Growth mindset intervention and the impact of conditional cash transfers on poor students

Last registered on April 22, 2021

Pre-Trial

Trial Information

General Information

Title
Growth mindset intervention and the impact of conditional cash transfers on poor students
RCT ID
AEARCTR-0007579
Initial registration date
April 19, 2021

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published
April 20, 2021, 6:37 AM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Last updated
April 22, 2021, 12:10 AM EDT

Last updated is the most recent time when changes to the trial's registration were published.

Locations

Region

Primary Investigator

Affiliation
Khon Kaen University

Other Primary Investigator(s)

PI Affiliation
Khon Kaen University
PI Affiliation
University College London
PI Affiliation
Universidad De Carlos III Madrid

Additional Trial Information

Status
In development
Start date
2021-01-15
End date
2022-03-03
Secondary IDs
Abstract
Evidences from recent studies point that households across socio-economic status differ in their view on future outlooks of their children – both of education and labour market achievement. That is, more well-off families, including parents and children themselves, tend to set higher expectations on returns to education and have higher career aspirations. In contrast, disadvantaged families are found not only to have lower income expectations but they are also constrained by a narrow view of career options. The disparity may contribute to more stagnant social mobility as limited information and low aspirations may lead to sub-optimal decisions over education and career choices. The experimental design intends to reduce the gaps of aspirations and income expectations across families of different socio-economic status by conducting a randomized controlled trial at school-level primary and middle school students in Thailand. The trial with the student subjects has two main intervention arms consisting of two semester-long, in-class interactive curricula on: (i) a course on career perceptive, and (ii) a course on Growth Mindset. We plan to measure skills (cognitive and non-cognitive), as well as subjective beliefs regarding educational and occupational choices before and after the intervention. And we will empirically test if our interventions alter the measured outcomes. Moreover, as in many developing countries, children’s beliefs and outcomes are highly influenced by parents. Therefore, the intervention also extends to a randomly selected group of parents who receive both curricula. Specifically for the parent subjects, we further random the assignment of a financial incentive to test whether a conditional reward (on student’s academic improvement) works differently from unconditional rewards.
External Link(s)

Registration Citation

Citation
Jirasatthumb, Norachit et al. 2021. "Growth mindset intervention and the impact of conditional cash transfers on poor students." AEA RCT Registry. April 22. https://doi.org/10.1257/rct.7579-1.1
Sponsors & Partners

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information
Experimental Details

Interventions

Intervention(s)
The experimental design intends to encourage educational and occupational aspirations as well as alleviate biases in subjective beliefs regarding returns to education among poor households. To do this, the design is to test whether providing information regarding career perceptive, or providing behavioural information on the malleability of intelligence along the principle of ‘Growth Mindset’ work more effectively in altering the belief biases.
The first target groups of the trial are students in two different grades, grade 4 and 7, enrolled in public schools in 3 provinces in the northeast of Thailand (Khon Kaen, Udon Thani, and Kalasin). The second target groups of the trial are parents of the student participants.
The trial with the student subjects has two main intervention arms consisting of two semester-long, interactive curricula on: a course on career perceptive, and a course on Growth Mindset. Students are assigned, at school-level, to receive three types of treatments: (i) both courses, (ii) only the course on career perceptive, and (iii) only the course on Growth Mindset. Students watch a video-based lesson in class under a teacher supervision and they follow the lessons with an accompanied lesson booklet. There are 8 lessons in total, which is extended throughout the semester.
The trial with the parent participants will randomly assign parents to receive both curricula. The parents will be invited to attend 5 sessions at the school and they follow the lesson with an accompanied lesson booklet. All parents in treatment group receive the same lessons. The additional intervention for parent participant in the treatment group is on the design on the financial reward, which may be conditional on the academic improvement of their children: (i) receive the lessons but no financial incentive, (ii) receive the lessons and receive unconditional financial reward prior to the start of the intervention with students, and (iii) receive the lessons, but may receive financial reward after the intervention is done conditionally on their children’s test performance.
Intervention Start Date
2021-05-17
Intervention End Date
2021-09-30

