Poster #55 - Academic achievement and profiles of father involvement among children experiencing homelessness

• In Event: Poster Session 01

Abstract

Experiences of family homelessness and associated economic, health, and psychosocial risk factors place children at risk for negative developmental outcomes (Perlman et al., 2012). The impact of homelessness on academic achievement has drawn particular interest, given the importance of success in school for achievements later in life (Masten et al., 2006; 2015). Previous research has identified the parent-child relationship as a key protective factor for academics in contexts of family homelessness (Cutuli & Herbers, 2014; Haskett & Armstrong, 2019; Perlman et al., 2012). However, this research has almost exclusively examined mother-child relationships and neglected the role of fathers, despite their noteworthy contributions to children’s development (Pattnaik & Medeiros, 2013). Generally, father involvement (FI) has been shown to positively predict academic achievement (Jeynes, 2014; Kim & Hill, 2015). No quantitative studies, however, have investigated how homelessness may influence FI, nor has research investigated how FI may moderate the association between family homelessness and academic achievement. The present study employed a quasi-experimental design and a person-centered approach to elucidate the relationships between experiences of family homelessness, FI during late childhood, and adolescent academic achievement.

Participants took part in the Fragile Families and Child Wellbeing Study (Reichman et al., 2001) of approximately 4,700 children and their families followed longitudinally from birth to age 15. We identified a subsample of 570 participants who had experienced homelessness at any point and a propensity score matched comparison group (Guo & Fraser, 2015; Thoemmes & Kim, 2011). Propensity scores were based on 11 covariates indicating sociodemographic risk and adversity. We then conducted a latent profile analysis to derive categorical profiles of FI using seven continuous indicators of FI constructed from parent reports at age 9: amount of time child resides with father, in-person contact, phone contact, financial involvement, educational involvement, routine involvement, and co-parenting relationship. We calculated overall GPA at age 15 by averaging self-reports of students’ grades across four subjects.

The model comparison approach indicated the three-profile solution as having the best balance of theoretical interpretability, BIC model fit, and classification entropy (.85). The largest profile comprised 50.8% of the participants and was characterized by below average involvement across all indicators (Low FI). The next largest profile included 39.3% of the sample with high levels of involvement across all indicators except in-person involvement (High FI). The final profile contained 10.0% of participants with FI average levels in between the two other profiles (Medium FI; see Figure 1). Contrary to our expectations, a multinomial logistic regression showed that homelessness status did not significantly predict FI profile membership beyond the poverty-related risks that contributed to propensity scores, χ2(4) = 4.102, p = 0.392. In a multiple regression model, homelessness and high FI-involvement profile membership each emerged as significant predictors of higher GPA scores.

Results suggest that homeless services and high levels of FI in late childhood may bolster academic achievement in adolescence. Limitations regarding data-collection procedures and definitions of homelessness as well as implications for emergency shelter policy and future research will be discussed.

Authors

• Jake Augustin Leonard, University of North Carolina at Chapel Hill
  Presenting Author

• Alexandra R Buccelli, North Carolina State University

• Janette E. Herbers, Villanova University