|
Results Descriptive statistics and bivariate correlations are presented in Table 1. The inctraclass correlations of the unconditional models for students’ plausible values in reading, mathematics and science controlling for school (SCHOOLID) were ρ = .24, .29, and .27 respectively, indicating that additional variance in plausible values in reading, mathematics and science may be accounted for by controlling for the effects of school at level 2. Fixed effects of the perceived usefulness of summarizing (METASUM), understanding and remembering (UNDREM), and their interaction (METASUM*UNDREM) were added to the unconditional models at the student level. A random intercept, random effect of the perceived usefulness of summarizing, and the random effect of perceived usefulness of understanding and remembering were entered into the models at the school level. Ultimately, three linear mixed models were analyzed to assess the proportion of variability in students’ plausible values in reading, math and science literacies accounted for by perceived usefulness of metacognitive strategies of students within school. All three final models fit the data well relative to their unconditional counterparts. Please refer to Table 2 for model fit indices and regression equations. Holding everything else constant the effects of METASUM, UNDREM and their interaction were all significant, positive predictors across all three criteria: students’ plausible values in reading, mathematics, and science. Together, METASUM, UNDREM, and their interaction accounted for a small proportion of the variation in students’ plausible values in reading, mathematics, and science in the models accounting for school (R2 = .17, .11, and .13 respectively). METASUM seems to make the most contribution in all three models holding everything else constant. Additional analyses were conducted to determine if the effects of the perceived usefulness of metacognitive strategies differ across three criteria: students’ plausible values in reading, mathematics and science. A comparison of the structure of the models for the three criterion variables was conducted by applying the models derived from students’ plausible values in reading to students’ plausible values in mathematics, from students’ plausible values in reading to students’ plausible values in science, and from students’ plausible values in mathematics to students’ plausible values in science and comparing the resulting “crossed R2” with the “direct R2” originally obtained for each criterion. The direct R2 and crossed R2 were not significantly different between any of the three criteria indicating no structural differences between the perceived usefulness of metacognitive strategies and its effect on students’ plausible values in reading, mathematics, and science. |