• 沒有找到結果。

Comparing satisfaction among DIANA, FUTS, and random composition groups

The second research question addressed whether the DIANA and/or FUTS grouping methods drew a higher number of positive subjective comments concerning group partners and the cooperative learning process. A 3 (DIANA, FUTS, Random)

×2 (elementary student, junior high student) mixed-design ANOVA was created for group partner and cooperative process satisfaction (descriptive statistics in Table 14, summary in Table 15). Data on mean member satisfaction for each grouping method across the two school levels are presented in Fig. 21a, mean member satisfaction for each school level across the three grouping methods are presented in Fig. 21b, mean process satisfaction for each grouping method across the two school levels are presented in Fig. 22a, and mean process satisfaction for each school level across the three grouping methods are presented in Fig. 22b. Statistical analysis results are shown in Table 16.

The main effect of grouping method on member satisfaction was significant (F(2, 457)=3.630, p=.027), suggesting that students grouped by DIANA (M=3.49)

perceived greater satisfaction with their fellow group members than students grouped randomly (M=3.27). In addition, the main effect of the grouping method on process satisfaction was significant (F(2, 457)=5.06, p=.007), suggesting that students grouped by DIANA (M=3.46) perceived more overall satisfaction with cooperative learning than FUTS (M=3.22) or randomly grouped students (M=3.19). No significant main effect for school level was observed on either member satisfaction or process satisfaction, but a significant interaction between school level and grouping method was observed for both member and process satisfaction.

Table 14 Means and Standard Deviations for Method Conditions as a Function of School Level condition (Satisfaction)

Member Process

Table 15 Summary of Two-Way Analysis of Variance for Grade and Grouping Method

Error 274.428 469 .585

Process

School Level .508 1 .508 .775

Method 6.642 2 3.321 5.060**

School Level * Method 8.492 2 4.246 6.469**

Error 307.827 469 .656

*p < .05 **p < .01 ***p<.001

The data in Tables 16 and 17 indicate that when grouping method is controlled for, junior high students grouped by the FUTS method had significantly more positive perceptions of their teammates (M= 3.49) and the cooperative process (M= 3.46) than their elementary school counterparts (teammates M= 3.16, F = 6.0, p< .05;

cooperative process M= 3.02, F = 9.37, p< .01). Random and DIANA grouping did not result in statistically significant differences between the elementary and junior high students. As shown in Tables 16 and 17 and Fig. 21 and 22, elementary and junior high students had significantly different perceptions across the three grouping methods. For elementary students, DIANA groups had significantly higher scores for satisfaction than FUTS groups (Member: DIANA 3.49 > FUTS 3.16, F = 3.30,

p < .05; process: DIANA 3.46 = random 3.31 > FUTS 3.02, F = 6.09, p < .01). Junior

high students in FUTS and DIANA groups had significantly higher perceptions of satisfaction than their randomly assigned counterparts (Member: FUTS 3.49 = DIANA 3.49 > random 3.20, F =4.36, p <.05; process: FUTS 3.46 = DIANA 3.45 >

random 3.07, F = 5.70, p <.01). The summarized information in Fig. 21 and 22 demonstrate that DIANA was the best grouping method for elementary students in this study and that junior high DIANA and FUTS group members had better perceptions of satisfaction than randomly assigned group members.

Table 16. Simple main effect (Member)

Simple Main Effect SS df MS F

School Level

RANDOM .95 1 .95 1.53

DIANA 1.718E-04 1 1.718E-04 .00

FUTS 3.38 1 3.38 6.00*

Method

Elementary Students 4.33 2 2.16 3.30*

Junior High Students 4.37 2 2.18 4.36*

*p < .05 **p < .01 ***p<.001 Table 17. Simple main effect (Process)

Simple Main Effect SS df MS F

School Level

RANDOM 2.491 1 2.491 3.35

DIANA 6.585E-03 1 6.585E -03 .012

FUTS 6.038 1 6.038 9.37**

Method Elementary

Students

7.846 2 3.923 6.09**

Junior High Students

7.662 2 3.831 5.70**

*p < .05 **p < .01 ***p<.001

(a)

2.9 3 3.1 3.2 3.3 3.4 3.5 3.6

Random DIANA FUTS

Grouping Method

Member Satisfaction

Elementary Students Junior High Students

(b)

2.9 3 3.1 3.2 3.3 3.4 3.5 3.6

Elementary Students Junior High Students

School Level

Member Satisfaction

Random DIANA FUTS

Fig. 21 (a) Mean member satisfactions for each grouping method across two school levels. (b) Mean member satisfactions for each school level across the grouping methods

(a)

Elementary Students Junior High Students

School Level

Process Satisfaction

Random DIANA FUTS

Fig. 22 (a) Mean process satisfactions for each grouping method across two school levels; (b) Mean process satisfactions for each school level across the grouping methods

In a like manner, a 2 (low vs. high density)×3 (DIANA vs. FUTS vs. random) mixed-design ANOVA was conducted on group partner satisfaction and cooperative process satisfaction (descriptive statistics in Table 18; results summary in Table 19).

