Critical Reading Skills and Translation Ability of Thai EFL Students : Pragmatic , Syntactic , and Semantic Aspects

Translation ability requires many language skills to produce an accurate and complete text; however, one important skill, critical reading in the research, has been neglected. This research, therefore, employed the explanatory sequential mixed method to investigate the differences in Thai-English translation ability between students with a high level of critical reading skill and students with a low level of critical reading skill, and to examine the relationship between translation ability and critical reading skill. Moreover, translation error analysis, with four main aspects, pragmatic, syntactic, semantic, and miscellaneous errors, was also conducted to explore the dominant translation errors committed by both groups of students. To this end, the participants of the study were 60 English major students from a Thai university. The results of the study showed that there was a significant difference in Thai-English translation ability between the students with a high level of critical reading skill and the students with a low level of critical reading skill, and a significant correlation between translation ability and critical reading skill. In the translation error analysis, the semantic aspect, particularly the wrong use of words, was found to be the most frequent error committed by both groups of students.


Introduction
In daily cross-linguistic communication, people always do translations in both spoken and written forms.Nevertheless, a formal or important communication requires a more effective translation process and expertise that guarantee the equivalence and correctness of the meanings of the messages between the original language and the target language.Therefore, formal translations are mostly done by well-trained translators.In consequence, many countries offer translation courses in the university level.The variety of beginners' translation courses to advanced course is designed to train students to be able to do effective translations.Notwithstanding this, although some of the students have completed some translation courses, their performance is still far from the basic requirement of translation, i.e., correctness and equivalence of meanings and forms between the original texts and the translated version.An inadequate translation ability could affect the final product and distort the understanding of the receivers of the messages.
Owing to the fact that translation ability requires many language skills to work simultaneously to achieve tasks effectively, various studies have been conducted to find out what the relationship is between translation ability and many language factors.Some affective factors, such as reading habit, and writing interests were proposed as the variables that might be related to translation ability (Hernandani, 2015;Ridhwan, 2014).Another group of researchers focused their attention on cognitive factors such as critical thinking ability (Jahromi & Suzani, 2016), vocabulary mastery (Khotimah, 2015), writing skill (Hernandani, 2015), and reading comprehension skill (Kavaliauskiene & Kaminskiene, 2009;Mashhadi, 2008;Rahemi & Jufri, 2013;Widiasari, Supraman, & Sudirman, 2015).
According to the studies noted above, translation error analysis might not be a new area of the study of second/ foreign language learning and teaching; however, it is still worth examining this classic problem, but with a fresh perspective.The present study was conducted with two distinctive aims.First, the focal point of the present study is the relationship between the critical reading skill (CR) and the translation ability of Thai university students because this relationship has not been studied before.Second, the aforementioned studies were mostly conducted using a sole research method that was either quantitative or qualitative.The present study utilized the explanatory sequential mixed method to illustrate the translation ability of English learners in the Thai context.Therefore, the present study was guided by the following research questions: 1).Are there any differences in translation ability between students with a high level of critical reading skill (the high CR students) and students with a low level of critical reading skill (the low CR students)?
2).Is there any relationship between the critical reading skill and the translation ability of a group of Thai university students?
3).What kind of translation errors occurred in students' translation from Thai to English? 4).What are the dominant errors made by the high CR students and the low CR students?

Critical Reading and Translation
There have been many attempts over several decades to investigate what makes a successful translation.Some scholars believe that cognitive factors make a strong contribution to developing translation ability.Therefore, the relevant studies focused on reading, writing skills, and vocabulary mastery.Most studies focused more on reading comprehension and investigated the correlation between reading comprehension and translation ability (Dwi, 2014;Kavaliauskiene & Kaminskiene, 2009;Mashhadi, 2008;Pham, 2017;Rahemi & Jurif, 2013;Widiasari, Supraman, & Sudirman, 2015).However, Nord (1997) argued that scholars should emphasize the pragmatic elements of translation as well as syntactic and semantic aspects.To understand the pragmatic sense of the messages, students require more than reading comprehension skill.Critical reading enables students to understand not only the literal meaning but the word choice and style of language (Wallace, 2003).Therefore, the underlying or between-the-line meanings need more advanced reading skill that is critical reading.

