|
Field
Abstract
|
Before
India has a large and comprehensive system of caste-based affirmative action in its public colleges and universities. Affirmative action in India’s higher education is meant to address historical inequities that resulted from centuries of caste-based oppression. The system has resulted in a large number of beneficiaries (students from “lower-caste” backgrounds) and non-beneficiaries (students from “upper-caste” backgrounds). Upper-caste students—who are the non-beneficiaries—might have beliefs about and preferences for affirmative action and other redistributive policies which are shaped, in part, by their misperceptions about caste. In this study, I propose an online survey experiment to capture these biases and misperceptions and conduct interventions to address them. I randomly assign 1600 college students in India to one of three arms: a 2-minute online intervention that provides students with facts about caste; a 2-minute online intervention that provides research evidence about caste-related discrimination; and a 4-minute online intervention that combines facts and research evidence about caste. The results of this study are critical to understanding caste-related biases and misperceptions among college students in India, and, more broadly, for designing interventions that promote diversity and inclusion at higher education institutions around the world.
|
After
India has a large and comprehensive system of caste-based affirmative action in its public colleges and universities. Affirmative action in India’s higher education is meant to address historical inequities that resulted from centuries of caste-based oppression. The system has resulted in a large number of beneficiaries (students from “lower-caste” backgrounds) and non-beneficiaries (students from “upper-caste” backgrounds). Upper-caste students—who are the non-beneficiaries—might have beliefs about and preferences for affirmative action and other redistributive policies which are shaped, in part, by their misperceptions about caste. In this study, I propose an online survey experiment to capture these biases and misperceptions and conduct interventions to address them. I randomly assign 1600 college students in India to one of three arms: a 10-minute online intervention that provides students with facts and research evidence about caste; 10 minute online intervention where students read a letter by an anonymous lower-caste student of the SAME gender and write a letter to a friend explaining what they learned from the letter (saying-is-believing, same gender); 10 minute online intervention where students read a letter by an anonymous lower-caste student of the OPPOSITE gender and write a letter to a friend explaining what they learned from the letter (saying-is-believing, opposite) gender; and a 10-minute online intervention that provides students information unrelated to caste. The results of this study are critical to understanding caste-related biases and misperceptions among college students in India, and, more broadly, for designing interventions that promote diversity and inclusion at higher education institutions around the world.
|
|
Field
Trial End Date
|
Before
October 31, 2023
|
After
July 01, 2025
|
|
Field
Last Published
|
Before
July 10, 2023 09:13 PM
|
After
March 23, 2025 06:57 AM
|
|
Field
Intervention End Date
|
Before
September 15, 2023
|
After
April 15, 2025
|
|
Field
Primary Outcomes (End Points)
|
Before
1. Preferences for affirmative action in college
2. Preference for supporting organizations that focus specifically on empowering lower-caste students
|
After
1. Preferences for affirmative action in college
2. Preference for supporting organizations that focus specifically on empowering lower-caste students
3. Attitudes toward caste
|
|
Field
Primary Outcomes (Explanation)
|
Before
1. Preferences for affirmative action in college: The outcome will be constructed based on respondents' answers to a question in the endline about their preference for affirmative action quotas in higher education in India (e.g., what percentage of seats should be reserved in colleges for students from lower-caste groups)
2. Preference for supporting organizations that focus specifically on empowering lower-caste students: this will be based on respondents' answers to a question in the endline about the choice of organization to which they would like to donate a part of the money they receive for filling out the survey (the question will provide a choice of two organizations: one that specifically supports education for lower-caste students and another generic education NGO. Respondents will also have a choice to donate to neither organizations).
|
After
1. Preferences for affirmative action in college: The outcome will be constructed based on respondents' answers to a question in the endline about their preference for affirmative action quotas in higher education in India (e.g., what percentage of seats should be reserved in colleges for students from lower-caste groups)
2. Preference for supporting organizations that focus specifically on empowering lower-caste students: this will be based on respondents' answers to a question in the endline about the choice of organization to which they would like to donate a part of the money they receive as part of a lottery -- they will have the option of not donating, donating to a caste-based organization, or donating to a gender-based organization.
3. Attitudes index: an index that uses responses from four attitudinal questions (attitudes toward lower-caste students) from the survey—this be z-scored. The questions will be asked on a standard 5-point Likert scale.
|
|
Field
Experimental Design (Public)
|
Before
The intervention will be in the form of an online survey experiment. The online survey experiment will take place as follows. Students will be randomized students into four equal groups (randomization is done automatically via Qualtrics): control/placebo (n = 400), fact treatment arm (n = 400), research arm (n = 400), and fact plus research evidence treatment arm (n = 400). The fact treatment arm (T1) will consist of providing students with information/facts about social equality especially related to caste. The second treatment (T2) arm will provide a short video about research evidence pertaining to the misperceptions about lower-caste college students and evidence about caste-based discrimination. The third treatment arm (T3) will combine the fact treatment with the research evidence intervention (fact + research evidence).
