The effect of receiving free educational support on the propensity to volunteer among secondary school students: An RCT in rural Bangladesh.

The ongoing RCT (AEARCTR - 0010647) primarily evaluates the effects of a free Interactive Voice Response (IVR) educational intervention on students' educational achievements and aspirations. This IVR education experiment was implemented between January 2023 and June 2023. In contrast, our study intends to examine whether the IVR educational intervention affects children’s inclination to volunteer. We do so by conducting an additional post-intervention survey in September 2023. In this follow-up survey, we: (1) measure students' propensity to volunteer using the “Bales Volunteerism-Activism Scale” questionnaire (Bales, 1996); and (2) elicit students’ preferences for volunteering through a “sign-up sheet” for actual volunteer activities. In October 2023, we will randomly assign invitation to volunteering activities to half of
the students who have signed up. Finally, a post-volunteering survey will be conducted in November 2023.

2023-01-01

2023-12-15

Outcome Variables

1. Primary outcomes from the post-IVR survey

1.1 Propensity to volunteer
Propensity to volunteer will be assessed using two instruments: “Bales Volunteerism-Activism Scale” and a “sign-up sheet”.
1.1.1 Bales Volunteerism-Activism Scale
1.1.2 Sign-up sheet

1.2 Preference over volunteer tasks

2. Primary outcomes from the post-volunteer survey

2.1 Attendance and participation in actual volunteer activities

2.2 Students’ well-being

2.3 Self-perceived happiness in comparison to others

1. Primary outcomes from the post-IVR survey
1.1 Propensity to volunteer
Propensity to volunteer will be assessed using two instruments: “Bales Volunteerism-Activism Scale” and a “sign-up sheet”.
1.1.1 Bales Volunteerism-Activism Scale
We will use “Bales Volunteerism-Activism Scale” (Table A1.1) to measure propensity to volunteer. Research has demonstrated that the likelihood of engaging in voluntary work is positively correlated with a higher score on the Activism scale (Bales,1996). There are 20 statements in the scale. Students will indicate to what extent they agree to each statement using the 5-point scale provided. The scale ranges from 1 ‘Strongly disagree’ to 5 ‘Strongly agree’. We will average them to obtain the score for each respondent.
1.1.2 Sign-up sheet
We will provide two types of volunteer tasks in a “sign-up sheet” (Table A1.3) to measure students’ preferences for volunteering. The two tasks are: (1) engaging in cleaning activities (cleaning); and (2) assisting primary school student with homework (tutoring). Additionally, students who don’t wish to participate in any volunteer activity could choose to sign up for neither task. A binary variable “volunteer” will be assigned a value of 1 if the student signs up for either activity, or 0 otherwise.
1.2 Preference over volunteer tasks
To elicit the students’ preferences over the two types of volunteer tasks, we will request students to indicate their preferred volunteer task (cleaning or tutoring) on the sign-up sheet. Two dummy variables will be created separately for whether cleaning is more preferred and whether tutoring is more preferred.
2. Primary outcomes from the post-volunteer survey
2.1 Attendance and participation in actual volunteer activities
We will record the attendance of the students who signed up and were randomly invited to perform volunteering activities. We will create two participation variables. The first is a binary outcome (1 for participation, 0 for non-participation) and the second is a continuous variable on a scale of 1 to 4, reflecting the hours of their participation.
2.2 Students’ well-being
We will also measure student well-being in the post-volunteer survey. We will employ the validated Adolescent Well-being Scale (18-item) developed by Birleson in 1980. The response scale ranges from 0 (most of the time) to 2 (never). This scale has been extensively utilized in research to assess various sides of adolescent well-being and offers a comprehensive framework for evaluating the effects of the intervention on students’ overall sense of well-being.
2.3 Self-perceived happiness in comparison to others
To assess participants’ subjective perception of their own happiness relative to others, we will employ the Subjective Happiness Scale (SHS) (4-item), as developed by Lyubomirsky and Lepper in 1999. This scale is a well-established instrument in the field of psychology and social sciences, designed to gauge individuals’ self-reported happiness levels and how they perceive their own happiness in relation to the general population.