Translation Error Analysis
Earlier studies using translation error analysis relied on the general error analysis (EA) framework as an effective tool to evaluate translation products and then improve the teaching and learning process of translation.Focusing on translation studies, another EA framework that has been widely applied for translation error analysis is derived from the framework of Corder (1973).He argued that errors in second or foreign language production might not only be caused by interlingual transfer and interference of L1.Therefore, he proposed another framework, which is a systematic description and classification of errors in L2 utterances and the reconstructed language.In addition, he believed that second or foreign language errors can be classified into four categories: omission, selection, addition and ordering of some language elements.Various researchers of translation errors analysis conducted their studies based on the Corder's EA framework (Farrokh, 2011;Komariah, 2013;Na, 2005;Presada & Badea, 2014).In addition, most of these studies analyzed the translation ability of their participants from a holistic viewpoint, utilizing the total score for the translation.Using the total score could only reveal the relationship between translation ability and cognitive language skills with a holistic overview, but it could not specify the problematic areas of translation.
In the Thai context, most of the researchers who conducted studies also relied on the EA framework.Sattayatham and Honsa (2007) studied the most frequent errors of first-year students' Thai-English translation at the sentence level.They focused on the syntactic and lexical levels.Pojprasat (2007) studied the translation errors made by Grade 12 students, analyzing 20 students' translation works, each paper consisting of 30 English-Thai sentences and 30 Thai-English sentences.His translation error analysis focuses on two aspects: syntactic and semantic errors.With the same focus, Supparajyothin (2010) conducted error analysis of translations done by Thai college students; moreover, she also studied the translation ability of 50 English major students at the university level.Moreover, the correlation between the Thai-English translation score and English-Thai translation score was investigated.Later Pornwiriyakit and Dandee (2015) carried out a study that aimed to identify errors in translation from English to Thai of some university students.Overall, the recent studies of Thai EFL learners have only paid attention to syntactic and semantic aspects, while the pragmatic aspect is still neglected.Although Zhang (2016) highlighted that pragmatic errors should be considered to be the highest level of translation errors, there have only been a few studies of translation errors that focused on the pragmatic aspect.One of those studies was done by Ardenshir and Zarafshan (2014).They analyzed the students' English-Persian and Persian-English translation works by utilizing Na Pham's (2005) Error Analysis Model which consists of pragmatic implication: pragmatic errors, inaccurate rendition of lexical items, distortion or change at the level of the meaning of the source text, literal translation, free translation, wrong lexical choice and incorrect focus.The results indicated that understanding the pragmatic senses was the most frequent problems of their students.

Translation Error Analysis Framework of the Study
Development of the translation error analysis framework used in this study was based on Liao's framework (2010), which focused on three types of errors that is rendition, language errors and miscellaneous errors.The first category included the errors that occurred when translation failed to deliver an accurate meaning from the original messages.Language errors referred to the problems in using grammar aspects of the target language.The third category, miscellaneous errors (omission), was counted when some parts of the original text were missing in the translation.
However, the present study re-classified Liao's language errors category by dividing these types of errors into two separate aspects: syntactic and semantic.The syntactic aspect only focused on the grammatical errors, whereas the semantic aspect paid more attention to the meanings at the word level.Therefore, the framework of the translation error analysis for the present study consisted of four types of errors, as follows:  Table 1 shows that there was no error type in the syntactic aspect of the translation.This was due to the intention of the researcher, who did not objectively pre-determine the fixed criteria when analyzing the students' translation errors for grammatical errors.This syntactic coding would then be developed from the raw data of the present study, and would be able to capture key errors that occurred specifically in this study context.

Method and Design
The present study utilized the explanatory sequential mixed method to illustrate the level of translation ability of English learners in a Thai university.The two-phase study started by collecting quantitative data to investigate the difference in translation ability between high CR and low CR students and the relationship between critical reading skill and translation ability.The quantitative results were used to plan the qualitative study done to follow up the findings.Consequently, the objective of the qualitative phase was to explore the translation errors committed by the students in more depth.

