Parent-Child Information Frictions & Human Capital Investment: Evidence from a Field Experiment
Key Finding
Parents tend to be overconfident about their child’s effort in school, and more frequent updates to parents can boost student effort and achievement.
When parents of students in a Los Angeles neighborhood were randomly selected to receive detailed, biweekly information about their child’s grades and missed assignments, their children missed fewer assignments and performed better academically. Automating this intervention could significantly boost student achievement for about $10 per student.
. . .
Other Findings
Nearly all parents underestimate how many assignments their child has missed. The more they underestimate missed assignments, the lower their child’s grade tends to be.
When parents are given more information, they monitor their children more. In turn, children put in more effort at school and earn higher grades.
Parents who received the study’s regular updates contacted the school 83% more often than those who did not. Parent-teacher conference attendance among this group also increased by 53%.
Achievement gains are on par with those achieved by the introduction of high-quality charter schools. Children of parents who received regular, detailed updates on grades and missed assignments missed 28% fewer classes, completed 25% more assignments, and earned higher GPAs and scores on standardized math tests.
. . .
Methodology & Data Highlights
Field experiment in a Los Angeles middle school, 2010-2011 school year.
Structural model to understand parent-child interactions.
Telephone survey of parents and caregivers to gather additional data. [1]
Summary
Parents want to help their children succeed in school, but doing so can become increasingly difficult as their child gets older. To motivate and guide their child, a parent needs information about what’s happening at school. When their child is old enough to withhold or manipulate that information, what’s a well-meaning parent to do?
This paper shares learnings from a field experiment in which parents of students at a Los Angeles school were randomly selected to receive frequent, detailed updates about their child’s grades and missed assignments. Studying school communication with parents in this population is especially important as low-income families are more likely to juggle irregular work schedules or encounter other barriers that make it more challenging to maintain regular communication with their child’s school.
Learning Collider founder Peter Bergman conducted this experiment and follow-up surveys with participating parents and students. He then developed a model of parent-student interactions where providing this information to parents significantly improves student effort and academic performance at a very low cost to schools.
These results, which for academic achievement are similar in magnitude to the introduction of high-quality charter schools, are driven by correcting a common parental bias: Parents often overestimate their child’s academic performance in school. With more information, however, they’re able to monitor and motivate their child and help them earn better grades.
Details on the field experiment
To conduct this experiment, Bergman randomly selected parents in a Los Angeles school to receive emails, text messages and phone calls listing students’ missing assignments. The information provided to parents came from teacher grade books gathered weekly from teachers. The gradebook information on grades and missed assignments, including details like specific pages of missing work, was pushed to parents several times a month over a six-month period.
Bergman then gathered data from the school on assignment completion, work habits, cooperation, attendance, and test scores. Finally, additional data was gathered via separate phone surveys to both parents and students at the end of the school year.
Parents want information–the intervention helped them get it
In parent surveys, Bergman found that parents in both the treatment and control groups care about their child’s schoolwork, asking their child whether they have completed their schoolwork nearly five times per week, on average. But there was an "information friction" at play: parents cared about their child's performance at school, but they could only get information from an unreliable source – their child.
Correcting this information friction by sending parents regular, detailed information increased two-way communication between parents and schools and translated into more effort on behalf of students.
When asked how often they’d heard from their child’s school in the last month, parents in the treatment group reported nearly twice as many updates, on average, than parents in the control group who weren’t receiving automated messages. Bergman’s findings also show that parents receiving the messages increased their own contact with the school. The treatment group contacted the school 83% more times, likely contributing to a 53% increase in the school’s overall parent-teacher conference attendance rate.
Student gains on par with the introduction of a charter school
As their parents increased monitoring of their children, the children did better in school, missing 28% fewer classes and completing 25% more assignments. Further, their likelihood of being cited for unsatisfactory work habits and unsatisfactory cooperation also decreased by around 25%.
Measures of academic achievement also went up significantly. The average GPA of students whose parents received messages was .19 standard deviations higher than students in the control group. Though there was no significant effect on English standardized test scores, math standardized test scores were .21 standard deviations higher.
Bergman posits there are several reasons for the limited effect on English standardized test scores, noting that math teachers in the experiment provided more frequent information on assignment completion, therefore allowing for more messages to parents. He also notes that 30% of students were classified as “limited-English proficient. ” While these students tended to perform better in English than math, the potential for increased information to parents to boost English scores for students with limited proficiency may be smaller.
Figure 3 from Parent-Child Information Frictions & Human Capital Investment
Overall, the achievement effects are comparable to the effects of introducing high-quality charter schools. Previous research on the nonprofit Harlem Children’s Zone suggests the program increased math scores by .23 standard deviations and English scores by .05 standard deviations. Research on KIPP Lynn charter schools suggests the schools increased math scores by .35 standard deviations and English scores by .12 standard deviations.
Why did it work?
To better understand how, specifically, the provision of more information generated results, Bergman built a model to measure when and how more information led to increased student effort. Bergman’s model suggests that the biggest driver of increased student effort was the fact that parents could more easily monitor their child’s progress (54% of the effect). Also important was the fact that the information parents received changed their previously-held beliefs about their child’s performance (42% of the effect).
Bergman then uses the model to study the potential effects of policies that target information frictions other than those described above. He finds that a policy that improves the quality of school reporting could also boost student effort, but only by half as much as the intervention he tested in Los Angeles: Simply providing more information.
Big academic gains for a very low cost
In Bergman’s experiment, messages to parents were written and sent manually. Bergman estimates that paying teachers to send these messages would cost schools (primarily in teacher overtime) $156 per student to achieve a .10 standard deviation increase in GPA or math scores.
If automated, Bergman estimates the Los Angeles intervention would cost about $10 per student to achieve the same academic gains. Other policies, which often give financial incentives to students and teachers, cost upwards of $500 per student to achieve the same results.
Footnotes:
[1] Learning Collider and its researchers use the term parents as inclusive of all caregivers and/or legal guardians in students’ households.