3. Secondary outcomes from the post-IVR survey

3.1 Helping with household chores.

3.2 Dimensions of volunteer attitude
3.2.1 A sense of effectiveness
3.2.2 Sociability and generability
3.2.3 Idealism or philosophical commitment
3.2.4 The ‘feel good’ factor

4. Secondary outcomes from the post volunteer intervention survey

4.1 Motivation to volunteer
4.2.1 Value
4.2.2 Understanding
4.2.3 Social
4.2.4 Career
4.2.5 Protective
4.2.6 Enhancement

4.3 Life satisfaction and perceived health condition

4.3.1 Perceived health
4.3.2 Life Satisfaction

3.3 Secondary outcomes from the post-IVR survey
3.3.1 Helping with household chores.
We will use data from the existing RCT endline survey regarding time spent doing household chores to see whether students in treatment groups are more helpful to their families. Student’s own assessment of minutes spent helping parents with household chores or outside work per day (average over last week).
The sum of minutes was reported by the participant for the following household activities: cooking, cleaning up the house, bringing clean water to your house, bringing groceries from a market, helping siblings to study, farming or gardening, parent’s income-generating activities (e.g., fishing, tailoring, etc.), other income-generating activities to support their family.
The specific survey question asks: “Do you help with [activity]? If yes, how much time in a day (average over last week)? [Answer: in minutes]”
3.3.2 Dimensions of volunteer attitude
We will use the four dimensions from “Bales Volunteerism Activism scale” to analyze the underlying features that create active volunteer attitude and see whether treatment group students score higher in any of the dimensions. According to the scale there are four clear dimensions which make up the active volunteer attitude. We will average scores of each dimension to obtain the score for each respondent. The scale has the following dimensions:
3.3.2.1 A sense of effectiveness
The perception of personal efficacy, control, and significance in tackling social issues is reflected in specific questions (b, c, f, m, p). Example: Potential volunteers will have very strong negative reaction to the statement, “Anything I do can’t really change the world’s problems.” This suggests that these respondents place great importance on their ability to make a profound impact in addressing societal challenges.
3.3.2.2 Sociability and generability
The perception that volunteer priorities global issues alongside their personal commitments. This is reflected in specific questions (d, h, j, o, r, t). For instance, potential volunteers will strongly disagree with the statement, “People with ordinary lives don’t feel the need to get heavily involved in social causes.”
3.3.2.3 Idealism or philosophical commitment
This demonstrates the inclination to make sacrifices for the betterment of society and their perception that addressing significant challenges involves incurring real costs. It is evident in the responses to specific questions (e, g, i, k, n). For instance, potential volunteers will strongly agree with statements such as “There will be peace only when there is justice”.
3.3.2.4 The ‘feel good’ factor
This dimension suggests that the potential volunteers believe that their actions not only benefit others but also contribute to their own personal growth and fulfillment. It is reflected in the responses to specific questions (a, 1, q, s). Potential volunteers will show a highly positive reaction to statements such as “The more you put into life, the more you will get out of it” and “I know that when I’m working to help others, I’m also helping myself.”
4. Secondary outcomes from the post volunteer intervention survey
4.1 Motivation to volunteer
Motivation to volunteer will be measured using the ‘Volunteer Functions Inventory’ (VFI) (Clary et al., 1998) (Table A2.1), which has been applied widely in the volunteering literature. There are 30 questions in the scale which consists of a set of six motivational functions served by volunteerism. Students will indicate to what extent each motive influences their decision to volunteer using the 7-point scale provided. The scale ranges from 1 ‘not at all important/accurate’ to 7 ‘extremely important/accurate’. Scale scores result from averaging scores on the five items of each motivational function, so that individuals’ scores on each scale can range from 1 to 7; higher score will reflect greater importance of the motivation .
4.1.1 Value
Engaging in volunteer service can serve the purpose of providing individuals with opportunities to express their values associated with altruism and humanitarian concerns for others. The factor ’Value’ is reflected in question no. 3, 8, 16, 19 and 22.
4.1.2 Understanding
Volunteering provides individuals with opportunities for new learning experiences and the chance to apply and develop their knowledge, skills, and abilities that might otherwise remain untrained. The factor ‘Understanding’ is reflected in question no. 12, 14, 18, 25 and 30.
4.1.3 Social
Another significant function served by volunteering revolves around motivations related to relationships with others. By engaging in volunteer work, individuals can strengthen social connections, seek approval from significant others, and derive satisfaction from contributing to the well-being of others within their social networks. The factor ’Social’ is reflected in question no. 2, 4, 6, 17 and 23.
4.1.4 Career
Volunteering is related to career-related benefits that individuals can obtain through their participation in volunteer work. By engaging in volunteer activities, individuals can gain valuable experience, develop transferable skills, and enhance their professional networks, all of which can contribute to their career advancement and growth. The factor ‘Career’ is reflected in question no. 1, 10, 15, 21, 28.
4.1.5 Protective
This function revolves around safeguarding the ego from negative aspects of the self. In the context of volunteerism, these motivations may serve to remove guilt over being more fortunate than others and to address one’s personal issues. By volunteering, individuals can find peace, alleviate guilt, and address personal challenges, thus providing themselves with a sense of ego protection and well-being. The factor ’Protective’ is reflected in question no. 7, 9, 11, 20 and 24.
4.1.6 Enhancement
Lastly, a sixth function proposed for volunteering roots from the enhancing positive mood. People often engage in helping activities to maintain or enhance their positive effect. The factor ’Enhancement’ is reflected in question no. 5, 13, 26, 27 and 29.
4.2 Life satisfaction and perceived health condition
In the post volunteer survey, we will inquire with participants from the existing RCT (IVR education intervention) about their subjective assessments of both their overall health condition and life satisfaction. These assessments will be gathered enabling participants to provide nuanced responses. From these responses, we will construct two distinct variables:
4.2.1 Perceived health: This variable will incorporate participants’ self-reported perceptions of their current health status. It will be based on their responses on the five-point scale and will serve as an indicator of how individuals view their own health.
Survey question: How would you rate your health at the present time? [Answer: 1 (poor) to 5 (excellent)
4.2.2 Life Satisfaction: This variable will capture participants’ subjective evaluations of their life satisfaction levels. Again, this assessment will be derived from their ratings on the seven-point scale, providing valuable insights into participants’ overall contentment and happiness.
Survey question: How satisfied are you at present with your life, all things considered? [Answer: 1 (completely dissatisfied) to 7 (completely satisfied)]
By incorporating these variables into our analysis, we aim to gain a deeper understanding of the perceived health conditions and life satisfaction of our participants after the educational program and volunteer intervention.

