The MPACT initiative: Using Behavioral Tools to Improve Children’s Math Skills

• In Event: Leveraging Behavioral Insights to Empower Parents to Support Young Children’s Development: Evidence from Experimental Interventions

Abstract

Compared to more advantaged parents, disadvantaged parents spend less time and especially less time in educational activities with their children (Guryan et al., 2008; Kalil et al., 2012). Research shows a causal relationship between the amount of time that parents spend in cognitively enriching activities and child outcomes (Del Boca et al., 2014; Fiorini & Keane, 2014; Hsin & Felfe, 2014). Therefore, understanding why these gaps in parental time investments in child skill development exist is important because they are closely related to the gaps in test scores.

We designed the Math for Parents and Children Together (MPACT) intervention to test the effectiveness of a low-cost, behaviorally informed treatment, an EdTech treatment, and informational treatments on parental engagement and child’s early math skills. MPACT was a 12-week intervention with four treatment groups and a control group (N=1,412). The treatments groups were: 1. A group that received the MKit, a booklet of math activities for parents to do with their children; 2. A group that received the MKit plus text messages intended to promote a growth mindset, which is the idea that children’s skills are malleable and therefore responsive to parental input; 3. A group that received the MKit plus text messages intended to overcome procrastination associated with present bias; and 4. A group that received a tablet loaded with math applications for children but not an MKit or messaging.

We assessed children’s math skills at baseline, at the end of the 12-week evaluation, 6 months after the evaluation ended, and 12 months after the evaluation ended for a subsample of children. We surveyed parents at baseline and immediately after the 12-week evaluation. Table 1 presents demographic characteristics of participant children as well as children in Head Start in Chicago. The majority of our sample is Hispanic and black non-Hispanic. Figure 1 presents baseline math scores measured by the Woodcock-Johnson IV Applied Problems (W-J). Children participating in MPACT scored more than a standard deviation below the national mean score. This is equal to about 27 percentile points below the national mean.

The results of the MPACT treatments on children’s test scores are as follows:

• At the end of the MPACT evaluation, children in the MKit + growth mindset treatment group had gained fewer percentile points on the W-J math assessment than children in the control group. No other treatment had a large or statistically significant effect on children’s assessment scores at the end of the evaluation.

• At six months after the end of the MPACT evaluation, children in the tablet group and in the MKit group had gained more percentile points than children in the control group. For both groups the difference is about 16% of the W-J standard deviation. Children in neither the MKit + growth mindset nor the MKit + present bias group had gained more than children in the control group at six months after the end of the evaluation.

Authors

• Susan Mayer, University of Chicago
  Non-Presenting Author

• Ariel Kalil, University of Chicago
  Non-Presenting Author

• William Delgado, University of Chicago
  Presenting Author