Primary Outcomes

Primary Outcomes (end points)
The primary outcomes for the student participants are the followings:
1. Students’ academic performance before and after the interventions. We measure academic performance from two modules: the cognitive module, and the non-cognitive module.
1.1. The cognitive module, both before and after tests use self-administered, grade-based cognitive test, which are aligned with Thailand’s centralised national examination. The test contains 15 multiple choices on language aptitude (Thai language), and 15 multiple choices on mathematic aptitude. Students take the test in class and are given 60 minutes to complete the test.
1.2. For the non-cognitive module, both before and after tests use two sets of questionnaires.
(A) the Thai-translated version of the Strength and Difficult Questionnaire (Goodman, 1997), provided by Thailand’s Ministry of Public Health (MoPH). The test is age-suitable and is used as a brief, self-reported behavioural screening questionnaire. The questionnaire contains 25 questions, with additional 4 questions added in the MoPH’s version to enquire further details on how might behavioural-emotional status affect a daily life. Students take the test in class and are given 60 minutes to complete the test.
(B) the Dweck Mindset Instrument. The questionnaire contains 8 questions and is adapted from Dweck, C. S. (2006), which surveys participants’ perception on the malleability of intelligence and behavioural traits. The Dweck Mindset Instrument is probed as a part of the surveys conducted before and after the intervention.
2. The responses of the quizzes taken immediately after each lesson finishes. There are 5 to 10 quizzes in each lesson, and students participate in 8 lessons in total.

The primary outcomes for the parent participants are the followings:
1. The responses in the quizzes taken immediately after each lesson finishes. There are 5 to 10 quizzes in each lesson, and parents participate in 5 lessons in total.
2. the Dweck Mindset Instrument. The questionnaire contains 8 questions and is adapted from Dweck, C. S. (2006), which surveys participants’ perception on the malleability of intelligence and behavioural traits. The Dweck Mindset Instrument is probed as a part of the surveys conducted before and after the intervention.
Primary Outcomes (explanation)
For the cognitive module, we will calculate the total score from the correct answers and convert the raw total score into a grade-specific standardised score (among the sample in our trial). The total scores will be available separately for each subject (Language, Maths), as well as the aggregate score of both subjects.
For the non-cognitive module, we follow the official guideline of how to use the SDQ and convert the itemised responses into 5 core scores (Conduct, Emotional, Peer Relation, Hyperactivity, pro-social). The scores from Conduct and Hyperactivity will also be grouped as ‘externalised behaviour score’, and the scores from Emotional and Peer Relation will be grouped as ‘internalised behaviour score’. We will use both the raw scores and the standardised scores in order to measure the changes of this non-cognitive measure before and after the intervention.
For the non-cognitive module, we follow the official guideline of how to use the SDQ and convert the itemised responses into 5 core scores (Conduct, Emotional, Peer Relation, Hyperactivity, pro-social). The scores from Conduct and Hyperactivity will also be grouped as ‘externalised behaviour score’, and the scores from Emotional and Peer Relation will be grouped as ‘internalised behaviour score’. We will use both the raw scores and the standardised scores in order to measure the changes of this non-cognitive measure before and after the intervention.
For the lesson-base quizzes, we will calculate the total correct responses within each ‘module’ (growth-mindset, career perceptive, controlled-lesson) at 3 time periods: before, immediately after each lesson, and one academic term after the lessons end. We will track the changes in each individual’s score over the 3 periods in order to evaluate if students and parents who receive the curricula maintain or alter their knowledge on the issue, and to what extent the knowledge lasts for a given duration.

