Students’ performance of English Writing for Business

The business writing performance measured by the online Criterion and CPIDR is respectively given in the Table 1. The results showed that the means of all the positive items for post-writing of the three target tasks such as writing score, idea, word and P-density, were better than those for prewriting. But, the mean frequency of writing errors for post-writing (.196) was smaller than that for the prewriting (.212). A paired sample t-test analysis showed that, except for P-density, the difference of the means for all the other mentioned items between the prewriting and the post-writing was significant.

Table 1 Results of business writing performance by online Criterion and CPIDR evaluation.

Mean of writing performance

Criterion CPIDR

score writing error per word idea word P-density Pre-task 1.92 0.212 50.2 102.4 0.4896 Post-task 2.52* 0.196** 70.4** 143.1** 0.4923

*: p<.05 and **: p<.01, significant difference between the pre-task and the post-task

A Pearson correlation analysis indicated that students with higher pre- and post-writing scores respectively made significantly fewer errors in their corresponding writing tasks

Shu-Chiao Tsai EFL Business Writing with Task-based Learning Approach:

A Case Study of Student Strategies to Overcome Difficulties

(r=-.552 and p=.000 for prewriting; r=-.390 and p=.003 for post-writing). In addition, although students’ P-density in the post-writing was greater than that in the pre-writing, there was no significant difference in P-density between the pre- and post-writing, a CPIDR indicator of writing quality. In general, writing requires conscious mental effort and is considered as a challenging language skill so that it is not easy for EFL students to significantly improve their P-density after only 12-weeks period of study.

Table 2 Results of business writing performance by vocabulary profiler (VP) Writing

performance

Word count of students’ writing from VP analysis

K1 word K2 word AWL word Off-list word

Pre-task 91.3 6.0 3.0 7.7

Post-task 124.6 7.0* 4.4** 8.9**

*: p<.05 and **: p<.01, significant difference between the pre-task and the post-task

The VP analysis indicated that students wrote more words in the four categories in the business post-writing, as shown in Table 2. A paired sample t-test analysis showed although there was no significant difference in the word count of the lowest VP English level, the word counts in the other three categories of the higher English level like K2, AWL and off-list words in the post-writing were significantly greater (p<.05) than those in the prewriting. Theses results showed that students wrote more words of higher levels in their post-writing. A further Pearson correlation analysis was conducted to investigate the relation among students’ business writing scores and all the items analyzed by CPIDR and VP, as shown in Table 3. The significantly positive correlation between the prewriting scores and the post-writing scores suggested that students who wrote better in prewriting still kept a better performance in post-writing in a significant way. Meanwhile, there was a significant positive correlation between students’ Criterion post-writing scores with two CPIDR items of idea and word, and all the four VP categories, suggesting that these three types of online evaluation were significantly correlated. These results indicated that based on the features of individual online assessment, EFL business writing performance in this study was simultaneously measured by the three different types of online assessments for a coherent and objective analysis. Of course, such an integrated writing analysis has to be further studied and verified in other professional or academic fields in order to find a more objective computational assessment of writing texts and avoid the fallibility of manual counts and the subjectivity of human raters’ intuitive judgments.

高應科大人文社會科學學報 ISSN 1815-0373

Table 3 Correlation of students’ post-writing scores with parameters analyzed by CPIDR and VP Pearson

CPIDR items VP categories

idea word P-density K1 K2 AWL Off-List

*: p<.05 and **: p<.01, significant correlation between the post-writing score and other parameters

In order to understand how the model texts influences EFL students’ post-writing, a software program was developed to compare and investigate two types of the words in students’ pre- and post-writing for the three tasks: (1) the additional words means the words that were only presented in the students’ post-writing, not in the pre-writing; and (2) the borrowed words means the additional words that were borrowed from the model texts. The comparison run by the program indicated that the mean counts for these two types of words respectively were 65.7 for the additional words (accounting for 46.9% of the words in the post-writing) and 26.5 for the borrowed words (accounting for 19.2% of the words in the post-writing). A Pearson analysis was conducted to study the correlation of students’

post-writing scores with their additional word count and their borrowed word rate which is the count of the borrowed words divided by the total words count in students’ post-writing.

The result indicated that students with higher post-writing scores significantly expressed more additional words in the post-writing, and meanwhile had fewer possibilities to borrow words from the model texts, as show in Table 4.

A further Pearson correlation analysis was also conducted to investigate the relation between students’ writing scores measured by Criterion and their English proficiency determined by the TOEIC-like test. The results showed that students’ English proficiency had a significantly positive correlation with their business prewriting scores (r=.315, p=.019<.05) and post-writing scores (r=.271, p=.045<.05), meaning that students’ English proficiency was an important intervening variable that affected business writing performance in this study.

Table 4 Pearson analysis between students’ post-writing score and their additional and borrowed words Pearson correlation

analysis

Parameters in the post-writing for three tasks Added word Borrowed word rate Post-writing

score

r .313(*) - .608(**)

p .020 .000

*: p<.05 and **: p<.01, significant difference between the pre-task and the post-task

Shu-Chiao Tsai EFL Business Writing with Task-based Learning Approach:

A Case Study of Student Strategies to Overcome Difficulties