CHAPTER 4: RESULTS, ANALYSIS AND DISCUSSION
4.3. Questionnaire Validity and Reliability
4.3.1. Exploratory Factor Analysis
Principal components analysis was used because the primary purpose was to identify and compute composite scores for the factors underlying the questionnaire adapted from Collaborative Education Lab guidelines for assessing collaborative learning in the classroom by Luis Valente (2016). Initially, eight factors with Eigenvalues greater than one were extracted, see Figure 4.1. A series of factor analyses were conducted, indicating that four factors gave the most interpretable solution. A varimax rotation was performed since there was no priori assumption of whether factors were expected to be correlated or not. The obtained rotated component matrix is displayed in Table 4.1. Only factor loadings above 0.3 are shown. Internal consistency for each of the extracted factors was examined using Cronbach’s alpha. Factor analysis of the questionnaire items used in the present study revealed four factors were sufficient to the explain the underlying structure of the impacts of collaborative learning on developing critical thinking.
Figure 4.1: Scree Plot
94 Table 4.1. Rotated Component Matrix
Questionnaire Items Componenta
1 2 3 4
1. I understand the importance of working in groups. .438 .354 .395
2. I get along with other team members in my group. .643
3. I respect / accept every team member in my group who is from different culture and background.
.702
4. I respect / accept every team member in my group who has different ability and learning style.
.715
5. I respect / accept different opinions in my group. .565 .300
6. I question the way other team members in my group do and try to think of a better way.
.624 .327
7. I feel that my ideas and suggestions are important to others. .652 .395
8. I feel excited and satisfied to work with my group. .631 .403
9. I like to help my team members in my group. .565 .313
10. I like to think differently in doing activities in my group. .434 .301 .304
13. I prefer to have a leadership role in my group. .462 .396
11. I like to share ideas and suggestions in my group. .305
12. I really enjoy working collaboratively with other students. .647
14. I am an important member in my group. .329
16. My teacher encourages us to work collaboratively in class. .354 21. My teacher monitors / controls students’ interaction in class. .329 .417 26. Working collaboratively with my group improves the content and the structure
of my writing.
.452 .450
29. Working collaboratively with my group makes me more creative. .521
31. Working collaboratively with my group makes me solve problems faster. .497 .414 32. Working collaboratively with my group makes me learn values and new
concepts.
.560
33. Working collaboratively with my group makes me learn and grow from other differences.
.536 .369
34. Working collaboratively with my group changes the way I look at myself. .540 .448 35. Working collaboratively with my group makes me feel better student. .577 .313
38. Working collaboratively with my group makes my communication skills better. .438 .395 40. Working collaboratively with my group creates better opportunity for my
learning.
.484
18. My teacher encourages us to think critically and solve problems. .649 19. My teacher encourages us to be independent and creative. .306 .456
95
Questionnaire Items Componenta
1 2 3 4
24. My teacher shares information that was collected from the group. .602 .385 27. Working collaboratively with my group makes me think differently. .588 .312 28. Working collaboratively with my group makes me think critically. .580
30. Working collaboratively with my group makes me learn new ideas. .389 .480
36. Working collaboratively with my group increases my desire to learn. .636 .331 37. Working collaboratively with my group is better than working individually. .408 .444
39. Working collaboratively with my group encourages me to be more responsible. .343 .605
15. I work hard and effectively in my group. .560
17. My teacher encourages us to discuss topics in class. .405
20. My teacher encourages us to reflect on our actions to see whether we could improve on what we did.
.445
22. My teacher asks useful questions to deepen the study and link to previous topics.
.315 .306 .607
23. My teacher uses differentiated questions that fit students’ abilities and learning style.
.623
25. My teacher treats us fairly and equally. .381 .577
Percentage of Variance 12.64 12.18 11.72 9.65
Eigenvalue 13.62 2.05 1.47 1.34
Cronbach’s alpha .890 .872 .863 .746
Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization. a. Rotation converged in 9 iterations. Note: Factor loadings < .3 are suppressed.
The rotated component matrix in Table 4.1 shows that the first factor was robust, with a high eigenvalue of 13.62, and it accounted for 12.64% of the variance in the data. Factor 2 had an eigenvalue of 2.05 and accounted for a further 12.18% of the variance. The eigenvalues for factors 3 and 4 were 1.47 and 1.34, respectively; together accounting for a further 21.38% of the total variance. Factor 1 consists of 11 items. This factor can be labeled “Individual Role” and demonstrated a high internal consistency (Cronbach’s α = 0.89). The second factor consisted of 14 items including nine items related to “Group Influence”, three items related to “Individual Role”, and two items related to “Teacher Role”. This factor was identified as “Positive Impact of Working Collaboratively on Students”, and reflected a high internal consistency (Cronbach’s α = 0.872). Factor 3 consisted of six items related to “Group Influence” and three items related to
“Teacher Role”, so it was labeled “Encouragement and Motivation”, and it had also a high
96
internal consistency (Cronbach’s α = 0.863). The fourth factor was called “Teacher Role”, and was considered sufficiently reliable (Cronbach’s α = 0.746). For the four factors, no substantial increases in alpha for any of the factors could have been achieved by eliminating more items.
Overall, the factor analysis of the adapted CEL questionnaire items revealed that all items loaded on at least one factor. To conclude, the four factors found were “Individual Role”, “Positive Impact of Working Collaboratively on Students”, “Encouragement and Motivation”, and
“Teacher Role”; which were considered subscales of the “Impact of Collaborative Learning in Relation to Critical Thinking”.
4.3.1.1. Descriptive Analysis of the EFA Extracted Four Factors of the
“Impacts of Collaborative Learning in Developing Critical Thinking” Scale
Composite scores were created for each of the four factors, using the mean of the items which had their primary loadings on each factor. Higher scores indicated greater agreement with the impact of collaborative learning on critical thinking. Encouragement was the main factor of collaborative learning that students reported being influenced by the most, with a negatively skewed distribution (skewness = -1.761). Descriptive statistics are presented in Table 4.2. The skewness and kurtosis were within the reference of substantial departure from normality proposed by West et al. (1996). The proposed reference is an absolute skew value > 2, and an absolute kurtosis value > 7. Examination of histograms indicated some departure from normality.
However, in large samples, a test of normality is more likely to be significant (Field, 2013).
Although a varimax rotation was used, strong correlations between each of the composite scores existed, as shown in Table 4.3.
Table 4.2. Descriptive Statistics forthe Four Factors of the “Impact of Collaborative Learning in Relation to Critical Thinking” Scale
Factors Mean Std. Deviation Skewness Kurtosis Minimum Maximum
Individual Role 4.10 .591 -1.081 .400 2.27 4.91
Positive Impact 4.12 .519 -1.481 1.976 2.29 4.86
97
Encouragement 4.15 .583 -1.761 3.210 1.67 5.00
Teacher Role 4.08 .565 -1.285 1.726 2.00 5.00
Table 4.3. Correlation of Composite Scores
Individual Role Positive Impact Encouragement Teacher Role Individual Role 1
Positive Impact .718** 1
Encouragement .638** .771** 1
Teacher Role .664** .656** .656** 1
**. Correlation is significant at the 0.01 level (2-tailed).
Overall, these analyses indicated that four distinct factors were underlying students’ responses to the adapted CEL questionnaire items and that these factors were highly internally consistent.
However, the original factor structure proposed by the researcher was retained. An approximately normal distribution was evident for the composite scores in the current study; thus the data were well suited for parametric statistical analyses.