Back to Top

Parent Skills and Information Asymmetries: Experimental Evidence from Home Visits and Text Messages in Middle and High Schools

By: Peter Bergman, Chana Edmond-Verley, and Nicole Notario-Risk

Economics of Education Review (volume 66, 2018)

Published version

Manuscript version (free)

Commentary by Gabriel Heller-Sahlgren

Parents’ investments in their children’s education are likely to play an important role for their outcomes. Such investments may be monetary, such as payments for private schooling or tutoring, but also include time spent on promoting outcomes, such as helping out with homework and making informed school choices.

Yet parents may also lack the skills and information necessary to help their children in the best way possible. For example, they may not be able to assist with homework as it becomes more complex at older ages. They may also be ill-informed about their children’s performance and therefore unaware of the support they need. If so, interventions seeking to increase parental skills and information may have positive effects on pupil outcomes.

In this paper, Peter Bergman and colleagues report the results from a one-year long randomised-controlled trial, which tested whether text messages and home visits could decrease information problems and improve skills among parents in ways that improve pupil outcomes. In total, 1,120 families with children attending a low-performing middle school or one of two low-performing upper-secondary schools in an anonymous low-income, urban school district in the American Midwest were randomly assigned to one of two interventions or a control group.

In one intervention, parents received text messages with detailed information on their children’s academic performance twice a month. In the second intervention, parents received these text messages as well as home visits to improve their ability to use school resources to track pupil progress, understand graduation requirements and university readiness, and improve the learning environment at home. Parents in the control group received neither text messages nor home visits.

Both treatments were delivered by members of community-based organisations, who were trained by the researchers to interpret pupil data, transmit that data in text messages to parents, phone calls, or emails, and to carry out home visits. Essentially all parents who were supposed to receive text messages did so, whereas, due to difficulties in scheduling, just over 50 per cent of parents who were supposed to receive home visits received at least one such visit.

Both treatments had large positive effects on the probability that pupils remained in the school district – by about 40 per cent compared with the control group – in the spring term in the school year during which the interventions were carried out. As pupils who leave the district are lower-performing relatively speaking, the results suggest the interventions helped retain some of the most at-risk pupils. Yet conditional on these effects, the interventions had no impact on absences or suspensions. Also, effects on district retention from one school year to the next are smaller and only statistically significant in the text-message treatment.

But did the interventions affect pupil achievement? Yes, they did. The text-message intervention increased pupils’ overall grade average by the equivalent of 13 PISA points compared with the control group, whereas the text-message plus home-visit intervention raised the average by the equivalent of 8 PISA points.

While only the former is statistically significant, the differences between the two treatments are not statistically significant so it is not possible to say that the text-message only intervention was more effective than the one that also included home visits.

At the same time, only the text-message plus home-visit intervention had effects on standardised test scores in reading and mathematics, raising pupil achievement by the equivalent of 12-13 PISA points.

In this case, the differences between the two treatments are also statistically significant, suggesting that the intervention that included both information and home visits was in this respect more effective than the one that only included information. This is especially true since only 50 per cent of families who were supposed to receive home visits actually did so; the effect sizes of actually receiving both text messages and at least one home visit amounts to 26 PISA points in mathematics and 19 PISA points in reading.

Interestingly, however, the effects are heterogeneous: learning gains appear to accrue mostly to pupils with a higher-than-average initial grade average in the sample, whereas retention effects also accrue to those with lower initial performance. Also, among initially higher-performing pupils, the text-message plus home-visit intervention increased attendance rates and decreased suspensions – but also appears to have had negative effects on chronic absences among lower-performing pupils. Still, even the higher-performing pupils had baseline mathematics and reading scores in the lowest 30th percent nationally, so it appears the clear the interventions benefited low-performing pupils in an absolute sense.

Overall, therefore, the interventions appear to have been successful, although in different ways. The per-pupil revenue increase for the district as a result of increasing pupil retention outweigh the total cost of the program by a considerable margin, and combined with the learning gains the authors find that the interventions pass a cost-benefit test, suggesting that similar information interventions may be a useful way to improve outcomes.

Still, it is difficult to extrapolate these findings to other contexts, especially more advantaged settings. It’s also not clear whether home visits by themselves would pass a cost-benefit test. The main benefit of the additional treatment arm appears to be larger positive effects on standardised tests, whereas the effects on the grade average and retention appear slightly larger in the text message-only treatment – and some research suggests that grade averages are more important than test scores for pupils’ long-term outcomes. It would certainly be fruitful to investigate these issues further in future research.


Gabriel Heller-Sahlgren is CfEE’s Lead Economist.

The research discussed here is also the ‘Editor's choice’ in the first (January 2019) issue of CfEE’s new Monthly Research Digest, of which he is the editor.

View it in flipbook, download a pdf, and subscribe here.

Blog Category: 
About the author