Hill, N., & Tyson, D. (2009). Parental involvement in middle school: A meta-analytic assessment of the strategies that promote achievement. Developmental Psychology, 45, 740–763.
Summary by Dr. Nancy Scammacca
Overview
Parental involvement, broadly defined, has been shown to play a critical role in children’s academic achievement. For elementary school students, some consensus exists about how to involve parents in their children’s education. At the middle school level, less is known about the best ways to keep parents involved. The multifaceted changes brought on by adolescence affect the parent-child relationship in many ways, including how parents are involved in their children’s education during the middle school years.
Recent attempts to synthesize the literature on parental involvement in education have not taken into account the ways in which the effectiveness of particular strategies might change between elementary school and middle school. Contemporary research on parental involvement in middle school has produced contradictory findings, making it difficult to determine which strategies to recommend. Hill and Tyson (2009) determined that a meta-analytic review (see below for a description of this type of study) would prove helpful in understanding how best to involve parents in the education of middle school students. They sought to understand the relationship between academic achievement and parental involvement in middle school and to isolate the strategies with the strongest relationship to academic success.
Hill and Tyson defined parental involvement as including activities that occur either at school (such as attending events at school or volunteering at school) or at home (such as helping with homework). The authors noted that parental involvement at school is often more difficult in middle school than in elementary school because middle schools are larger and more challenging to navigate, with students typically having many more teachers. This difficulty is compounded by the desire of middle school students to have more freedom from direct parental involvement during adolescence. As a result, parents may withdraw from many of the ways they pursued involvement during their children’s elementary school years and attempt to be involved more indirectly, such as by sharing their expectations for their children’s academic achievement, showing how their children’s learning relates to the world around them, helping their children with learning strategies, and helping to plan their children’s educational experience in high school and beyond. Hill and Tyson called these strategies for indirect involvement “academic socialization” and expected that they would be the most effective method for parental involvement during middle school.
Study Design
Hill and Tyson conducted a meta-analysis to determine which strategies for parental involvement have the greatest effects on the academic achievement of middle school students. A meta-analysis is a “study of studies.” It takes a comprehensive look at research published in a particular area and compares the effects found in each study, using a statistic known as an effect size. Based on the meta-analytic results, researchers can develop recommendations for best practices that reflect the knowledge gained from many different studies. A meta-analysis also allows researchers to test how different factors influence the strength of effects across a group of studies. The Hill and Tyson meta-analysis examined differences in effects, based on ethnicity and type of parental involvement (home based, school based, and academic socialization).
Another important factor in understanding the results of a meta-analysis is the types of studies included. Hill and Tyson included only the following types of studies published between 1985 and 2006:
- Observational studies that reported correlations (see below for a description of this statistic) between parental involvement and academic achievement
- Experimental studies that tried to promote parental involvement
- Studies that used data from publicly available national datasets, such as the 1988 National Education Longitudinal Study
A total of 27 observational studies with 92 correlations between parental involvement and academic achievement were included in Hill and Tyson’s analysis, along with 5 experimental studies and 3 studies that used national datasets.
Hill and Tyson reported their results by using the correlation coefficient (r) as the effect size statistic for the 27 observational studies. Correlation coefficients can range between -1.0 and +1.0. In the context of this study, a large negative correlation (e.g., -0.70) would mean that high parental involvement is associated with low academic achievement for middle school students. A correlation of 0 would indicate no relationship between academic achievement and parental involvement. A large positive correlation (e.g., 0.70) would mean that high parental involvement is associated with high academic achievement. It is important to note that a correlation coefficient does not mean that one variable causes the other. Rather, it just indicates a relationship between the two variables. In a meta-analysis, effect sizes from individual studies are averaged in a way that gives more weight to studies with a larger number of people in them. Weighted average effect sizes are reported with a 95% confidence interval (CI), which is a range of values within which the true effect size most likely lies.
Key Findings
Across the 27 correlational studies, Hill and Tyson found an average correlation of 0.18 (95% CI = 0.12, 0.24), indicating an overall positive relationship between parental involvement and academic achievement in middle school that is significantly different from 0. The researchers found that the average correlation was similar for European American parents and for African American parents. Next, the authors looked at whether one type of parental involvement had a stronger relationship with academic achievement than another. The findings indicated the following:
- The average correlation for home-based involvement was 0.03, which is not significantly different from 0 (95% CI = -0.02, 0.11).
