On the distribution of errors in transcription of primary school students in Vietnam

JOURNAL OF SCIENCE OF HNUE 2011, Vol. 56, N◦. 1, pp. 113-122 ON THE DISTRIBUTION OF ERRORS IN TRANSCRIPTION OF PRIMARY SCHOOL STUDENTS IN VIETNAM Nguyen Thi Cam Huong (∗) Hanoi National University of Education Yusuke EDA Wakayama University, Japan Nguyen Thi Hoang Yen The Vietnam Institute of Educational Sciences (∗)E-mail: nch19381@yahoo.com Abstract. In this paper we study the development of handwriting ability of primary school students by investigating the number of errors appearing in their transcription. More precisely, 1557 primary school students were asked to copy a paragraph in a period of time and then the errors in their samples were collected and statistically analyzed. The results yielded that: The fre- quency of errors in the transcriptions tended to increase in grade; the upper grade students tend to make more errors than the lower ones, in particular, among all students, the number of students who made errors in grade five is the most and the number of the ones in grade two is significantly the least. Furthermore, there is a gender discrimination in the transcriptions. Among the second, fourth and fifth grades, the proportion of boys who made errors are significantly more than that of the girls. Our results imply the necessity to consider the relationship among handwriting errors, speed and legibility in Vietnamese language. Keywords: transcription, error, handwriting, primary school students. 1. Introduction The development of writing ability is considered as an essential ingredient for success in school and the failure to attain handwriting competency may affect both academic success and self-esteem of students [2]. A lot of previous researches on handwriting have focused on the two most important factors in handwriting performance - speed and legibility. ( quite right) They reported that the handwriting speed increased along with the age and grade of students; however this relationship was not linear [14,6,10]. Graham et al. [6] also indicated that the student’s handwriting legibility improved during the primary grades. While Blote & Hamstra-Bletz [1] who studied handwriting legibility of students in primary grades demonstrated that the legibility of handwriting tended 113 Nguyen Thi Cam Huong, Yusuke EDA and Nguyen Thi Hoang Yen to decrease with the speeding up of handwriting and students tended to simplify their handwriting in higher grades. Furthermore, in consideration that handwriting error is an aspect which affects the legibility of handwriting, Kono [10] did a study on handwriting errors of students in primary schools and his results indicated that the errors did not increase with the grades of students. On the other hand, the effect of gender on handwriting was also studied and in- dicated as an important factor in handwriting development. Spear [13] who studied the elements differing with the written work of boys and girls reported that girls’ work tends to be neat and well presented while that of boys is often untidy and poorly presented. Harley [7] afterwards found out that this difference was clearly recognized in seven to eight year-olds. Moreover, Graham et al. [6] with their study on handwriting of students in Grade 1 through to Grade 9 indicated that girls could write not only faster but also more legible than boys. So once again, the gender discrimination between boys and girls was also supported by a study of Demie [2] which showed that girls tended to perform at higher levels than boys in educational achievements. Vietnamese language is a phonetic symbol language, thus, there is an one-to- one relationship between phonetic and letters/signs in writing. The current Viet- namese language uses a writing system which is an adapted version of the Latin alphabet with additional diacritics for tones and certain letters. There are 29 letters and five intonation marks being used in the Vietnamese writing system. Besides the difference in phonology, language variation, grammar, the characteristic of writing system should have made Vietnamese handwriting different to English handwriting by using more scripts in writing. On the other hand, although like Korean and Japanese, in Vietnamese there is a lot of Chinese vocabulary and grammatical in- fluence, the current Vietnamese writing system uses no Chinese scripts but only Romanized styles. Handwriting skills are practiced every week in every grade of primary school. The models of letters and writing speed in primary education level are issued by the Ministry of Education and Training. For example, students at grade 1, 2, 3, 4, 5 are required to be able to write 30/15, 50/15, 60-70/15, 80-90/20, 100/20 words/minutes, respectively. Nguyen Thi Hanh [11] investigated the relations of handwriting errors and dictation rules and found that most errors were of following types: incorrect, omitted and misplaced letters. Additionally, Do Xuan Thao [3] reported about 60% of students at grade 1 who took the test made mistakes in their writings. 