3.1. Results of Descriptive Analysis
Before testing the hypothesis, a descriptive analysis was carried out on the research data. For this reason, firstly, the results of the analysis of teachers participating in planning collegial discussions, implementing collegial discussions, and evaluating collegial discussions are shown. For more details, the results of the analysis of the frequency of carrying out instructional supervision with collegial discussions are presented in
Table 3.
Based on
Table 3, it can be seen that the majority of teachers simply follow the supervision of teaching through collegial discussions at the planning, implementation, and evaluation stages. This is supported by the results of the mean analysis, which showed the mean values of 3.15, 3.31, and 3.14, with standard deviations of 1.04, 1.01, and 1.04, respectively. The planning is preparing a program of collegial discussion activities, implementation is carrying out collegial discussion activity programs, and evaluation is evaluating the results of collegial discussion activity programs. The frequency of utilizing ICT in the implementation of collegial discussions is presented in
Table 4.
Referring to
Table 4, it can be known that most teachers simply use ICT to conduct collegial discussions. This is supported by the mean value obtained at 3.19, with a standard deviation of 1.01, while for the variations in the utilization of media or communication devices, the average value is 3.17 with a deviation value of 2.28. It was seen from the large standard deviation that there is heterogeneity between teachers, from using only one medium to 10 communication media, with an average of using three communication media.
Meanwhile, the quality of the teacher’s instructional process, whether at the planning, implementation, or evaluation stages of the instructional process, is presented in
Table 5.
Based on
Table 5, it can be concluded that the quality of the teacher’s instructional process was on average in the good category. It can be seen that most of the teachers meet this criterion in planning, implementing, and evaluating the instructional process. This is confirmed by the results of the average analysis obtained as 4.28, 4.20, and 4.12. All of the points go towards frequent execution. Thus, it can be concluded that the average teacher has carried out ideal activities in planning, implementing, and evaluating the instructional process.
On the other hand, the student learning outcomes are divided into three components, namely knowledge, skills, and attitudes. The mean value of the knowledge obtained is 80.22, the mean of skill is 81.40, the mean value of attitude is 82.66, and the overall mean value is 81.55. All the mean values are above 80.00. Thus, it can be concluded that the average value of students’ learning outcomes is in a good category.
3.2. Hypothesis Test
Based on the research design to test the research hypothesis, structural equation modeling analysis techniques were used. The hypothesis that will be answered is that there is a structural influence of instructional supervision through discussion and the utilization of ICT in collegial discussions on the teacher’s instructional process, utilizing ICT in the teacher instructional process, and student learning outcomes.
Regarding the results of the analysis, the chi-square value is 91.24 with a
p-value = 0.007. Looking at the chi-square value, it has not shown a fit result. For that, it is necessary to look at other goodness of fit criteria. The results of other goodness of fit analyses are presented in
Table 6.
Based on the values of the goodness of fit criteria presented in
Table 6, it can be concluded that the model of the structural influence of exogenous and endogenous variables proposed in this study fits with the data in the field. GFT, AGFI, NFI, and NNFI values are all above 0.9, and RMSEA values are below 0.08. Thus, the null hypothesis is rejected, and the working hypothesis is accepted. The proposed hypothetical model fits the data in the field.
In an outline, the structural relationship model between these variables is presented in
Figure 1.
Information:
CD = Collegial discussion;
X1 = Collegial discussion planning;
X2 = Collegial discussion implementation;
X3 = Collegial discussion evaluation;
ICTC = Utilizing ICT in collegial discussion;
X4 = Frequency of utilizing ICT in collegial discussion;
X5 = Types of media devices were used in collegial discussion;
TQ = The quality of the teacher’s instructional process;
Y1 = Lesson planning;
Y2 = Teaching implementation;
Y3 = Teaching evaluation;
ICTT = Utilizing ICT in the instructional process;
Y4 = The frequency of utilizing ICT in the instructional process;
Y5 = Types of media devices for utilizing ICT in the instructional process;
SA = Student’s learning outcomes;
Y6 = Knowledge learning outcomes;
Y7 = Skill Learning outcomes;
Y8 = Attitude learning outcomes;
Y9 = Total learning outcomes.
