Poster #74 - Examining the relationship between Success For All and student engagement

• In Event: 2-110 - Poster Session 07
  In Poster Session: PS 07 Section - Education, Schooling

Integrative Statement

There has been an upsurge of research interest in the concept of student engagement in the past two decades (Dunleavy, Milton, & Crawford, 2010) because of the positive academic and life outcomes associated with increased engagement. Student engagement is commonly understood as students’ behavioral, emotional, and cognitive involvement in and with their learning activities (Fredricks, Blumenfeld, & Paris, 2004). Considering the importance of student engagement, this study examines the impact of one of the most widely used whole school reform programs in the U.S., Success For All, (SFA; Slavin, Madden, Chambers, & Haxby, 2009) on young children’s engagement and the mechanisms through which it may achieve this impact. The current study seeks to answer:

1. Are there differences in engagement among students in SFA and non-SFA schools over time?
2. To what extent does participation in SFA relate to growth in engagement over time?
3. To what extent is the effect of SFA on students’ engagement mediated by academic achievement?

Method
Data
Data come from a quasi-experimental study of whole school reform programs. For this study, there were 977 students (SFA=469, control=508) students within 54 schools (SFA =29, control = 25) who had outcome data in the first year of data collection. Data were collected longitudinally over three years from 2000-2004 when students were in kindergarten to the end of second grade.

Measures
Engagement. The teacher-completed survey included eleven items on students’ engagement behaviors. Sample items include “this student usually pays attention in class” and “this student is eager to learn.” The measure demonstrated high internal consistency (=0.96).
Achievement. The norm-referenced TerraNova scale scores for reading were used as the main measure of achievement. Kindergarten students took the Woodcock-Johnson Tests of Achievement as a pretest in the fall, but subsequent measurements were with the TerraNova.
School-and student-level information. Various aspects of school context such as proportion of families on welfare, percent free lunch students, total school enrollment, and percent minority students were measured. Student demographics such as grade, race, and sex were included.

Analytic model
For the first research question, two-level multilevel models were conducted using Mplus with students at level 1 and schools at level 2. For research question 2, the following three-level multilevel model was run, with time as the first level. Propensity score methods were used to balance the school data in both multilevel models. To answer research question 3, an auto-regressive cross-lagged (ARCL) panel model was used to test mediation over time.

Results
The results indicated that teacher-reported engagement was significantly different between treatment and control groups, with kindergarten students in SFA scoring higher (=0.18, ES=+0.27) than those in comparison schools. Interaction effects run to address whether SFA could affect the rate of growth in non-cognitive outcomes (research question 2) indicated that this interaction term was not significant. Finally, the ARCL panel models indicated that achievement did not significantly mediate SFA’s effect on engagement. However, significant cross-lagged relations between achievement and engagement were found, affirming the reciprocal development of these two constructs over time.

Authors

• Elizabeth Kim, Johns Hopkins University
  Presenting Author

• Robert Slavin, Johns Hopkins University
  Non-Presenting Author