穨Review. Pdf

Table 4.7 Paired Samples Test of 5 Individual Grading Criteria of Control 
Group (N=30) 
Items 
Pre 
Post 
M 
SD 
t 
p 
Appropriateness 13.00 
13.80 
.80 
1.85 
1.66 
.11 
Vocabulary 
12.66 
13.06 
.40 
1.99 
.77 
.45 
Grammar 
12.00 
14.26 
2.26 
1.83 
4.75 
.00** 
Intelligibility 
12.60 
13.33 
.73 
1.66 
1.70 
.11 
Fluency 
13.60 
15.05 
1.46 
1.80 
3.14 
.00** 
** p < .01 
As a sharp contrast to the significant gains in all of the five items in the 
experimental group, as shown in Table 4.6, there were only two items that gained 
significantly in the control group. As Table 4.7 indicated, the control group 
progressed significantly in terms of grammar and fluency.
The Independent Samples Test was used to compare the inter-group difference in 

---

89
the scoring of the five items. The first comparison was on the result of the pre-test 
of oral performance between the two groups. 
Table 4.8 Independent Samples Test of Pre-Oral Task (N=60) 
Items 
Exp 
Control 
MD 
SD 
t 
p 
Appropriateness 13.33 
13.00 
.33 
.81 
.40 
.68 
Vocabulary 
14.60 
12.66 
1.93 
.85 
2.26 
.03 
Grammar 
13.66 
12.00 
1.66 
1.83 
.57 
.00** 
Intelligibility 
12.93 
12.60 
.33 
.77 
.43 
.66 
Fluency 
12.86 
13.60 
.73 
.87 
.83 
.41 
According to Table 4.8, there was only one item (grammar) that displayed 
statistical significance in the pre-test between the two groups. The mean score of the 
experimental group was 13.66 while that of the control group was 12.00 in their 
scores on grammar. The mean difference was 1.66, which was statistically 
significant because the p-value was as low as .00. The other four items did not show 
any significant difference between the two groups. 
The next section examined the inter-group difference in the post-test oral task.
As shown in Table 4.9, the experimental group outperformed the control group 
significantly on four items with the p-value lower than .05: (1) the appropriateness 
(p<. 01), (2) vocabulary (p <. 01), (3) grammar (p<.05), and (4) intelligibility (p<. 01).
The only item that did not show any statistical significance on the post-test oral task 
between the two groups was fluency (p=. 84).