublished master's thesis. Kangwon National University, Chuncheon, Korea. Lennon, P. (1991). Error: Some problems of definition, identification, and distinction. Applied Linguistics, 12(2), 180-196. Park, J. E. (2001). Korean EFL Learners' Vocabulary Learning Strategies. English Teaching 56(4) 3-30. Peyton, J. K. (1988). Mutual conversations: Written dialogue as a basis for student-teacher rapport. In J. Staton, R. W. Shuy, J. K. Peyton, & L. Reed (Eds.), Dialogue journal communication: Classroom, linguistic, social, and cognitive views (pp. 1830201). Norwood, NJ: Ablex. Reid, J. M. (1993). Teaching ESL writing. Englewood Cliffs, NJ: Prentice Hall Regents. Staton, J. (1988). Contributions of the dialogue journal research to communicating, thinking, and learning. In J. Staton, R. W. Shuy, J. K. Peyton, & L. Reed (Eds.), Dialogue journal communication: Classroom, linguistic, social, and cognitive views (pp.183-201). Norwood, NJ: Ablex.