Figure 1 strengthens the results of the goodness of fit analysis of the model of the influence of exogenous and endogenous variables, and between endogenous variables and endogenous variables fit with data in the field. The frequency of teachers participating in collegial discussions has a direct effect on the quality of the teachers’ instructional process. The use of ICT in collegial discussions has a direct effect on the quality of the teachers’ instructional process and the use of ICT in the teachers’ instructional process. The quality of the teachers’ instructional process and the use of ICT in the teachers’ instructional process have a direct effect on students’ learning outcomes. X is the dimension of the observed exogenous variables, while Y is the dimension of the observed endogenous variables. The score of each variable dimension is obtained from the score of the answers to the items of the measured variable dimensions. The arrow indicates the direct effect. In addition, the indirect effect is obtained from the output of the linear structural equation analysis in more detail. Hence, the direct and indirect effects between variables are presented in
Table 7.
Based on
Table 7, it can be concluded that the frequency of teachers carrying out collegial discussions has a direct effect on the quality of the teachers’ instructional process, with a coefficient of 0.222. The more actively the teacher participates in group discussions, the higher the quality of the teacher’s instructional process. The frequency of teachers utilizing ICT in collegial discussions has a direct effect on the quality of the teacher’s instructional process, with a coefficient of 0.225, and the utilization of ICT in the teacher’s instructional process, with a coefficient of 0.367.
The quality of the teachers’ instructional process has a direct effect on student learning outcomes with a coefficient of 0.292. Therefore, the higher the quality of the teachers’ instructional process, the higher the learning outcomes achieved by the students. The utilization of ICT in the teachers’ instructional process has a direct effect on the students’ learning outcomes, with a coefficient of 0.281. This means that the more the teacher utilizes ICT in the instructional process, the higher the learning outcomes achieved by the students.
The frequency of teachers carrying out collegial discussions has an indirect effect on the students’ learning outcomes, with a coefficient of 0.065. Likewise, the utilization of ICT in collegial discussions also has an indirect effect on the students’ learning outcomes, with a coefficient of 0.169. The value of direct, indirect, and total effects is obtained from the results of the Lisrel (Linear Structural Relation) program analysis.
We can notice from the support for each of the observed variables on the latent variable that it also shows strong results. In other words, the exogenous and endogenous variable measurement model fits the data in the field. Thus, the results of the measurement model analysis are presented in
Table 8.
Based on
Table 8, it can be underlined that in the collegial discussion variable, the planning factor obtained a value of λ = 0.933, the implementation obtained a value of λ = 0.901, and the evaluation of collegial discussion obtained a value of λ = 0.930. The three lambda values are quite high, so this shows that the three observed variables are the main indicators of the construct of collegial discussion variables. From the results of the analysis, it is also known that the variable of ICT utilization in collegial supervision, the frequency factor of ICT utilization, obtained the value of λ = 0.934, while the type of ICT devices factor obtained the value of λ = 0.552. The two lambda values are also quite high, indicating that the two observed variables are the main indicators of the construct of utilizing ICT in collegial discussions.
Based on
Table 8, it is also known that the teacher instructional process variable, teaching planning, obtained the value of λ = 0.885, the teaching implementation factor obtained the value of λ = 0.761, and the teaching evaluation obtained a coefficient of λ = 0.898. The three lambda values are quite high, so the two observed variables are the main indicators of the construct of the teacher’s instructional process. From the results of the analysis, it is also known that, for the variable of ICT utilization in the instructional process, the frequency factor for using ICT obtained the value of λ = 0.872, while the factor of variation in the type of ICT devices obtained the value of λ = 0.787. The two lambda values are also quite high, indicating that the two observed variables are the main indicators of the construct of utilizing ICT in the teacher’s instructional process.
Based on
Table 8, it is also known that the student learning outcomes variable, namely knowledge learning outcomes, obtained the value of λ = 0.891, the skill learning outcomes factor obtained a value of λ = 0.867, the attitude learning outcomes obtained a coefficient of λ = 0.705, and total learning outcomes obtained a coefficient of λ = 0.923. The four lambda values are quite high, so the four observed variables are the main indicators of the construct of student learning outcomes.