Participants
The participants of the present study were 60 English-major students enrolled in Thai-English translation and critical reading courses at a Thai university.

Translation Texts
For the qualitative data analysis, the 60 students' translation papers were divided into three groups according to their critical reading scores: those with a level of high critical reading skill, those with a moderate level of critical reading skill, and those with a low level of critical reading skill.After that, 40 papers were purposively selected, 20 from a group with a high CR skill and another 20 from the group with a low CR skill.This Thai -English translation paper consisted of 25 items with the length of each item varying from a single sentence to a short paragraph.

Instrument
The translation error classification model adapted from Liao's (2010) model was employed for classifying and exploring the dominant errors committed by the students.

Procedures for Collecting Quantitative Data
The researcher started collecting the data by asking for the lecturers' permission to have access to the scores of a Thai-English translation course and their exam papers, and asked for the same access to scores and exam papers of the English critical reading course.The translation scores were derived from the in-class exam paper, in which the students were asked to translate Thai sentences into English sentences, and they were allowed to use an off-line dictionary during the exam.The critical reading scores were also from the in-class exam papers, in the forms of short-answer questions and short-paragraph writing.

Procedures for Collecting Qualitative Data
After receiving the translation papers, the researcher and the second rater (one of the researcher's colleagues who has expertise in teaching both Thai-English and English-Thai translation modules) had a discussion about the overview of the translation error analysis process.The researcher explained the objectives of the study and the overview of translation error analysis to the second rater.Then she created the coding without seeing the initial categories.When both the researcher and the second rater had finished coding the papers, the comparison was conducted and the ambiguity of the criteria was revised.Subsequently, the researcher and the second rater did the coding separately.

Data Analysis
The translation and critical reading scores were processed by using the Statistical Package for Social Science (SPSS) for the descriptive data of the participants' scores, followed by an independent-t-test analysis to compare the means scores of the high CR students and the low CR students.The total number of participant was 60, and these were divided into three groups according to their level of critical reading proficiency: there were 20 students in the high CR group, 20 students in the moderate CR group and 20 students in the low CR group.Then only 40 students' scores from high CR and low CR groups were selected and analyzed.The Pearson correlation coefficient was run to investigate the relationship between translation ability and critical reading skill.Finally, after diving the papers into two groups (high CR group and low CR group), the papers were analyzed by means of a translation error classification model adapted from Liao's (2010).

Trustworthiness
The procedures that were used to assess the trustworthiness of the data analysis included inter-rater reliability check and coding consistency.
The translation papers were scored by two lecturers teaching the translation courses with the agreed rubric, then an Intra-class correlation coefficient (ICC) with a two-way random effects model was used to assess the inter-rater reliability.It was found that the ICC of the translation test at the 95% confidence interval was 0.98.This meant the inter-rater reliability of the translation test was high, which indicated that there was almost agreement between the two raters (Koo & Li, 2016).
The in-class critical reading scores had been graded by the lecturer and then were marked again by the researcher in order to increase the reliability of the scores, and the ICC was calculated for the degree of agreement.However, before scoring the papers, the researcher consulted with the lecturer teaching the critical reading class regarding the scoring criteria and employed the same rubric for scoring.After that, the inter-coder reliability was checked.The ICC of the critical reading test at the 95% confidence interval was 0.96, and this showed that there was strong agreement between the two raters.
The independent parallel coding was conducted to achieve coding consistency of the translation error analysis.Moreover, the ICC was calculated to measure the degree of agreement between the coders.The ICC of the translation error analysis at the 95% confidence interval was 0.85, and this indicated a substantial agreement between the two coders.As shown in Table 3, the critical reading scores of the high CR students were significantly higher than those the students with the low CR students.Moreover, the means score of the total scores of the high CR students was significantly higher than the mean score of the low CR students.

