Alan-Ertac Grit and Growth Mindset Intervention Study 2 (replication)

How to cite

Alan, Sule. 2018. "Alan-Ertac Grit and Growth Mindset Intervention Study 2 (replication)." AEA RCT Registry. AEARCTR-0003327. https://doi.org/10.1257/rct.3327-1.0

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

Sule Alan*

* Primary investigator

Registry ID
AEARCTR-0003327
DOI
https://doi.org/10.1257/rct.3327-1.0
First registered
Last updated
Status
Completed
Trial dates
2015-04-01 – 2016-04-30
Countries
Turkey (Istanbul)
JEL codes
C91, C93, D03, I28
Registry record
http://www.socialscienceregistry.org/trials/3327

Abstract

The purpose of this trial is to replicate the findings of Alan-Ertac Grit and Growth Mindset Intervention Study 1 (AEA registry number: AEARCTR-0003317)

We evaluate the same program described in Trial AEARCTR-0003317 with respect to the same outcomes. The program involves an intensive teacher training program, supported by a specially designed curriculum. The curriculum consists of animated videos, mini case studies and classroom activities that highlight i) the plasticity of the human brain against the notion of innately fixed ability, ii) the role of effort in enhancing skills and achieving goals, iii) the importance of a constructive interpretation of failures and therefore perseverance, and iv) the importance of goal setting.

In this replication, we randomly assign the same grit intervention as in Trial AEARCTR-0003317 across a new set of schools in Istanbul. This sample consists of 16 schools. While the intervention follows the same procedures (same curricular materials and teacher training approach), there are a couple of important differences in the way the study was conducted. These changes were made to alleviate potential issues with the design of the first study. First, in the replication study, the treatment schools were not subject to any other treatment than grit. This allows us isolate the effect of the grit
intervention. Second, we administer our own math and verbal tests both at baseline and at
follow-up.

In addition to the experimental outcomes measuring grit and test scores, we evaluate this replication with respect to social preferences. In particular, we implement an altruism experiment, where we manipulate the donation context. Using a variant of the dictator game, we ask children whether they would like to donate a fraction of their experimental earnings (gifts) to anonymous children. A random half of both treated and control children are given the information that the recipient has no gifts because he/she failed at the real-effort task, while the other half are given the information that the recipient has no gifts because his/her school was not visited.

By implementing one context where the recipient’s poorness is certainly due to bad luck (not having been visited) and one where it may be due to a lack of effort, we compare the responsiveness of giving on the part of treated and untreated children to the potential reasons for the recipient’s poorness.
With respect to the effect of the program on experimental measures of grit as well as real outcomes, we replicate all our results documented in Study 1.

With respect to social preferences, we find that treated children, holding significantly more optimistic beliefs about the role of effort in success, are significantly less likely to donate their earnings to those who failed at a real effort task. These results highlight possible unintended social effects of growth mindset interventions.

Intervention

Teachers in the treatment group received full-day training on the delivery of the curriculum. This is the same Grit curriculum described in Study 1 (AEARCTR-0003317).

Experimental design

We sampled 16 schools and collected baseline information. We randomly allocated 8 schools to treatment, 8 to control. Teachers in the treatment group received training on our grit curriculum and nothing else. After the intervention ended, there were two follow-up data collection phases. In the first follow-up, experimental measures on grit, competition, and altruism were collected, along with objective test scores. In the second follow-up, longer-term effects on objective test scores were measured.

Design details

Randomization method
Teachers in designated schools were called in a random manner. A willing teacher was assigned 50% probability to be in treatment group. Using Stata’s uniform random number generator, we performed the assignment.
Randomization unit
School
Clusters
16 schools.
Observations
Official student number is 1499
Arms
Official student number for Treatment= 8 schools (683 students), Control= 8 schools (816 students), Total = 16 schools (1499 students). Official sample size corresponds to all registered students in participating schools. Sample size for experimental outcomes is lower because these outcomes were collected from students who were present at the time of the classroom visits.
Primary outcomes
We have two sets of outcomes: 1) Experimental outcomes 2) Achievement outcomes

Cited by

No citation data available for this registration yet.

Keywords