25% of them wrote the letter in incorrect size and/or form. The main types of errors were omitted stroke and extra stroke. However, to the best of our knowledge, it is still lack of quantitative informa- tion and objective studies about the development of Vietnamese primary students’ handwriting. The concrete analysis on the differences in handwriting with respect to gender and age of students is still omitted. 114 On the distribution of errors in transcription of primary school students in Vietnam In order to build up more understanding of handwriting ability of primary school students in Vietnam, in this study we investigated the distribution of their handwriting errors. More precisely, students from first to fifth grades in regular classes were asked to copy a sample paragraph in a fixed period of time. We collected the data concerning the errors in their products and then performed a statistical analysis. We predicted that the frequency of errors changed through five grades and the proportion of students who made errors changes from the one grade to another. Furthermore we conjectured that there would be a difference in making errors in the transcription between boys and girls. 2. Content 2.1. Methods 2.1.1. Participants There were 1557 students from grade 1 to 5 of the three primary schools participating in the study during February and March 2010. These schools were randomly selected among all the primary schools located in the centre of Hanoi and the surrounding area. Among 1557 students, six of them (four boys and two girls) did not properly participate as guided, so their performances were eventually removed from our sam- ples. The final data contained samples from 1551 students: 793 boys and 758 girls. There are about 150 boys and 150 girls in each grade. The concrete information about the number of participated students in each grade is shown in Table1. Table 1. The number of participants categorized by gender and grade Unit: person Grade 1 2 3 4 5 Total Male 163 159 151 138 182 793 Female 146 159 167 170 116 758 Total 309 318 318 308 298 1551 In addition, out of the 1551 students, there was only one left-handed student (0.06%). It is thought that in Vietnam, most left-handed children are trained to write with their right hand at home before attending school or at a very early stage in the first grade. 2.1.2. Sample paragraphs for transcription For each grade, we selected one paragraph from the corresponding textbooks Vietnamese Language, which was used in primary schools until 2003 (old pro- gramme), to make a sample for transcription. The fact that students did not have any memory about the contents of the selected paragraphs was taken into consider- 115 Nguyen Thi Cam Huong, Yusuke EDA and Nguyen Thi Hoang Yen ation. Besides, other aspects such as grammar, dictation rules and theme were also considered in order that the chosen texts were suitable for the knowledge of students who were studying the current programme. All sample paragraphs for transcription are shown in Table 2. Table 2. The topic, theme and source of model sample paragraphs and their length Grade Theme Source 1 Duoi anh trang Vietnamese Language Grade 1(II)* 2 Canh dong lua chin Vietnamese Language Grade 2(I)** 3 Duong len ban Hmong Vietnamese Language Grade 3(I) 4 Nhung canh buom ben bo song Vietnamese Language Grade 4 (I) 5 Buoi sang o thanh pho Ho Chi Minh Vietnamese Language Grade 5 (I) *: II is the text book used in the second semester **: I is the text book used in the first semester The letters in model paragraphs were printed in a similar font-style to the ones in the current textbooks. More precisely, the chosen font-styles were VnAvant for grade 1 and VnArial Narrow for all the others. All of the model paragraphs were printed in size 14 font and placed at the top of a lineless A4 paper. 2.1.3. Procedures Students were asked to copy a short model paragraph in two minutes as quickly as possible and without making any mistakes. Students wrote on standard papers which are being used in daily learning at school (the paper with vertical - horizontal 4 lines and 2.5mm × 2.5mm square). Students were allowed to use their most favorite pen (not pencils) while sitting at their usual seat in the classroom. They were asked to not use an eraser, if there was a mis-spelling they had to leave it as it is and write the correct letter next to it. 2.1.4. Measurement The 1551 samples of students were checked to find out the errors. Any sign (a letter, an intonation sign, a punctuation sign, a word, etc...) in the transcription of each student which is different to the sample is considered as one handwriting error. Furthermore, in order to count errors properly, we classified them into various types and sub-items (see Table 