|
After
The intervention will be in the form of an online survey experiment. The online survey experiment will take place as follows. Students will be randomized students into four equal groups (randomization is done automatically via Qualtrics): control/placebo (n = 400), fact treatment arm (n = 400), letter opposite gender arm (n = 400), and letter same gender arm (n = 400). The fact treatment arm (T1) will consist of providing students with information/facts and research about social equality especially related to caste. The second treatment (T2) arm will provide students a letter from an anonymous lower-caste students of the same gender and then prompt students to write a letter to a friend describing what they learn. The third treatment arm (T3) will also be a letter but from a lower-caste student of the opposite gender. T4 will be the control/placebo arm with information unrelated to caste or gender.
|
|
Field
Intervention (Hidden)
|
Before
The intervention will be in the form of an online survey experiment. The online survey experiment will take place as follows. Students will be randomized students into four equal groups (randomization is done automatically via Qualtrics): control/placebo (n = 400), fact treatment arm (n = 400), research arm (n = 400), and fact plus research evidence treatment arm (n = 400). The fact treatment arm (T1) will consist of providing students with information/facts about social equality especially related to caste. The second treatment (T2) arm will provide a short video about research evidence pertaining to the misperceptions about lower-caste college students and evidence about caste-based discrimination. The third treatment arm (T3) will combine the fact treatment with the research evidence intervention (fact + research evidence).
Specifically, the fact treatment arm (T1) will contain facts about the proportion of socially disadvantaged groups in India’s population versus advantaged groups (socially disadvantaged groups constitute 68% of India’s population versus 32% of socially advantaged groups). Students will also see facts about the gaps in income, education, and college admission criteria between upper and lower-caste groups. These facts will be shown to participants in a visually appealing manner (e.g., through visually appealing graphs and plots).
The research evidence intervention (T2) will describe research evidence pertaining to misperceptions about lower-caste college students and evidence about caste-based discrimination. For example, the research evidence video will comprise of evidence from correspondence studies to show caste-based discrimination manifests itself in the labor market. Finally, in the facts plus research intervention (T3), participants will get both the facts intervention and the research evidence intervention. The placebo (control arm) will be comprised of information and a short video unrelated to caste or caste-based discrimination. After the intervention, I conduct two endline surveys: one immediately after the intervention and another two weeks after the intervention. In both the endline surveys, I will capture students’ endline attitudes toward caste and preference for redistributive policies such as affirmative action in higher education in India.
|
After
The intervention will be in the form of an online survey experiment. The online survey experiment will take place as follows. Students will be randomized students into four equal groups (randomization is done automatically via Qualtrics): control/placebo (n = 400), fact
= research arm (n = 400), letter opposite gender arm (n = 400), and letter same gender arm (n = 400). The first treatment arm (T1) will combine the facts about caste and research evidence about caste-based discrimination. These facts will be shown to participants in a visually appealing manner (e.g., through visually appealing graphs and plots). The research evidence video will comprise of evidence from correspondence studies to show caste-based discrimination manifests itself in the labor market. In the second treatment arm (T2) respondents will read a letter from an anonymous college student of the same gender. In the letter, the anonymous college student will describe their experiences with caste-based discrimination. The respondents will then be prompted to write a letter to their friend, describign what they learned in the letter. In third treatment arm, students will also read and respond to the letter. However, they will read a letter of an anonymous student of the opposite gender. The The placebo (control arm) will be comprised of information and a short video unrelated to caste or caste-based discrimination. After the intervention, I conduct two endline surveys: one immediately after the intervention and another two weeks after the intervention. In both the endline surveys, I will capture students’ endline attitudes toward caste, preference for redistributive policies such as affirmative action in higher education in India, and willingness to donate to caste-based philanthropies.
|
|
Field
Secondary Outcomes (End Points)
|
Before
1. My first outcome variable will measure attitudes toward lower-caste students (an index that uses four attitudinal questions from the survey—this be z-scored).
2. My second outcome variable will measure preferences for affirmative (an index that uses four questions from the survey related to preferences for affirmative action in college —this be also z-scored).
|
After
1. Preferences for knowledge
2. Implicit Association Test (IAT) for caste
|
|
Field
Secondary Outcomes (Explanation)
|
Before
1. Attitudes index: an index that uses responses from four attitudinal questions (attitudes toward lower-caste students) from the survey—this be z-scored. The questions will be asked on a standard 5-point Likert scale.
1. Preferences index: an index that uses responses from four questions from the survey that ask respondents about their preferences for affirmative action in higher education—this be z-scored. The questions will be asked on a standard 5-point Likert scale.
|
After
1. Preference for knowledge: based on whether students sign up to receive information and insights about caste especially in higher education
2. I will calculate the mean of difference in IAT scores for those who received the treatment and those who did not
|