Data on mean member satisfaction for each grouping method across the two sample density levels are presented in Fig. 23a, mean member satisfaction for each density level across the three grouping methods are presented in Fig. 23b, mean process satisfaction for each grouping method across the two density levels are presented in Fig. 24a, and mean process satisfaction for each density level across the three grouping methods are presented in Fig. 24b. Results from a statistical analysis are shown in Table 20.

The main effect of grouping method on member satisfaction was significant (F(2, 337)=9.557, p=.000), suggesting that students grouped by DIANA (M=3.46) or FUTS (M=3.31) perceived greater satisfaction with their group members than students grouped randomly (M=3.08). The main effect of grouping method on process

satisfaction was also significant (F(2, 377)=10.469, p=.000), suggesting that students grouped by DIANA (M=3.45) perceived the highest level of satisfaction with the cooperative learning process, followed by FUTS (M=3.22) and randomly assigned students (M=3.04). No significant main effect for sample density level was observed for either member or process satisfaction. Statistically significant interaction was noted between sample density level and grouping method for both member and process satisfaction.

Table 18 Means and Standard Deviations for Method Conditions as a Function of Density Level condition (Satisfaction)

Member Process

Table 19. Summary of Two-Way Analysis of Variance for Sample Density and Grouping Method (Satisfaction)

Source SS df MS F

Process

Density 1.018 1 1.018 1.612

Method 13.222 2 6.611 10.469***

Grade * Method 6.635 2 3.317 5.253**

Within cells 209.021 331 .631

Total 3813.675 337

Member

Density .505 1 .505 .855

Method 11.276 2 5.638 9.557***

Grade * Method 5.271 2 2.635 4.467*

Within cells 195.268 331 .590

Total 3897.293 337

*p < .05 **p < .01 ***p<.001

As the analysis of simple main effect data presented in Tables 20 and 21 show, low density sample students grouped randomly had significantly better perceptions of teammates and of the cooperative process than their high density counterparts

(teammates M= 3.22 versus M= 2.72, F = 6.888, p< .01 and cooperative process M=

2.62 versus M= 3.21, F = 9,999, p< .01) when grouping method was controlled for.

No signficant differences were noted between low and high density samples in terms of FUTS or DIANA grouping. The data in Tables 20 and 21 and Fig.23 and 24 indicate significantly different perceptions for the high density samples across the three grouping methods. FUTS and DIANA groups had significantly better

perceptions than the randomly assigned groups (member: DIANA 3.49 = FUTS 3.45

> random 2.72, F = 12.822, p <.001; process: DIANA 3.45 = FUTS 3.39 > random 2.63, F =12.398, p <.001). In summary, members of DIANA and FUTS groups in high density classes had better perceptions of fellow team members and the overall process than their randomly assigned counterparts (Fig. 23 and Fig. 24).

Table 20. Simple main effect (Member) Simple Main

Effect

SS df MS F P

Density

RANDOM 4.477 1 4.477 6.888** .010

DIANA 9.995E-02 1 9.995E-02 .180 .672

FUTS .806 1 .806 1.376 .243

Method

Low 1.557 2 .779 1.161 .315

High 11.477 2 5.739 12.822*** .000

*p < .05 **p < .01 ***p<.001

Table 21. Simple main effect (Process) Simple Main

Effect

SS df MS F P

Density

RANDOM 5.966 1 5.966 9.999** .002

DIANA 7.507E-04 1 7.507E-04 .001 .972

FUTS 1.084 1 1.084 1.582 .211

Method

Low 3.232 2 1.161 2.337 .099

High 13.045 2 6.522 12.398*** .000

*p < .05 **p < .01 ***p<.001

(a)

Fig. 23 (a) Mean member satisfactions for each grouping method across two density levels. (b) Mean member satisfactions for each density level across the grouping methods

(a)

Fig. 24 (a) Mean process satisfactions for each grouping method across two density levels; (b) Mean process satisfactions for each density level across the grouping methods

相關文件