Copyright 2020 The Author(s). Published by vgtu press

Partial eta 
squared
Corrected model
4140.412
a
2
2070.206
6.252
.003
.108
Intercept
41161.480
1
41161.480 124.306
.000
.547
Creativity Index A
4135.745
1
4135.745
12.490
.001
.108
Group
239.731
1
239.731
.724
.397
.007
Error
34106.494
10
3
331.131
Total
1237758.000
106
Corrected total
38246.906
105
a. R-Squared = .108 (adjusted R-Squared = .091)

---

282
M. Khalid et al. Enhancing creativity and problem solving skills through creative problem...
3.1. Results of structured observation
Table 6 shows the mean and standard deviation (SD) for each construct of creativity collected 
through the checklists that were completed by thirty-six observers during the intervention 
of cycles 1 to 3.
The data shows that overall, students’ creativity increases from moderate to high from 
the beginning until the completion of the project. Each construct or creativity, 
i.e.
original-
ity, elaboration, fluency and flexibility also increase from moderate to high from the first to 
the third cycle. In cycle 1, the construct of fluency gives the highest mean while the lowest 
mean is from the construct of flexibility. It can clearly be observed that fluency still shows 
the highest mean in cycles 2 and 3, while flexibility still scores the lowest mean. One-way 
analysis of variance was performed to investigate if the changes are significant from cycle 
to cycle and also according to different schools. Table 7 shows the details of the analysis for 
Table 6. Items in observation checklist (source: created by authors)
Items
Cycle 1
Cycle 2
Cycle 3
Mean
SD
Mean
SD
Mean
SD
Originality
2.921
.644
3.272
.773
3.857
.687
Elaboration
2.907
.701
3.386
.711
3.954
.570
Fluency
2.983
.638
3.405
.673
4.028
1.671
Flexibility
2.861
.767
3.241
.639
3.708
.656
Average score overall
2.910
.708
3.313
.685
3.874
.615
Table 7. Analysis of variance for the comparison of mean in different constructs of creativity for the 
3 cycles (source: created by authors)
Sum of 
squares
Degree of 
freedom
Mean 
square
F-test
Significant
meanOri
Between cycles
16.083
2
8.041
16.182
.000
Within cycle
53.170
107
.497
Total
69.253
109
meanEla
Between cycles
19.755
2
9.877
22.350
.000
Within cycle
47.287
107
.442
Total
67.041
109
meanFlu
Between cycles
29.834
2
14.917
12.395
.000
Within cycle
128.777
107
1.204
Total
158.611
109
meanFlex
Between cycles
12.967
2
6.484
13.669
.000
Within cycle
50.754
107
.474
Total
63.721
109
meanAll
Between cycles
16.888
2
8.444
20.804
.000
Within cycle
43.430
107
.406
Total
60.317
109

---

Creativity Studies, 2020, 13(2): 270–291
283
different cycles. It can be observed that the overall change as well as the changes in originality, 
elaboration, fluency and flexibility are all significant.
Answering research question 2
– whether there are any changes in students’ problem solv-
ing ability.
The result from Table 8 shows a 24% increase in the post-test for the treatment group as 
compared to 15% increase for the comparison group. It can also be observed that there are 
more data from the comparison group for the problem solving test (2 classes) as compared 
to the creativity test (1 class).
Table 8. The marks for problem solving test of students from different schools (treatment and compari-
son group) (source: created by authors)
Treatment
Comparison
School
Gender
Number
Pre- (%) Post- (%)
Number
Pre- (%)
Post-(%)
SMKSS
– All
14
60
95
–
–
–
– Male
7
73
95
–
–
–
– Female
7
48
95
–
–
–
SMKTS
– All
25
37
64
16
44
65
– Male
14
37
59
71
11
47
65
– Female
11
36
5
38
64
SMKHC
– All
22
53
70
–
–
–
– Male
9
44
73
68
–
–
–
– Female
13
59
–
–
–
SMKSG
– All
29
44
68
31
33
45
– Male
15
40
56
55
16
28
39
– Female
14
29
15
39
53
Total
– Male
44
46
68
27
36
50
– Female
46
41
67
20
39
56
Grand total
90
44
68
47
37
52
Table 9. Paired samples test for treatment and comparison group (source: created by authors)
Paired differences
t-test
Degree 
of 
Free-
dom
Sig. 
(2-tailed)
Mean Standard 
deviation
Standard 
error
95% Confidence 
interval of the 
difference
Lower Upper
Pair treat-
ment
Percent 
(post-, pre-)
24.20
20.71
2.18
19.86
28.54 11.09
89
.00
Pair com-
parison
Percent 
(post-, pre-)
15.13
18.74
2.73
9.63
20.63
5.54
46
.00

---

284
M. Khalid et al. Enhancing creativity and problem solving skills through creative problem...
Paired t-test in Table 9 shows a statistically significant result for both the treatment and 
comparison groups, suggesting significant improvement in problem solving skills for both 
groups. Since the increase for the treatment group is larger than the comparison group, AN-
COVA was run to determine if the difference is due to the intervention. The ANCOVA shows 
a statistically significant result suggesting that the larger increase in the treatment group may 
have been due to the intervention (see Table 10). The R-Squared value suggests that 27.3% of 
the change in problem solving marks may be attributed to the intervention.
Meaningful learning refers to how new information can be used effectively when needed. 
When solving mathematics problems in groups and applying the four stages of Polya problem 
solving (2004, p. 41), the participants were actively involved with the cognitive, affective and 
psychomotor domains of learning.
Answering research question 3 
– is there a relationship between creativity and problem 
solving ability?
Table 10. Analysis of covariance on tests of between-treatment effects (source: created by authors)
Dependent variable: percent post-
Source
Type III sum of 
squares
Degree of 
freedom
Mean 
square
F-test
Sig.
nificant