Child Physical Health, School Engagement, and School Success: A Structural Equation Model

Background

Children’s physical health is an identified and overlooked predictor of school success. Existing evidence suggests that health can influence children’s relationships with peers and teachers that may, in turn, affect their academic achievement. However, more research is needed using (a) ecological, multi-domain models, and (b) samples of younger children. This structural equation modeling analysis used an ecological framework and attachment theory to posit that children’s physical health problems would be negatively related to their school engagement and academic achievement. Data from two time points (fall 2009, spring 2010) were used to test a semi-longitudinal model.

Methods

Sample. Data were collected in 2009–2010 using the Elementary School Success Profile, a well-validated and widely-used school assessment measuring student, teacher, and parent perceptions of children’s social environments. The data set comprised 339 5th grade students in four schools. The diverse sample was 50.3% female and 55.9% White, 17.4% African American, 13.8% Latino, and 13.0% other.

Measures. “Academic Achievement” was measured with two teacher-reported observed items assessing current reading and math performance. School-based “Behavioral Engagement” was measured with four teacher-reported items regarding students’ recent classroom/school-based ability to stay on task and pay attention. School-based “Emotional Engagement” was measured with four child-reported items assessing enjoyment of, and positivity towards, school and learning. The “Physical Health” outcome was measured with seven child-reported items assessing basic health problems such as sleepiness, body aches, and hearing and vision problems. Three dichotomous control variables (“Race,” “Gender,” “Free/Reduced Lunch”) were included to help control for selection bias.

Analysis. Mplus 7.3 was used to test the (a) four baseline measurement, (b) four single wave structural, and (c) two multi-wave structural models for both reading and math. Weighted Least Square (WLSMV) was used with a polychoric correlation matrix of the ordinal observed variables and a correction for the clustering in schools. Full information maximum likelihood (FIML) allowed for the inclusion of missing values.

Results

The final multi-wave models demonstrated good fit. Reading: χ² (371, N = 289) = 407.984, p = .0902; CFI = .969; TLI = .964; RMSEA = .019 (90% CI .00, .03). Math: χ² (371, N = 289) = 410.018, p = .0794; CFI = .967; TLI = .961; RMSEA = .019 (90% CI .00, .03). Thirteen and 15 of 19 hypothesized paths were statistically significant in the reading and math models, respectively. Overall, children’s physical health demonstrated direct effects on engagement and indirect effects on academic achievement. The total standardized positive indirect effects of physical health on reading were .163 (p < .05) and on math were .171 (p < .05). The final models explained 51.3% and 59.5% of the variation in wave two reading and math, respectively.

Conclusions

Results suggest that children with fewer physical health problems will have higher school engagement and achievement. These findings have important implications for social work researchers and practitioners within the domains of Adolescent and Youth Development (ADOL) and School Social Work (SSW).