|
Field
Trial Title
|
Before
Improving parent-teacher communication on student well-being
|
After
Understanding Teacher Decision-Making: Parent-Teacher Alignment in Multidimensional Skill Development
|
|
Field
Trial End Date
|
Before
March 31, 2024
|
After
June 30, 2025
|
|
Field
Last Published
|
Before
January 02, 2024 10:56 AM
|
After
March 12, 2025 05:14 PM
|
|
Field
Intervention (Public)
|
Before
Within each grade-school pair (e.g., within 1st grade sections at a given school), class teachers will be randomly selected to be in the treatment or control, with equal probability. The intervention is an information treatment. Immediately after the baseline survey, I will provide treatment teachers information about their student capabilities and QOL collected from the parent survey. I have developed a web-portal through which the information will be accessed by teachers. I will provide teachers individual log-in information, through which they will have access to reports for students in their classroom, and no others.
|
After
The intervention is an information treatment delivered through a custom website (wiseteaching.net). Treated teachers receive:
Individual student reports showing parent preferences and current ability ratings (0-100) across nine dimensions of human capital grouped into academic, social, and emotional categories
Classroom-level summaries of parent priorities
Pedagogical strategies for improving non-cognitive skills based on feedback from teachers
Teachers access this information through individual login credentials that limit their access to only their classroom's data. The information intervention reduces uncertainty about parent preferences but does not change the weight teachers place on parent preferences or their own preferences.
|
|
Field
Intervention Start Date
|
Before
January 15, 2024
|
After
April 30, 2024
|
|
Field
Intervention End Date
|
Before
March 15, 2024
|
After
March 01, 2025
|
|
Field
Primary Outcomes (End Points)
|
Before
Student outcomes: grades, attendance, parents’ perceptions of student ability
Teacher outcomes: beliefs about parent preferences, own preferences
Parent outcomes: satisfaction with schooling, life satisfaction
|
After
Teacher belief updating:
Correlation between teacher beliefs about parent preferences and actual parent preferences
Student academic outcomes:
School-provided grades in academic subjects
Student attendance rates
Student non-cognitive outcomes:
Parent ratings (0-100) of child's abilities across nine dimensions of human capital
Teacher assessments of student capabilities in social and emotional domains (report card)
Teaching practices:
Self-reported teacher implementation of suggested pedagogical strategies
|
|
Field
Primary Outcomes (Explanation)
|
Before
Grades and attendance are provided by schools. All other outcomes are collected in the parent or teacher surveys.
|
After
Teacher belief updating is measured by the correlation between teachers' perceptions of parent preferences and actual parent preferences. I examine how this correlation changes after the intervention for treatment versus control teachers. Higher correlation indicates more accurate teacher understanding of what parents want for their children.
For student outcomes, I use:
Academic performance comes directly from school-provided grades and attendance records
Non-cognitive skills are measured through parent ratings on a 0-100 scale for nine dimensions of human capital grouped into academic, social, and emotional categories
Parent and teacher outcomes are captured through:
Current ability ratings: Parents rate each student's current abilities across nine skill dimensions on a 0-100 scale
Importance rankings: Both groups rank which skills they consider most important for the student to improve (a subset for teachers)
For teaching practices, I collect self-reported data on:
Implementation of pedagogical strategies (for the treatment group)
|
|
Field
Experimental Design (Public)
|
Before
This study will be done in conjunction with Delhi Public School Newtown (DPSN) and Maharashtra Institute of Technology Vishwashanti Gurukul Schools (MITVGS). Schools will administer surveys for teachers and parents of students from upper kindergarten (UKG) to 9th grade. Within each grade, students are grouped into sections, each of which is associated with a class teacher. Each section is approximately 40 students.
Within DPSN, there are an average of 10 sections in each grade, and each section contains approximately 40 students. Therefore, the total possible sample size for DPSN is 4000 students (40 students x 10 grades x 10 sections). MITVGS has 7 campuses, each of which contains an average 2 sections in each grade, and each section contains approximately of 30 students. Therefore, the total possible sample size for MITVGS is 4200 students (30 students x 10 grades x 2 sections x 7 schools).