Secondary Outcomes

Secondary Outcomes (end points)
The secondary outcomes are commonly measured student and parent participants in the trial and are probed during a short survey, which is conducted before and after the interventions:
1. Subjective expectations of income along different education attainments for self (student) or for children (parent), for a top-level student, for a low-level student
2. Educational aspirations for self (student), or for children (parent).
3. Occupational aspirations for self (student), or for children (parent).
4. Time-use in a given day (weekend and weekday are reported separately)
5. Expenditure in a given month, with the focus on education-related expenditure
6. Frequency of communication on issues related to future plans (career and occupation)
Secondary Outcomes (explanation)
1. For all secondary measures, the analysis will compare the responses given before and after the intervention, and thus calculate the changes of each itemised measure.
And there are additional indicators:
2. For subjective expectations of income: we construct the income trajectory along different educational attainment (from low education to high education), and calculate the slope (or the shape) as an additional indicator. Therefore, we will also consider the changes in the slope of the expected income trajectory before and after.
3. We pair up the responses given by the student and the parent and check for the discrepancy (‘gap’) in the response for each household. Therefore, the ‘household gap’ of income expectation is an indicator.
4. Educational aspirations: We plan to additionally convert education level into its equivalent years of schooling in the regression analysis so as to obtain a cardinal measure. In addition, we pair up the responses given by the student and the parent and check for the discrepancy (‘gap’) in the response for each household. Therefore, the ‘household gap’ of income expectation is an indicator.
5. Occupational aspirations: We plan to additional convert occupation into an ordinal format using a occupational prestige score (for example CAMSIS score) or relate the occupation to the average earnings (which can be derived from the Thailand’s Labour Force Survey). In addition, we pair up the responses given by the student and the parent and check for the discrepancy (‘gap’) in the response for each household. Therefore, the ‘household gap’ of occupational aspirations is an indicator.
6. Time-use in a given day: We use the raw number of minutes spent in each activity (8 listed groups of activities) and the standardised values for the time spent in a given weekend and a given separately.
7. Expenditure in a given month. We use the raw value of expenditure in each item (10 listed items), as well as the standardised values.
8. Frequency of communication on issues related to future plans: We use the frequency reported in each activity (7 items) to calculate the pattern of communication based on the partners (no one, with parents, with teacher, with friends)

Experimental Design

Experimental Design
To identify the effects of these courses on our targeted outcomes, we create a multi-arm of study involving four cross-cutting sets of treatments, three sets of treatments for students only and one set of combination treatments between students and parents. Three sets of treatments for students consist with 1) receiving growth mindset course only treatment, 2) receiving career perspective course only treatment, and 3) receiving both growth mindset and career perspective courses treatments. A set of combination treatments between students and parents is a combination between students and parents receiving both curricula. In addition to curricula, two types of incentive are also tested in this arm. These two types of incentives are financial conditional and unconditional rewards. For the former one, parents will receive reward if scores from standardize tests after the intervention of their children are improved by at least 30 percent compared to those before the intervention. On the other hand, for unconditional reward, parents will receive reward after the end of intervention no matter whether standardize tests scores of their children are improved or not. From this design, there are therefore seven experimental groups in total, six treatment groups and one control group.
The sample was randomized from public schools in Khon Kaen and Udonthani provinces. The criteria for school selection are as follow. Schools with the number of students in grade 4 and/ or grade 7 and the number of extra-poor students in grade 4 and/ or 7 at least at the average of all schools in the province were kept for randomization. These schools were then grouped by districts in each province.
Districts with less than 7 schools (determined by the number of experimental groups) were dropped out from our school list. From these criteria, 346 schools were eligible for randomization, which can be divided by province as follows: 201 schools in Khon Kaen and 145 schools in Udon Thani. Due to budgetary constraints and study duration, the research team sampled 245 schools from both provinces as samples for the study. Khon Kaen province has 145 schools or approximately 58 percent of the total schools in the study. And Udon Thani has 100 schools or approximately 40 percent of the total number of schools in the study. In addition, the research team added 6 schools from Kalasin province to be the sample resulting in 251 schools in total.

