The Effects of Information Provision to Parents on Student Outcomes – A Randomised Controlled Trial

Schools are assigned to the four experimental conditions: control and three treatment conditions.

2018-10-01

2018-11-30

We measure financial literacy by a test based on nine questions. The questions measure financial knowledge as well as financial behaviour.

The financial literacy test also measures financial attitudes and intertemporal preferences.

Schools that registered for participation were randomised to the aforementioned four experimental conditions. We assessed the level of financial literacy of all students before as well as after followed the course. Students assigned to the control group completed the same tests at the same time as students in the treatment groups, without receiving any treatment between the tests.

Students in the control group completed a pre-test at the start of the experiment and a post-test after four weeks, but did not receive any treatment. Students in all treatment groups completed the same pre-test before the intervention and the same post-test at the end of the classroom intervention.
Students in all treatment groups received a homework assignment before the classroom intervention. Students in the first treatment group (‘class’) received a standard homework assignment without parental involvement. Students in the second treatment group (‘parental involvement’) and in the third treatment group (‘information’) received a modified version of the homework assignment with parental involvement. Students in the third treatment group (‘information’) were given before the class an information brochure for their parents.
Students in all treatment groups followed a 4-hour financial education class on saving and investing. The course material was designed as an adaptive online learning path that students completed in groups of two. All course material was developed by high school teachers, tailored to the age and the ability of the target group. After the classroom intervention and the post-test at school, the intervention ended for the first treatment group. In the second and third treatment group, students received a second homework assignment which consisted of a family task to be completed with a parent.
Parents of the students in all treatment groups received a pre-test after the student completed the pre-test at school. A post-test for parents was given to parents of students in all treatment groups after students completed the post-test at school.

Schools were randomly assigned to the different experimental conditions by a random number generator in STATA.

The treatment was clustered at school-level.

Yes

28 schools
Given that the intervention has already been completed and data have been collected for both the pre-test and the post-test, the numbers reported refer to the observations for which both the pre-test and the post-test are available, i.e. the final sample for the analysis. The calculation of the minimal detectable effect size is also based on this final sample.

1,253 students

Control group = 305 pupils, 9 schools
Treatment group 1 = 418 pupils, 10 schools
Treatment group 2 = 209 pupils, 4 schools
Treatment group 3 = 321 pupils, 5 schools
Average number of schools per condition = 7
Average number of pupils per condition = 313.25
Average number of pupils per school = 44.75

The computation is based on List et al. (2011) and accounts for intracluster correlation in the calculation of the minimal detectable effect size. In our experimental setting, there are on average 7 schools in each experimental condition. Each school has on average 44.75 students. Computation in Stata based on the post-test shows that the intracluster correlation in the final sample equals 0.1. In the analysis, this intracluster correlation can be reduced by controlling for baseline characteristics of schools and students. Using the conventional power of 0.8 and a significance level of 0.1, the calculation results in a minimal detectable effect size of 0.52 standard deviations in case we would not control for students’ characteristics. Details of the calculation: According to List et al. (2011), in a clustered design, the minimum number of observations in each experimental group can be computed as follows: n=2(t_(α/2)+t_β)²(σ/δ)²(1+(m-1)ρ) This implies that the minimum detectable effect size is equal to: δ=σ/√(n/(2(t_(α/2)+t_β)²(1+(m-1)ρ))) Or the minimum detectable effect size expressed as a fraction of a standard deviation is equal to: δ/σ=1/√(n/(2(t_(α/2)+t_β)²(1+(m-1)ρ))) δ/σ=1/√(313.25/(2(1.96+0.84)²(1+(44.75-1)0.1)))=0.52 Reference List, J., Sadoff, S. and Wagner, M. (2011), So you want to run an experiment, now what? Some simple rules of thumb for optimal experimental design, Experimental Economics 14, 439-457

Targeted parental involvement is increasingly employed to activate parents to use their supporting role to enhance learning effects of students. Given the low cost, one popular measure is information provision to parents. However, the evidence on such measures remains limited. This study provides causal evidence on the effects of information provision to parents in a financial education course where parental involvement was stimulated. Based on a randomised controlled trial with 1253 students from grade 8 and 9 in Flanders, we identify the effects of the classroom intervention separately from parental involvement prompted in homework and the provision of subject-specific information to parents on students’ learning outcomes. The estimates reveal that, overall, the treatment effectively increases financial knowledge and skills. However, despite its popularity in schools, the value added of prompting parental involvement and providing parents information is not significant overall, but only for financial skills.

Maldonado, J.E., and De Witte, K. (2021). The Impact of Information Provision to Parents: Experimental Evidence on Student Outcomes. Journal of Behavioral and Experimental Finance 31.

https://doi.org/10.1016/j.jbef.2021.100548