The experimental design involves two waves of surveys for parents and teachers. The waves are separated by 2.5 months. Baseline surveys are scheduled to be take place during the week of January 15th, 2024 after students return from winter break. Endline surveys will take place 2.5 months later the week of March 18th, 2024, coinciding with the end of the Indian academic school year. The information treatment will be administered immediately after the baseline surveys conclude.
|
After
This study will be done in conjunction with four schools in four states across India. Schools will administer surveys for teachers and parents of students from 1st to 10th grade. Within each grade, students are grouped into sections, each of which is associated with a class teacher. Each section has approximately 30-40 students. Students will also be surveyed at endline.
Within the largest school in the sample, there are an average of 10 sections in each grade (100 total sections), and each section contains approximately 40 students (~4000 students total). In the smallest school, there are 3-4 sections in each grade, totaling ~1100 students. Across the schools, there are ~8150 student-parent pairs.
Having completed the baseline surveys, I have collected responses from ~3,300 student-parent pairs, representing 43.5% of the full student body.
The experiment randomizes at the classroom (teacher) level following a two-stage design:
1. Random split of grades with either odd or even grades getting information on pedagogical strategies
2. Within each group:
• Grades that get strategy info: 80% receive parent preferences and ratings information
• Grades that do not get strategy info: 20% receive parent preferences and ratings information
This design minimizes the threat of spillovers by assuming teachers within a grade communicate, thus assigning strategy information at the grade level, while parent information is specific to each class. The within-grade randomization proportions were chosen to maximize the power for comparing control and the combined treatment with both strategy and parent information.
Data collection occurs across multiple waves. Baseline surveys took place in April 2024 for the first school, and in September 2024 for the three other schools. The information treatment began in the week after the baseline surveys were completed for teachers. Endline surveys will take place in April 2025 for all schools, with a follow-up wave planned for August 2025. The timing aligns with the Indian academic calendar, with academic years starting in April.
|
|
Field
Planned Number of Observations
|
Before
5000 students
|
After
3300 students
|
|
Field
Sample size (or number of clusters) by treatment arms
|
Before
120 classrooms control, 120 classrooms treatment
|
After
96 classrooms control, 24 classrooms receive strategy information, 24 classrooms receive parent preferences and ratings, 96 classrooms receive both
|
|
Field
Intervention (Hidden)
|
Before
Within each grade-school pair (e.g., within 1st grade sections at a given school), class teachers will be randomly selected to be in the treatment or control, with equal probability. The intervention is an information treatment. Immediately after the baseline survey, I will provide treatment teachers information about their student capabilities and QOL collected from the parent survey. I have developed a web-portal through which the information will be accessed by teachers. I will provide teachers individual log-in information, through which they will have access to reports for students in their classroom, and no others.
Teachers in the treatment group will have access to four types of information about students in their classroom:
• A table of the most important skill for each student
• Student-specific levels and rankings
• Classroom-averaged levels and rankings
• A graph of levels against importance for a user-selected set of abilities
The first table contains each parent’s most important skill, i.e., the skill with the highest estimated marginal utility based on that parent’s tradeoffs. The idea underlying this table is to provide teachers a sense of what each parent values most. The second and third types of information display the full ranking and levels of skills for each student, and averaged across the classroom, respectively. The final type of information presents the information in graphical form. Teachers are able to select which capability(ies) to view, and the tool plots each student’s level against the importance ranking. The motivation for this view is to make it easier for teachers to visually group similar students. Therefore, in theory, teachers can better target instruction, and identify needs for improvement by grouping students along two dimensions: by current skill-level, and by how valuable that capability is.
|
After
The intervention is an information treatment delivered through a custom website (wiseteaching.net). Treated teachers receive:
Individual student reports showing parent preferences and current ability ratings (0-100) across nine dimensions of human capital grouped into academic, social, and emotional categories
Classroom-level summaries of parent priorities
Pedagogical strategies for improving non-cognitive skills based on feedback from teachers
Teachers access this information through individual login credentials that limit their access to only their classroom's data. The information intervention reduces uncertainty about parent preferences but does not change the weight teachers place on parent preferences or their own preferences.
|