3). Errors in type 1 are any letters which had mistakes relating to their formation. In type 2, errors concerned misspelled words whose mistakes were related to letters. Errors in type 4 and 5 were mistakes in mark intonation and the punctuation marks. Thus, errors in type 1, 2, 4, 5 were counted by the number of mistakes in letters and signs. However in the transcriptions, 116 On the distribution of errors in transcription of primary school students in Vietnam students, for some reason, repeatedly, took out some of the words. The letters in these words were in correct formation, the spelling of these words was also correct but these letters made students’ copied sample different to the model one. These mistakes were counted for errors in type 3. Therefore when counting the errors in type 3, it was not necessary to be concerned about the number of letters but the number of words. The unit of errors in type 3 was not the same as the other types. Table 3. Types of error and sub-items of errors in transcription of students Types of error Sub-items of each type and their content Type 1 Errors on letter 1-1 Collapsed letter (letter’s shape is collapsed so it is hard to read or unreadable) 1-2 Mirror writing letters 1-3 Incorrect stroke (omitted or extra added stroke) Type 2 Errors on spelling 2-1 Extra added letter to word (repeated letter in word is included) 2-2 Omitted letter from word 2-3 Letters in the wrong order (letter is added to different position) Type 3 Errors on word 3-1 Different word 3-2 Word in wrong order (correct word but word’s position is wrong) 3-3 Omitted word 3-4 Extra added word (repeated word is included) Type 4 Errors on intonation marks 4-1 Different mark 4-2 Mark in wrong order 4-3 Omitted mark Type 5 Errors on punctuation marks 5-1 Different mark 5-2 Omitted mark 5-3 Extra added mark The quantitative aspects of errors were analyzed by Direct Probability Calcu- lation Methods and the Chi-square Test. Statistic software called JavaScript-Star version 5.5.0j (Original programming by Satoshi Tanaka) was used to handle those statistical processing. 117 Nguyen Thi Cam Huong, Yusuke EDA and Nguyen Thi Hoang Yen 2.2. Results and discussion The number of students who made errors compared to the numbers who did not make errors when copying a paragraph were categorized by gender and grade and are shown in Table 4. The number of errors is shown in Table 5. Besides, the number of students who made errors, medians number of their errors, ranges of errors were categorized by gender and grade are shown in Table 6. Table 4. The number of students who made error and the one who made no error M: Male; F: Female. Unit: person Grade 1 2 3 Gender M F Total M F Total M F Total Made error 74 55 129 69 47 116 69 65 134 Made no error 89 91 180 90 112 202 82 102 184 Total 163 146 309 159 159 318 151 167 318 Grade 4 5 Total Gender M F Total M F Total M F Total Made error 71 60 131 103 51 154 386 278 664 Made no error 67 110 177 79 65 144 407 480 887 Total 138 170 308 182 116 298 793 758 1551 Table 5. The number of errors categorized by gender and grade of students Grade 1 2 3 4 5 Total Error made by male 110 142 125 153 166 696 Error made by female 75 73 98 102 87 435 Total of errors 185 215 223 255 253 1101 Table 6. The frequency of errors on handwriting (count for one person) M: Male; F: Female N: number of students who made error. Grade 1 2 3 4 5 Gender M F M F M F M F M F N 74 55 69 47 69 65 71 60 103 51 Mdn 1 1 2 1 1 1 2 1 1 1 Range 1-4 1-5 1-7 1-6 1-7 1-8 1-11 1-5 1-8 1-13 118 On the distribution of errors in transcription of primary school students in Vietnam 2.2.1. Frequency of handwriting errors and the number of students who made errors In order to compare the number of students who made errors and that of the one who did not make errors, the data was analyzed by the Direct Probabil- ity Calculation Method. As a result, the number of students who made errors is statistically significantly less than the number of the ones who did not make errors with 99% accuracy (p = 0.0000 ** (p <.01)). It is that 42.81% of students made errors in their transcription. More concretely, the proportion of students in the first grade who made errors is 41.75%. This proportion is smaller than the one which was reported by Do Xuan Thao [3]. In order to recognize the difference among the number of students who made errors in grades, the data was analyzed by the Chi-square test and the results are statistical significantly with 99% accuracy (χ2(4) = 14.990, p< .01, Phi = 0.098). The residual analysis result (see Table 7) performed that among the all students who made errors, the numbers who made errors in grade 2 was significantly the lowest; in contrast, the number of ones who made errors in grade 5 was significantly the highest. This result meant that the number of students who made errors tends to increase though the grades. Table 7. The difference between the