 |
Asia-Pacific Forum
on Science Learning and Teaching, Volume 10, Issue 2, Article 4 (Dec., 2009) |
In this section, descriptive statistics of the students’ environmental knowledge and the differences seen in the groups’ the pre-test, post-test and monitoring scores are presented.
Table 1. Descriptive Statistics Concerning the Environmental Knowledge Pre-test, Post-test and Monitoring Scores of the Students in the Control and Experimental Groups
Grups
N
Mean
Std. Deviation
Pre-test
Control
36
7.75
1.872
Experimental 1
35
7.25
1.686
Experimental 2
30
7.46
2.255
Total
101
7.49
1.926
Post-test
Control
36
8.66
1.690
Experimental 1
35
10.20
1.745
Experimental 2
30
10.00
1.701
Total
101
9.59
1.834
Monitoring test
Control
36
8.33
1.414
Experimental 1
35
9.77
1.733
Experimental 2
30
9.56
1.794
Total
101
9.19
1.755
As can be seen in Table 1, the environmental knowledge pre-test scores of the students are close to each other (
(control)=7.75,
Figure 1. Profile Plots showing the environmental knowledge pre-test, post-test and monitoring test scores of the control and experimental students
As can be seen above, throughout the study the environmental knowledge scores of the three groups increased. Yet, a higher score increase favouring the experimental groups can clearly be seen in Figure 1. Scores of the monitoring test administered three months after the study ended indicate little decrease in the scores of the three groups. To determine whether the score differences among the groups are statistically significant, a two-way ANOVA for mixed measures was employed and the results are presented below.
The covariance homogeneity of the groups was tested to determine the suitability of the variance analysis aiming to elicit the significance of the changes in the scores, and covariances were found to be homogenous (F(12-43777, 077)=1.197; p>.05) (Table 2).
Table 2. The Results of Box's Test of the Equality of Covariance Matrices
Box's M
15.037
F
1.197
df1
12
df2
43777.077
Sig.
.278
Results of Levene statistics were examined to test the hypothesis of the equality of the variances belonging to the groups, and for three variables the hypothesis was confirmed (F(2-98)=1.096, .014 and .559; p>.05, respectively).
Table 3. The Results of Levene's Test of Equality of Error Variances
F
df1
df2
Sig.
Pre-test
1.096
2
98
.338
Post-test
.014
2
98
.986
Monitoring test
.559
2
98
.574
As can be figured out in Table 4, the environmental knowledge scores of the experimental groups trained according to green class model and those of the control group students exhibited significant changes from the beginning of the study to the end of monitoring process; that is, the effects of the model were found to be significant (F(4-196)=7.227; p<.001). On the other hand, F(2-196)=72.532 value is significant at the level of 0.001, indicating that there are significant differences among the environmental knowledge scores of the students obtained before the study, after the study and after the monitoring process.
Table 4. The Results of Tests of Within-Subjects Effects
Source
Type III Sum of Squares
df
Mean Square
F
Sig.
Factor1
Sphericity Assumed
257.622
2
128.811
72.532
.000
Greenhouse-Geisser
257.622
1.156
222.907
72.532
.000
Huynh-Feldt
257.622
1.185
217.461
72.532
.000
Lower-bound
257.622
1.00
257.622
72.532
.000
Factor1 * Grups
Sphericity Assumed
51.341
4
12.835
7.227
.000
Greenhouse-Geisser
51.341
2.311
22.212
7.227
.001
Huynh-Feldt
51.341
2.369
21.669
7.227
.001
Lower-bound
51.341
2.000
25.671
7.227
.001
Error (Factor1)
Sphericity Assumed
348.078
196
1.776
Greenhouse-Geisser
348.078
113.262
3.073
Huynh-Feldt
348.078
116.098
2.998
Lower-bound
348.078
98.00
3.552
Besides the findings presented above, the results of the within-subjects effects tests show that there are significant differences between the total score mean (sum of pre-test, post-test and monitoring test scores) of the experimental groups and the control group, stemming from the environmental education given to the experimental groups in line with green class activities and the environmental education given to the control group (F(2-98)= 3.792; p<.05).
Table 5. The Results of Tests of Between-Subjects Effects
Source
Type III Sum of Squares
df
Mean Square
F
Sig.
Intercept
23203.629
1
23203.629
3975.706
.000
Grups
44.262
2
22.131
3.792
.026
Error
571.963
98
5.836
The results of LSD multiple-comparison test were used to find the source of the environmental knowledge score differences between the groups. The results are presented in Table 6.
Table 6. LSD Analysis-based Multiple-comparison Results Concerning the Environmental Knowledge Scores of the Control and Experimental Groups.
(I) Grups
(J) Groups
Mean Difference (I-J)
Std. Error
Sig.
Control
Experimental 1
-.826(*)
.331
.014
Experimental 2
-.761(*)
.344
.030
Experimental 1
Control
.826(*)
.331
.014
Experimental 2
.065
.347
.852
Experimental 2
Control
.761(*)
.344
.030
Experimental 1
-.065
.347
.852
* The mean difference is significant at the .05 level.According to the results of the multiple-comparison test, the environmental knowledge score means of experimental group 1 and experimental group 2 are significantly better than that of the control group students (Table 6). The difference between the environmental knowledge mean scores of the experimental groups was not found to be significant. These results show that the environmental education given according to green class model is more effective in enhancing the environmental knowledge and its retention.