The Findings
It was found that among the three aspects of translation scores, the most significant difference in the translation scores of both groups was in the word-choice aspect (high CR = 32.71,low CR = 28.91,t = 4.96).Furthermore, the mean score of the overall meaning of the high CR students was 38.80 (S.D. = 3.36), and the mean score of the low CR students was 35.85, (S.D. = 1.90).However, the mean score of grammar aspect was the least significantly different aspects in both groups of students.The mean score for the grammar of the high CR students (mean = 26.02,S.D. = 4.66) was significantly higher than the mean score of the low CR students (mean = 22.19, S.D. = 4.06).As shown in Table 4, overall there were significant positive correlations between critical reading and translation ability in every aspect, ranging between .504 and .654,and among three aspects of the translation score, the correlation was highest between the critical reading skill score and the word-choice score, r = .631.

Translation Error Analysis Results
The errors were classified into four main categories: pragmatic (rendition), syntactic (grammar), semantic (vocabulary and word choice) and miscellaneous (omission).Table 5 indicates that the most frequent errors that occurred in the students' Thai-English translations were of the semantic type, followed by syntactic and pragmatic ones.The least frequent errors found in the students' translations were of the miscellaneous type.Table 6 shows, overall there was a smaller number of errors made by the high CR students (1702 errors) than the numbers of errors made by the low CR students (1886 errors).The high CR students committed fewer errors in nearly all aspects.The greatest difference between the numbers of errors committed by the high CR students and those made by the low CR students was found for "inappropriate word choice" in the semantic category (high CR = 398 and low CR = 466).
However, the most interesting findings were that regarding semantic errors, the low CR students (138 errors) could use prepositions more correctly than the high CR students (166 errors).This was similar to the use of inappropriate register, in relation to which the low CR students committed lower numbers of error (32 errors) than those of the high CR students (19 errors).
In addition, the big differences of the numbers of errors committed by both groups of students were also found in the following error types: misinterpreting the source text, excessive rending, which differentiates the translation from the original text, subtle difference in meaning between the source and target texts; insufficient accuracy, excessive literal translation, which leads to ambiguous translation, part of speech, pronoun, articles, and missing translation.

Pragmatic Errors
According to the frequencies and percentages of Thai-English translation errors committed by high CR students and the low CR students, the most dominant errors found in the pragmatic category was excessive literal translation, which leads to ambiguous translation (R6).

Example
Source text: อุ ณหภู มิ อาจจะสู งกว าปกติ ราวๆ ปลายเดื อน Translation: Temperatures will probably be/rise above average by the end of the month.
Student's translation: The temperatures may rise more than the normal around the end of the month.
This example revealed that the error occurred because the students translated the sentences word-for-word regardless of the meanings in English.This led to the source text being translated into the target text in a form that is ambiguous.

Syntactic Errors
Of the syntactic errors, tenses was the one most commonly made, and both groups of students committed the same numbers of errors with very high frequencies.In addition, the errors found in the tenses aspect could be classified into three main areas: simple, continuous, and perfect.When all errors concerning the tenses aspect were considered, it was indicated that the simple aspect was the most difficult element for the students when translating texts, followed by the perfect aspect and the continuous aspect.Translation: Many international companies are now asking all their staff/ employees to communicate only in English.

Student's translation: Nowadays, many multinational companies ask employees to communicate only in English
According to the three examples above, it could be argued that adverb of time might be one cause of the problem.
Because the students misunderstood the meaning of time in the words "annually", "past" and "now" in the original text, the meanings of time in their translations were wrong.

Semantic Errors
Nearly half of all the errors found in this study were from the semantic category.The high CR students and the low CR students committed these errors with the highest frequency.However, most of the semantic errors made in this study did not affect the main messages when the translations were done, and serious distortion of meaning rarely happened.Furthermore, although most of the cases were considered to be errors, some cases might have been just a slip of the pen.Misspelling (W1) could have been that type of the mistake, and this was found in the translations of both groups of students, even though they were allowed to use their off-line dictionary during the translation process.
The appropriateness of word choice (W2) was found to be the major issue in the translations in the present study.This type of error was found in the translations of every student.Below is an example of a student's translation, which represented various types of errors from the semantic category such as misspelling (W1), inappropriate word choice (W2) and inappropriate register (W5) errors.Student's translation: The pearl of the Andaman sea also known as Phuket province is thailand's largest island and one of the most popular tourist attractions.It is widely well-know in terms of the wonder terrain, the identity of architecture and numerous local culture.

