Page 43 - Educational Leadership in a Changing World
P. 43

Unearthing Context

            used. For this purpose, the Hayes’ process Procedure for spss Ver-
            sion 4.0 (Hayes, 2022) was chosen. This approach provides a robust
            framework for testing indirect and interaction effects through boot-
            strapped confidence intervals, improving upon traditional stepwise re-
            gression methods (Baron & Kenny, 1986). Descriptive statistics and
            Pearson correlation were computed before the main analysis to assess
            variable relationships and potential multicollinearity. Statistical signif-
            icance was set at p <0.05.
              Mediation analysis was conducted using process Model 4 to deter-
            mine whether escs mediates the relationships between instructional
            leadership, school autonomy, and students’ science achievement. This
            model estimates:
              • Path a: The direct effect of instructional leadership and school au-
                tonomy on escs.
              • Path b: The direct effect of escs on science achievement, control-
                ling for instructional leadership and school autonomy.

              • Path c : The direct effect of instructional leadership and school au-
                tonomy on science achievement after controlling for escs.
              • Path c: The total effect of instructional leadership and school au-
                tonomy on science achievement. The total effect is the sum of the

                direct and the indirect effect (c = c +a×b).
              • Indirect Effect (a × b): Bootstrapping with 5,000 resamples was
                used to estimate bias-corrected confidence intervals (95 ci) for
                the indirect effect. Mediation was considered statistically signifi-
                cantifthe bootllci and bootulci didnotincludezero(Preacher
                & Hayes, 2008).

              The mediation model follows the equations:
                  M = β 0 +β 1 X +ε,
                  Y = β 0 +β 2 X +β 3 M +ε,

            where M represents escs, X represents instructional leadership or
            school autonomy, and Y represents students’ science achievement. The
            indirect effect (ie) was calculated as β 1 × β3. The overall mediation
            model of this study is presented in Figure 2.2.
              To assess whether escs moderates the relationships between in-
            structional leadership, school autonomy, and science achievement, we
            employed process Model 1. This model tests:


                                                            43
   38   39   40   41   42   43   44   45   46   47   48