The analysis of the test scores of the two groups was to answer this sub-question:
1. To what extent does the use of metacognitive and cognitive strategies affect students’
reading performance?
As mentioned in the previous chapter, the analysis of the scores was conducted at two stages in order to answer the above research sub- question.
4.2.1 Stage 1: before the treatment
At this stage, the mean scores of the pretests taken by both experimental and control groups were compared to detect students’ levels of reading comprehension. As for a study with small number of participants, non-parametric statistic tests are appropriate as they do not rely too much on the normal distribution (Larson-Hall, 2010).
53
As can be seen from the table below, the minimum of the pretest scores of the experimental group is 4.57 while it is 5.14 with respect to the control group. As for the maximum score of the pre- test, both the two groups reached the maximum score of 8.57. Also from the table, the mean of pretest for experimental group (6.75) is slightly higher than the mean score of the control group (6.51). Thus, it can be seen from the table that the reading levels of the two groups do not vary much. However, a more powerful test should be used to confirm this.
Table 4.1: The means of pretest score of control and experimental group
An independent-samples t-test was used to investigate levels of the two groups in terms of reading performance before the treatment.
The table below highlights that the means of two groups did not vary as sigma = 0.358 which is higher than p value (p=0.05), the null hypothesis that there is no difference in means of the pre-test score could be rejected. Thus, it can be inferred from non-parametric t-test that the means of the two groups are the same which leads the conclusion that the levels of the two groups before the treatment did not vary much. This conclusion lays a firm bridge for the comparison after the treatment.
Test Statisticsa
Pretest
Mann-Whitney U 117.500
Wilcoxon W 288.500
Z -.919
Asymp. Sig. (2-tailed) .358 Exact Sig. [2*(1-tailed
Sig.)] .365b
a. Grouping Variable: Class b. Not corrected for ties.
Table 4 2: the non-parametric t-test of pretest scores
Count Mean Minimum Maximum Column N % Class Experimental Pretest
16 6.75 4.57 8.57 47.1%
Control Pretest 18 6.51 5.14 8.57 52.9%
54 4.2.2 Stage 2: after the treatment
At the second stage, the mean scores of the posttests of two groups were compared to identify the differences between the two groups.
The table below shows the descriptive score of the posttest for experimental and control group. Both the minimum and maximum score of the experimental group are higher than control group. Also the mean score of the posttest of the experimental group is higher than the mean score of the control group. It can be inferred from the table above that there is a difference between the two groups in terms of the score of the post test. As for the same posttest, the students in the experimental group performed better and achieves the means score of 7.34 while the members of the control group only completed the posttest with the means of 6.38. The difference of the post test score between the two groups is 0.96 (7.34 - 6, 38 = 0.96) which accounts 13 percent of the experimental group’s mean (0.96 / 7.34) and 15 percent of the control group’s mean. In addition, it is worth noting that the experimental group exceled the control group at both the minimum and maximum score of the post test, 5.12 and 9.71 respectively.
Mean Minimum Maximum Median
Standard Deviation
Class Experimental Posttest 7.34 5.12 9.71 7.36 1.44
Control Posttest 6.38 4.85 8.28 6.21 1.01
Table 4. 3 : description of the posttest scores
To find out whether there is statistical difference in the scores of the post test, a nonparametric t-test was used to examine this. The result is displayed below.
The result (sigma) is 0.043 which is lower than p-value set at 0.05. This mean the null hypothesis that there is no difference of the mean scores between experimental and control group can be rejected. The difference is a statistical significance.
55 Test Statisticsa
Posttest
Mann-Whitney U 85.500
Wilcoxon W 256.500
Z -2.024
Asymp. Sig. (2-tailed) .043 Exact Sig. [2*(1-tailed
Sig.)] .042b
a. Grouping Variable: Class b. Not corrected for ties.
Table 4.4: the non-parametric t-test of the posttest scores
Next the mean scores of the pretest and posttest of the experimental group were compared to find out if there was a statistical difference. A non-parametric dependent sample t- test was conducted to investigate this issue.
The difference in students’ reading performance before and after the treatment in the experimental groups was tested with a nonparametric paired-samples t-test. A Mann-Whitney U test was used because the data of this the thesis did not meet the requirements of a parametric test. The p-value was set at 0.05 in order to accept or reject the null hypothesis. The table below displayed the result. The null hypothesis that there is no difference in students’ reading performance could be rejected as there is significant difference in means (sig=0.007< 0.05).
Thus, it could be concluded that students’ reading performance did improve after the strategy training course.
Test Statisticsa
Posttest - Pretest
Z -2.692b
Asymp. Sig. (2-
tailed) .007
a. Wilcoxon Signed Ranks Test b. Based on negative ranks.
Table 4 5: the non-parametric dependent sample t-test of the experimental group
56
In addition, a nonparametric paired-samples t-test was also conducted to see if students’
reading performance after the course and their performance before differed or not. From the table below, it can be seen that the null hypothesis that the true difference in means was zero could not be rejected (sigma= 0.138 > 0.05). Hence, it could be drawn that the reading performance does not vary much before and after the course as for the control group.
Test Statisticsa
Posttest - Prestest
Z -1.483b
Asymp. Sig. (2-
tailed) .138
a. Wilcoxon Signed Ranks Test b. Based on positive ranks.
Table 4.6: the non-parametric dependent sample t-test of the experimental group