Sanders, J.E. & Fallon, B. (2018). Child welfare involvement and academic difficulties: Characteristics of children, families and households involved with child welfare and experiencing academic difficulties. Children and Youth Services Review, 86, 98-109.
Academic difficulties are among the most frequently identified child functioning concerns for children involved with child welfare services. This paper used data from the Ontario Incidence Study of Reported Child Abuse and Neglect (OIS-2013), which was based on a large sample of child welfare investigations (n = 4033), involving children aged 4-15, in 17 agencies selected in a multi-stage sampling design. Data from individual maltreatment assessment forms, completed by child welfare workers at the conclusion of child investigations, were analyzed to examine possible relationships between academic difficulties and 1) maltreatment categories, 2) transfer to ongoing child welfare services.
Results: In the total sample, academic difficulties were noted for 26% of the boys and 16.7% of the girls. When occurrence was individually crosstabulated by 12 child, household and maltreatment variables, Chi-Square was highly significant (p < .001) in every case. “Indigenous students were identified as having academic difficulties more often than any other ethnicity”.
1) Maltreatment categories found to be most strongly associated with academic difficulties were neglect, physical abuse and emotional abuse. Categories least strongly associated were sexual abuse and exposure to interpersonal violence.
2) Unsafe housing and caregiver functioning issues were found to be the strongest predictors of a decision to transfer a case for ongoing child welfare services. Academic difficulties alone were not a significant predictor; however, children with academic difficulties often had associated problems (e.g., depression, anxiety, drug abuse) and the combination could result in such a transfer.
The authors point out that schools are in a position to identify predictors of academic difficulties (e.g., unsafe housing, caregivers with deficits and few social supports). They argue that early identification of students at risk could allow for targeted supports, such as extra tutoring, before academic difficulties become entrenched. Earlier intervention could minimize impact, reduce distress, and improve outcomes for children and families.
Data for the OIS-2013 were collected over a 3 month sampling period (Oct. 1 -- Dec. 31), which did not include variations due to seasons, or academic year. Presence of academic difficulty was assessed only by the investigating workers, not verified by an independent source. Information was only tracked for the first 30 days of an investigation.
Data analysis was complicated. The relationship between academic difficulties and individual case variables was assessed using crosstabs plus chi-square tests of statistical significance. The relative effects of maltreatment type, academic difficulty and other case variables on a decision to transfer a child to ongoing child welfare services were assessed using multiple logistic regression.
The original OIS dataset had been weighted twice, to compensate for 1) a purposeful agency- selection procedure, and 2) the use of only Oct–Dec data. The article describes these procedures, which resulted in a total weighted N = 95,565. Actual and weighted sample sizes were used in various statistical tests -- weighted data in the crosstabs, unweighted data in the logistic regressions. This was a complex process, and the description wasn’t always easy to follow.