Example
In this case, misspellings were found in two places, "thailand" and "well-know," whereas inappropriate word choice errors were found in many places.The words the students chose did not match perfectly with the original text.For example, the word "numerous", indicates a great number of something but does not suggest the vast differences within that thing; therefore, the word "diverse" was more appropriate in order to describe the differences between the native cultures noted in the original text.
The inappropriate register (W5) error differed slightly from the inappropriate word choice error.It could be said that this error type focused on the perfect transfer of meaning from the original to the translated texts.For instance, the word "terrain" in the above example did not seem to fit with the word "ภู มิ ประเทศ" in this context.

ภู มิ ประเทศ ) ที ่
งดงาม( in the sentence represented the description of the landscape of Phuket, and the focus was on the beauty of the environment of the area.The word "terrain," however, is normally used to describe the physical features of an area of land.In contrast, the word "scenery" could have been used instead to describe the generally beautiful appearance of the area. Preposition (W3) and article (W4) were also important issues in this study.These two types of errors were classified as being in the semantic group because they would affect not only the grammatical sense but also the meaning of the translated text.Errors involving these two features were also committed with very high frequencies by both groups of the students.However, these errors could be divided into three points.The first is the wrong use of prepositions and articles, as shown in the following example.

Example
Source text: ในระยะเวลา 3 ป ที ่ ผ านมา รั ฐบาลได เปลี ่ ยนนโยบายการคลั งมาหลายครั ้ งแล ว Translation: In the past three years, the government has changed its fiscal policy several times Student's translation: For three years ago, the government has changed a fiscal policy several times.
The second and the third points of the preposition and article errors were the overuse of preposition and articles and the omission of prepositions and articles.
Student's translation: The one guest of the hotel who stay on the second floor said he heard something fell down from the top floor on Saturday morning.

Student's translation: A guest of hotel who stay second floor say that she heard something fall from the top floor on Saturday morning
The words "the" and "on" in the first student's translation and the word "of" in the second student's translation indicated the overuse of prepositions and articles.In contrast, the underlined words in the second student's translation showed where there was a missing article and a missing preposition.

Miscellaneous Error
This type of error refers to where some part of the original text is missing in the translation.However, no serious omissions were found in this study.Most of cases involved the omission of non-essential descriptive words or modal words.
Student's translation: Some students of Duangduen's class aren't enthusiastic.
In this case, the overall meaning of the translation was not distorted but the student was unable to render all the meaning of the original text.

Comparison of the Translation Errors Committed by the Students With High and Low Critical Reading Skill
The results of the independent t-test independent analysis revealed that overall the high CR students committed lower numbers of errors than the low CR students.Moreover, in all aspects of translation, that is, overall meaning, word choice, and grammar, big differences were found between both groups.In addition, it was found that the most significant difference in the translation scores of both groups was in the word-choice aspect.
Comparisons between a high level and a low level of language proficiency, i.e. critical reading skill, in relation to translation ability have rarely been conducted, so the findings of the present study indicated a new area of study.

Relationship Between Critical Reading and Translation Ability
Although the present study focused on the relationship between the translation ability and the critical reading skill of a group of Thai university students, the findings of this study agreed with the results of other studies on the correlation between translation ability and reading comprehension skill in EFL contexts.Therefore, critical reading skill might be considered as one of the factors that relate to translation ability, just as general reading comprehension does.
Furthermore, the results of the present study were in accordance with studies mentioned in the literature review section, in that it revealed a moderate positive correlation between overall translation ability and critical reading skill (r = .65,p.01).However, the earlier studies found a wide range of degree of relationship between translation ability and reading comprehension skill (r = .30-.91, at least ps.05) (Aminafshar, 2016;Dwi, 2014;Jahromi & Suzani, 2016;Kavaliauskiene & Kaminskiene, 2009;Khanmohammad & Kehtari, 2015;Mashhadi, 2008;Pham, 2017;Rahemi & Jurif, 2013;Tavakoli, Shafiei & Hatam, 2012;Widiasari et al., 2015).This big difference in degrees of correlation between translation ability and either reading comprehension or critical reading skill could indicate some concerns.First, the participants of the studies varied from high school students to university students, and even the English major students who were well-trained in translation skills.Furthermore, although the aforementioned studies all investigated translation ability regarding the target language and English, the translation scores were different.This might have been due to the different source language (different L2 as the original language).Moreover, most of the studies employed a multiple-choice test for the reading comprehension measurement while the present study derived the scores from short-answer questions and short-paragraph translation tests.Tavakoli et al. (2012) suggested that an open-ended translation test is more reliable and valid than multiple-choice one.Furthermore, most studies assessed the translation ability using the holistic scores, while the total translation scores of the present study was compiled from three parts, overall meaning, word choice, and grammar, in order gain a deeper understanding of translation problems.Hence, these proposed issues might affect the degree of relationship between the studied variables.