The sample size (schools, students, and parents) by grade is the following (These numbers could be changed after the interventions implemented.):
Treatment 1 (TS1): Students receive a course on career perspective, no parent intervention
• Schools in TS1: schools with grade 4 only= 14, schools with grade 7 only = 2, schools with grade 4 and 7=17
• Students in TS1: Grade 4 = 837, Grade 7 = 994
• Parents in TS1: 0
Treatment 2 (TS2): Students receive a course on Growth Mindset, no parent intervention
• Schools in TS2: schools with grade 4 only = 14, schools with grade 7 only = 4, schools with grade 4 and 7=18
• Students in TS2: Grade 4 = 909, Grade 7 = 1216
• Parents in TS2: 0
Treatment 3 (TS3): Students receive a course on career perspective and on Growth Mindset, no parent intervention
• Schools in TS3: schools with grade 4 only = 15, schools with grade 7 only = 8, schools with grade 4 and 7=17
• Students in TS3: Grade 4 = 1001, Grade 7 = 1803
• Parents in TS3: 0
Treatment 4 (TS4): Students receive a course on career perspective and on Growth Mindset, parents receive a course on career perspective and on Growth Mindset, parents receive no financial incentive
• Schools in TS4: schools with grade 4 only = 13, schools with grade 7 only = 5, schools with grade 4 and 7=18
• Students in TS4: Grade 4 = 736, Grade 7 = 1285
• Parents in TS4: Grade 4 = 166, Grade 7 = 243
Treatment 5 (TS5): Students receive a course on career perspective and on Growth Mindset, parents receive a course on career perspective and on Growth Mindset, parents receive conditional financial incentive
• Schools in TS5: schools with grade 4 only = 14, schools with grade 7 only = 4, schools with grade 4 and 7=15
• Students in TS5: Grade 4 = 772, Grade 7 = 1115
• Parents in TS5: Grade 4 = 171, Grade 7 = 170
Treatment 6 (TS6): Students receive a course on career perspective and on Growth Mindset, parents receive a course on career perspective and on Growth Mindset, parents receive unconditional financial incentive
• Schools in TS6: schools with grade 4 only = 10, schools with grade 7 only = 8, schools with grade 4 and 7=16
• Students in TS6: Grade 4 = 885, Grade 7 = 1575
• Parents in TS6: Grade 4 = 208, Grade 7 = 300
Control 1 (C1): No interventions on students and on parents
• Schools in C1: schools with grade 4 only = 14, schools with grade 7 only = 7, schools with grade 4 and 7=19
• Students in C1: Grade 4 = 876, Grade 7 = 2051
• Parents in C1: Grade 4 = 188, Grade 7 = 237
Experimental Design Details
Randomization Method
Schools in each district are randomly assigned to experimental groups. Randomization are done using computer where numbers 1 to 7, represented experimental groups, are randomly given to each school.
Randomization Unit
The randomization takes place at the school level.
Was the treatment clustered?
Yes

Experiment Characteristics

Sample size: planned number of clusters
251 schools.
Sample size: planned number of observations
Total students in the trial would be 16,082 students; Total parents in the trial would be 1,683 parents.
Sample size (or number of clusters) by treatment arms
The sample size (schools, students and parents) by grade is the following:
• Treatment 1 (TS1): 16 schools, 1831 students, 0 parents
• Treatment 2 (TS2): 18 schools, 2125 students, 0 parents.
• Treatment 3 (TS3): 23 schools, 2804 students, 0 parents
• Treatment 4 (TS4): 18 schools, 2048 students, 409 parents.
• Treatment 5 (TS5): 18 schools, 1887 students, 341 parents.
• Treatment 6 (TS6): 18 schools, 2460 students, 508 parents.
• Control 1 (C1): 21 schools, 2927 students, 425 parents.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
IRB

Institutional Review Boards (IRBs)

IRB Name
The Khon Kaen University Ethics Committee in Human Research
IRB Approval Date
2021-03-25
IRB Approval Number
HE633012

Post-Trial

Post Trial Information

Study Withdrawal

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Intervention

Is the intervention completed?
No
Data Collection Complete
Data Publication

Data Publication

Is public data available?
No

Program Files

Program Files
Reports, Papers & Other Materials

Relevant Paper(s)

Reports & Other Materials