The effect of a pen pal program on child outcomes: Evidence from a randomized controlled trial in China

How to cite

Feng, Shuaizhang; Heckman, James; Han, Yujie. 2023. "The effect of a pen pal program on child outcomes: Evidence from a randomized controlled trial in China." AEA RCT Registry. AEARCTR-0011159. https://doi.org/10.1257/rct.11159-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Shuaizhang Feng* , James Heckman , Yujie Han

* Primary investigator

Registry ID
AEARCTR-0011159
DOI
https://doi.org/10.1257/rct.11159-1.0
First registered
Last updated
Status
On going
Trial dates
2021-12-05 – 2023-11-30
Countries
China (Sichuan Province)
Registry record
http://www.socialscienceregistry.org/trials/11159

Abstract

Our aim is to study the effect of pen pal programs on child development. We conduct a randomized controlled trail on students in grade 7 (first-year middle school students) in China, in collaboration with the BlueLetter, an NGO specialized in organizing pen pal programs for many years. The program matches each student in the treatment group with a pen pal volunteer, and allow them to communicate by mail for one and half years anonymously and under the supervision of BlueLetter. We plan to (1) examine the causal effect of pen pals on a range of development outcomes, including test scores, cognitive skills, and noncognitive skills (Big Five personlity factors) as well as economic preferences, both immediately after the program ended and longer term effects; (2) study heterogeneous effects for different subgroups. In particular, we will focus on left behind children whose parents were away from home for work in cities. Left behind students are especially likely to suffer from emotional issues, thus could potentially benefit from pen pal programs which provides emotional support; (3) analyze the underlying mechanisms through which the program impacts children. We have collected extensive information regarding the children, their parents and schools before the intervention, and will do so after the RCT. Meanwhile, we will analyze the text contents of the letters exchanged between the student and pen pal volunteers.

Experimental design

We sample all 9 middle schools in Mianzhu county of Sichuan province in China, each of which includes 4-14 classes in grade 7. In the grade 7 of each school, we randomly assign half of the classes in each school into either the treatment or control group. As a result, there are 33 and 31 classes in the treatment group and control group, respectively. For all students in the treated classes, we match them with a pen pal volunteer with whom to have mail correspondence for about one and half years.

Design details

Randomization method
random sampling of classes using stata program on a computer.
Randomization unit
class
Clusters
64 classes
Observations
1952 students
Arms
33 treated classes and 31 controlled classes
Primary outcomes
Academic achievement measured by test score of Chinese, math, and English. Cognitive skills measured by IQ. Noncognitive skills measured by the Big Five personality traits. Economic preferences measured by time preference, risk preference, and altruism.

Cited by

No citation data available for this registration yet.

Keywords