Translation Error Analysis
According to the translation error analysis, the most dominant types of errors that occurred in the students' Thai-English translation were from the semantic category.These types of errors dominated nearly half of the total errors found in the students' papers.The findings of the present study were in accordance with many previous studies (Cuc, 2018;Farraokh, 2011;Jajanshani & Kafipour, 2015), which all confirmed that the wrong selection of words was the most frequent errors of EFL students.
Beside from the errors of inappropriate word-choice, the wrong use of prepositions was also found as one of the major errors frequently committed in the present study.This result also shared some similar findings with some of the previous translation error analysis of EFL students (Cuc, 2018;Mohaghegh, Mahmoudi, & Mohammad, 2011;Popescu, 2013;Supparajyothin, 2010).However, the results of the present studies need to be studied further in order to confirm the similarities in more detail.However, the causes of using the wrong preposition in the present study is still unclear since although the students were allowed to use an off-line dictionary while they translated the texts, the problem still occurred.Possible answers might be that the errors occurred because the students were unaware of contexts when selecting the appropriate prepositions or they might have lacked the knowledge about some aspects of collocations, i.e., "adjective + preposition" or "verb + preposition."Another possible cause could be in accordance with the study of Suranakkharin (2017).He found that collocation competence was one of the problematic learning areas of Thai university students.
In addition, the wrong use of articles was also one of the main problems in the translations.The present study found that the students in both groups used articles incorrectly a lot in their translations.This result supported the findings of Tafazoli, Golshan and Piri's study (2013) and that of Sharif and Hassani (2016), which found that the wrong use of articles was the most noticeable error.Focusing on the Thai EFL context, Sattayatham and Honsa (2007) also found that using an article incorrectly was one of the major sources of frequent errors made by Thai university students.
Moving to the syntactic error category, the results of the study also agreed with the research that used translation error analysis, especially research with its main focus on grammatical or syntactic errors, that the incorrect use of tenses was the most frequently occurring y.The early studies of translation error analysis also indicated that tenses were the dominant source of errors in translation (Farrokh, 2011;Mohaghegh et al., 2011;Pornwiriyakit & Dandee, 2015;Presada & Badea, 2014).
Although these studies mutually reported the misuse of tense in translation as the major source of the syntactic errors, Presada and Badea (2014) gave further details by adding that this problem resulted from the misuse of the continuous aspect and the confusion between the present perfect and past tenses.The findings of Presada and Badea agreed with the present study's results, in which the main problem concerning the tenses occurred in the translation of the simple aspect, i.e., the present simple and the past simple.The confusion is mostly found in the translation of, firstly, the present simple and the present continuous and, secondly, the past simple and the present perfect.
Interestingly, these findings still could not claim that the wrong use of tenses was the most frequent source of syntactic errors because some researchers conducted translation error analysis and discovered that although tenses might be one of the most frequent errors committed by the participants, they were not the most significant errors.The study of Supparajyothin (2010) found that errors in the translation of the passive voice and serial verbs were the most dominant error types in the participants' translation, while wrong use of tenses was found the least.
Regarding errors of the pragmatic types, which refers to the differences in meaning between the source texts and the translated version, ranging from significantly misinterpreting the original message to minor distortion of the translation.The most frequently made errors concerning the pragmatic aspect arose from excessive literal translation.Besides noting excessive literal translation, the present study also discovered that subtle difference in meaning was the dominant source of errors committed by the participants.These findings shared some similarity with the study of Jahanshani and Kafipour (2015), which found that subtle difference in meaning was the most dominant source of errors, followed by excessive rendering, which differentiate the translation from the original text.Moreover, only small numbers of insufficient rendering errors were found the last two studies mentioned.
Concerning overviews of the errors committed by EFL students, Khdabandeh (2007) focused his study on Persian graduate students and discovered that the participants' chief difficulties in Persian to English translation were grammatical errors, followed by discoursal and lexcical types.However, Farrokh (2011) found that semantic errors (word-choice) were the most dominant errors.Therefore, in accordance with the finding mentioned earlier, there was no agreement on the type of error that was the most frequently committed by the EFL students who translated their native language text source into an English version.One possible explanation could be the different error categories.For instance, some studies classified articles and prepositions in the grammatical type of error, while some studies, including the present study, classified articles and prepositions as being in the semantic aspect.Furthermore, there was no consensus on the number of categories for translation error analysis in each study.Some studies focused on one or two aspects, i.e., grammatical or grammatical and word errors.
Although semantic errors was found to be the most frequent error in the present study, it was unclear whether pragmatic, syntactic or sematic errors were the most frequent errors committed by EFL students in native (L1) to English (L2) translation of EFL students.This might be due to the difference in the classification of errors types between the earlier studies and the present study.Nevertheless, the present study could confirm some of the same results as the aforementioned studies and highlighted some areas of errors such as word choice, tenses, prepositions, and articles to the teachers of translation courses as well as to the researchers in the translation error analysis field.It also emphasized the importance of pragmatic such as 'excessive literal translation' and 'subtle difference of meaning between the source and target texts' as the areas of interest to the field of translation error analysis.