number of students who made error and the ones who made no error Unit: person Grade 1 2 3 4 5 Number of students who made error 129 116 2 134 131 154 4 Number of students who made no error 180 202 4 184 177 144 2 2: statistically few 4: statistically many Table 8. The gender discrimination of students who made error Unit: person Gender Male Female Number of students who made error 3864 2782 Number of students who made no error 4072 4804 2: statistically few 4: statistically many On the other hand, in order to find out whether the difference in the number of errors in five grades was statistically significant or random we analyzed the data by Chi-square test. It showed that the difference was statistically significant (χ2(4) 119 Nguyen Thi Cam Huong, Yusuke EDA and Nguyen Thi Hoang Yen = 14.946, p< .01) with 99% of accuracy and then a multiple comparison shows that the number of errors made by students in grade 1 was statistically significantly less than the one made by students in grade 4 and grade 5 with 99% of accuracy. This fact leads to a conclusion that the frequency of errors tends to incresase through five grades. This is not consistent with the previous research by Kono [10] which reported that the number of errors which occurred in transcription of primary students in Japan did not increase together with the grade. In addition, although the medians of the number of errors made by boys in second grade and fourth grade was both 2, the medians of other grades were 1; moreover, the ranges of errors were spread through one grade to the next (see Table 6). Those results mean that among the students who made errors most of them made only one error in two minutes, furthermore in the same period of writing time, the number of students who made many errors tended to increase with grades. The above findings matched our prediction about the change of the frequency of errors and the proportion of children who made errors. One likely explanation for why upper grade students tend to make more errors than the lower ones involves how frequently students practise their handwriting. Handwriting is not taught as a seperate subject from grade four onwards and time for practicing handwriting is less than before. It may also be assumed that the less students are instructed on how to write and modified their handwriting, the looser their attention to handwriting is, resulting in an increase in errors, although their handwriting skill becomes more habitual, organized and is available as a tool to facilitate the development of ideas. Siegel [12] demonstrated that working memory of students developed from the age of 6 through to 19 years of age, as a straight line and following a report by Igarashi and Kato [9], the capacity or efficiency of management of working memory relating to language rises rapidly by the age of 10. In our present study, we observed that most of the first grade students looked at words one by one and spelt out words while they were writing letters one by one carefully, however, the ones in upper grades such as grade 4 and 5 looked and remembered a group of words then wrote them down without spelling them out. Those representations of students expressed their development of working memory. This characteristic of cognitive processes in writing could explain why the students at lower grades wrote fewer letters but made less errors than the ones at higher grades. However, according to Hoang Van Thung [8], when making a transcription, if students spelt words while they were writing, their handwriting speed might be slowed down and easily lead to mistakes. He also suggested that based on reading quickly without analyzing words to spell, students have to perceive words with letters in sequence as a complete block of image. His theory and our above conclusion may bring up a hypothesis that there should be other aspects besides working memory and reading mechanism affecting students’ handwriting. One of those aspects should be dictation rules in Vietnamese which 120 On the distribution of errors in transcription of primary school students in Vietnam were reported that they made a lot of effect on handwriting errors [3,11]. Another aspect should be handwriting speed which were reported by Kono [10] that the more letters made, the more errors appeared in students’ handwriting. This implies that it is necessary to find out the relationship between the handwriting error and writing speeds in Vietnamese language. Additionally, in our study, broken-unreadable letters are considered as hand- writing errors. This kind of error may affect handwriting legibility. As our obser- vation, the unreadable letters which were made by students in grade 1 were mostly the letters with miss-strokes. This seems to be that the students still did not easily recognize forms of letters and/or how to write those letters so they wrote them with incorrect strokes. Meanwhile t