- The average correlation for school-based involvement was 0.19, which is significantly different from 0 (95% CI = 0.10, 0.21) and is significantly larger than the effect size for home-based involvement.
- The average correlation for academic socialization was 0.39, which is significantly different from 0 (95% CI = 0.26, 0.44) and significantly larger than the effect sizes for home-based and school-based involvement.
To better understand why home-based involvement was not associated with academic achievement, Hill and Tyson looked at the effect sizes for two types of home-based involvement: help with homework and activities at home (such as providing books and other educational resources and taking trips to museums and libraries). The effect size for help with homework was negative (-0.11; 95% CI = -0.04, -0.25); the effect size for activities at home was positive (0.12, 95% CI = 0.05, 0.19) and significantly different from the effect size for help with homework. Hill and Tyson pointed out that parents might be more likely to provide homework help for children whose academic achievement is poor than for children who are doing well. If true, one possible explanation for the negative relationship is that low academic achievement encouraged more parental involvement than high academic achievement, rather than parental involvement leading to low achievement.
Hill and Tyson looked next at the longitudinal and experimental studies to determine the direction of the causal relationship between parental involvement and academic achievement (in other words, to answer the question: “Do different types of parental involvement cause different achievement outcomes?”). The longitudinal studies showed that when parent involvement was high at the first measurement time point in the study, academic achievement was high at a later time point in the study. The experimental studies primarily focused on improving parents’ ability to help with homework. They resulted in an average effect size that was not significantly different from 0. Therefore, no clear causal relationship between parental help with homework and achievement in middle school could be verified.
Recommendations
Recommendation 1: Parents should remain involved in their children’s education during the middle school years.
Because the studies available for inclusion in this meta-analysis were almost all correlational, a clear causal link between parental involvement and academic achievement during middle school cannot be made. However, it is reasonable to infer from this meta-analysis that at least some forms of parental involvement are beneficial to middle school students. School-based involvement, through activities such as attending school events and volunteering at school, positively related to academic achievement. Home-based involvement, through means such as providing educational resources and books, also positively related to achievement. Additionally, a clear positive relationship was found between academic socialization and academic achievement. These types of activities include discussing academic strategies and future educational plans and communicating expectations regarding a child’s ultimate educational attainment.
Recommendation 2: Successful parental involvement during middle school likely is different from successful parental involvement during elementary school.
During middle school, when children typically desire more autonomy from their parents and the school environment becomes more complex for parents to navigate, some types of involvement are likely to work better than others. In particular, the nature of the relationship between providing help with homework and achievement is unclear, suggesting that this form of involvement may not be of much benefit to middle school students. Additionally, although school-based involvement was associated with positive results, many adolescents do not want their parents to spend time at their school, and building relationships with the many teachers who instruct a middle school student can prove impossible. Middle schools may need to create new opportunities to allow parents to become involved in the school in meaningful ways that are acceptable to students and manageable for parents. These opportunities could include promoting parent-child discussions about careers and educational options for the future, providing resources so that parents can reinforce at home the learning strategies taught at school, and helping parents to articulate their academic expectations to their children.
Recommendation 3: In research syntheses such as meta-analyses, to obtain the most accurate findings, research questions must be specific.
When considering parental involvement, these researchers first identified three broad categories of involvement (home based, school based, and academic socialization). Home-based involvement was found to “have no effect” on academic achievement. However, when the researchers looked at a more specific research question (help with homework separately from activities at home), a more nuanced picture emerged. Help with homework continued to have “no effect,” but activities at home had a small positive effect that was statistically different from both 0 and the effect of help with homework. These results suggest that some types of home-based parental involvement do, in fact, positively relate to academic achievement—a finding that is different from the focus of initial question, which looked at home-based involvement as a single variable.
Recommendation 4: More rigorous research is needed to determine the nature of the relationship between academic achievement and parental involvement during middle school.
As mentioned above, correlational studies do not prove causation. It is just as likely that the high academic achievement of a middle school student leads to greater parental involvement as it is that greater parental involvement leads to high academic achievement. The small number of experimental studies available for inclusion in this meta-analysis did not allow for the nature of the causal relationship between involvement and achievement to be made clear. An example of a rigorous research study that would allow for conclusions to be made about causality is a study that randomly assigns one group of parents to receive training in involvement strategies and another group to receive no training and then measures the academic achievement of their children before the parents receive training and after the parents have implemented the strategies for a period of time. If this type of study were to show that the children whose parents received training had significantly higher achievement than the children whose parents did not, evidence for a clear causal link would be established.