Conclusion
Overall, the results of the study indicated a significant difference in translation ability between the students with high CR and the low CR students and the positive relationship between critical reading skill and translation ability.
Translation error analysis revealed that the dominant type of errors that occurred in the students' Thai-English translation were from the semantic category, particularly the errors involving inappropriate word-choice.However, the big differences between the two groups of students was found in the pragmatic aspect, i.e., misinterpreting the source text, "excessive rending, which differentiates the translation from the original text, and excessive literal translation, which leads to ambiguous translation.These findings made by using translation error analysis confirmed the results from the quantitative phase, which were that the students with a higher level of critical reading skill tended to have higher translation scores for translations of Thai texts to English versions because they not only understood the literal meaning of the text, but they could go along, between and beyond the lines of the texts and consequently did better in translation.
Pragmatic errors (rendition) R1: Misinterpreting the source text R2: Insufficient rendering, which differentiates the translation from the original text R3: Excessive rendering, which differentiates the translation from the original text R4: Subtle difference in meaning between the source and target texts; insufficient accuracy R5: Misinterpretation due to unawareness of terms R6:Excessive literal translation, which leads to ambiguous translation R7parts in the target text; omission.
3.2.1The Differences in Translation Ability Between the High CR Students and the Low CR StudentsThe descriptive results for the critical reading and translation scores are summarized in rendering, which differentiates the translation from the original text R3: Excessive rendering, which differentiates the translation from the original text R4: Subtle difference of meaning between the source and target texts; insufficient accuracy

Table 1 .
Classification of error types

Table 2 .
Table 2 below.Descriptive statistics for the critical reading and translation scores Table2shows that the mean score for critical reading was 52.35 out of 70.The highest score was 65.00 and the lowest was 30.00.For the translation, the mean score was 91.83 out of 150.The highest score was 116.00 while the lowest was 75.00.

Table 3 .
Comparison of translation ability between the high CR students and the low CR students

Table 4 .
Correlation coefficient between critical reading skill and translation ability

Table 5 .
Overall frequency of errors in the students' Thai-English translation

Table 6 .
Frequencies and percentages of Thai-English translation errors made by the high and the low CR students Pearl of the Andaman Sea" or PhuketIsland is the largest island in Thailand and one of the most popular tourist destinations.It is known throughout the world for its picturesque scenery, unique architectural settings and diverse native cultures.