Volume 1 - Issue 1 - DBU Journal for K-12 Educational Leadership - Page 25

Journal for K-12 Educational Leadership 23 Research Question 3: Does a relationship exist between scores on the engagement indicators of the Gallup Student Poll and the number of behavior removals at middle school campuses? The current study was conducted in a school district in the Dallas-Fort Worth area, which includes three high schools, seven middle schools, 21 elementary schools, two early childhood centers, an alternative high school, and an advanced technology complex. The enrollment of more than 26,000 students is culturally diverse, composed of approxi- mately 52% White, 32% Hispanic, and 12% African Amer- ican students, with 43% of students considered to be eco - nomically disadvantaged (ED) and 14% English language learners (Texas Education Agency, 2013b). The unit of study in the current investigation was the school. The six middle schools within the district which par - ticipated in the Gallup Student Poll from 2010 through 2013 constituted a convenience sample. Research Question 1 Research Question 1 addressed the potential relationship between student engagement and mathematics perfor - mance. To determine whether a relationship exists between these variables, a correlation analysis was performed using the difference in Gallup Student Poll grand mean scores between the Fall 2010 and the Fall 2013 as the measure of student engagement, and the difference in the percentage of students passing the STAAR mathematics assessment be- tween the Spring 2012 and the Spring 2014 as the measure of mathematics performance. Scatter plots were generated to visualize the relation - ship between gains in engagement and math performance for each student group. The plot for the all students group suggests that a linear relationship may exist for five of the six campuses, and one campus may represent an outlier in the data set. Based on this evidence, the correlation analy- sis was re-run with the apparently aberrant data point re - moved, which resulted in a statistically significant correla - tion ( r = .974, t (3) = 7.45, p = .003). The results suggest that a true relationship may exist between student engagement and math performance. Figure 1 displays the scatter plot of Gallup Student Poll gains and STAAR math gains for all students. Research Question 2 Research Question 2 focused on the potential relationship between student engagement and school attendance. A sta - tistically significant relationship between student engage - ment and attendance was not detected for all students ( t (4) = 0.812, p = .462), economically disadvantaged ( t (4) = 0.185, p = .862), and LEP ( t (4) = 0.020, p = .986) student groups. Since all p-values quantified were reported at greater than .05, the statistical test failed to reject all null hypotheses.