NEW UPDATE: Completed trials may now upload and register supplementary documents (e.g. null results reports, populated pre-analysis plans, or post-trial results reports) in the Post Trial section under Reports, Papers, & Other Materials.
Parents’ Beliefs and Children’s Education: Experimental Evidence from Malawi
Initial registration date
April 13, 2018
April 17, 2018 2:18 PM EDT
University of Chicago
Other Primary Investigator(s)
Additional Trial Information
Information about children's school performance appears to be readily available. Do frictions prevent parents, particularly poor parents, from acting on this information when making decisions? I will conduct a field experiment in Malawi to test the following research questions: First, are parents' baseline beliefs about their children's academic performance inaccurate? Second, does providing parents with clear and digestible academic performance information cause them to update their beliefs and correspondingly adjust their investments? Finally, if these frictions exist, are they worse among the poor?
I will provide academic performance information to randomly selected parents and measure the effects on educational investments and decisions. Specifically, I will provide the treatment group with report cards describing the academic performance of two of their children (one report card per child). The reports will show children’s performance on all test administered in the most recent school term, specifically: the percent score, the corresponding grade on the Malawian grading scale, and the within-class percentile ranking for math, English, and Chichewa.
Intervention Start Date
Intervention End Date
Primary Outcomes (end points)
(1) "Experimental outcomes" (outcomes measured as part of the experiment):
• Choice between three levels of difficulty for free math and English textbooks: beginner, average, or advanced
• Willingness to pay for remedial textbooks in math and English
• Parental allocation of nine lottery tickets across two children, in which lottery prize is four years of government secondary school fees for one child in every 100 households
(2) Non-experimental outcomes:
• Dropouts and expenditures, collected from a second endline survey of parents 1 year after intervention; • Administrative data on attendance 1 month after intervention
Primary Outcomes (explanation)
Secondary Outcomes (end points)
Transfers across schools and non-monetary investments
Secondary Outcomes (explanation)
The study will work with 39 schools in two districts (Machinga and Balaka) in Malawi. Schools will be selected randomly from the universe of primary schools, oversampling schools with high and low expected levels of parent education to increase heterogeneity in parent education within the sample. Since one of the outcomes to be examined is inter-sibling tradeoffs, multiple-sibling households will be used as the sampling frame.
Based on the test score data from schools and sibling data, a sample of households with at least two children enrolled in grades 2-6 with test score data will be drawn. For households with more than two children, two will be randomly selected. I will randomly assign half the households in the sample to a treatment group that receives information about children’s test scores, and half to a control group, which will not. Note that half the treatment groups will also be assigned to receive an add-on intervention designed to test a hypothesis intended for another study: that providing more detailed information would increase parental engagement, as measured through monetary and non-monetary investments. This group will receive additional skill information (e.g. whether child can add 3-digit numbers). In my primary analysis, I will ignore the add-on treatment and pool the treatment households. The randomization will be stratified on a test score measure (between sibling score gap) and a proxy for parent education (estimated literacy rate in household’s village).
Experimental Design Details
Randomization done in office by a computer.
Was the treatment clustered?
Sample size: planned number of clusters
Sample size: planned number of observations
5,268 parent/child pairs (2,634 households with 2 observations per household)
Sample size (or number of clusters) by treatment arms
Control: 1,327 households
Treatment: 1,307 households
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
INSTITUTIONAL REVIEW BOARDS (IRBs)
IRB Approval Date
IRB Approval Number
Post Trial Information
Is the intervention completed?
Intervention Completion Date
July 31, 2012, 12:00 AM +00:00
Is data collection complete?
Data Collection Completion Date
July 31, 2013, 12:00 AM +00:00
Final Sample Size: Number of Clusters (Unit of Randomization)
Was attrition correlated with treatment status?
Final Sample Size: Total Number of Observations
Final Sample Size (or Number of Clusters) by Treatment Arms
1327 control, 1307 treat
Reports, Papers & Other Materials
Information about children’s school performance appears to be readily available. Do frictions prevent parents, particularly poor parents, from acting on this information when making decisions? I conduct a field experiment in Malawi to test this. I find that parents’ baseline beliefs about their children’s academic performance are inaccurate. Providing parents with clear and digestible academic performance information causes them to update their beliefs and correspondingly adjust their investments: they increase the school enrollment of their higher-performing children, decrease the enrollment of their lower-performing children, and choose educational inputs that are more closely matched to their children’s academic level. These effects demonstrate the presence of important frictions preventing the use of available information, with heterogeneity analysis suggesting the frictions are worse among the poor.
Dizon-Ross, Rebecca. "Parents’ Beliefs About Their Children’s Academic Ability:
Implications for Educational Investments." Working Paper, April 2018.
REPORTS & OTHER MATERIALS