We use the experimental design of an existing RCT. The existing IVR education experiment (AEARCTR - 0010647) used a two-stage cluster randomized controlled trial with three experimental arms (two treatment arms and one control arm). In this RCT, 208 secondary schools were selected from a list of schools in the sub districts where the implementing organization, Global Development Research Institute (GDRI), operates.
In the first stage, these 208 schools were randomly allocated to one of three experimental arms (without stratification): T1 (self-help group), T2 (assisted group), and the control group. Each of the treatment arms (T1 and T2) comprises 69 schools, while the control group includes 70 schools. In the second stage, within-school randomization was conducted to select a subgroup of 9th grade students within T1 and T2 schools to receive the treatment. 15 students were selected on average per school. The students assigned to the ‘Self-help’ group (T1) were provided audio lessons which could be accessed through the IVR system by calling a registered phone number anytime during the 6-month intervention period. The students assigned to the ‘Assisted’ group (T2) received biweekly phone calls from a tutor to discuss the content covered in the IVR audio lessons in addition to access to the IVR audio lessons. The students in the control group did not receive any intervention.

The experimental design contains three study arms:
a) T1: Self-help group (69 schools, 1058 students) –was provided with information on how to access IVR-based lessons and how this program could result in improved educational outcomes (such as higher marks on the SSC exams). Participants will receive regular text messages from the IVR system.
b) T2: Assisted group (69 schools, 1022 students) – received the same information as students in T1. Additional treatment will be provided in the form of biweekly phone calls from a tutor. Participants will also receive regular text messages from the IVR system.
c) T3: Control (70 schools, 1050 students) - did not receive any intervention.
The focus of our study is to use the existing RCT design to examine the inclination to volunteer among students in the treatment and control groups.

We use the experimental design of an existing RCT. The existing IVR education experiment (AEARCTR - 0010647) used a two-stage cluster randomized controlled trial with three experimental arms (two treatment arms and one control arm). In this RCT, 208 secondary schools were selected from a list of schools in the sub districts where the implementing organization, Global Development Research Institute (GDRI), operates.
In the first stage, these 208 schools were randomly allocated to one of three experimental arms (without stratification): T1 (self-help group), T2 (assisted group), and the control group. Each of the treatment arms (T1 and T2) comprises 69 schools, while the control group includes 70 schools. In the second stage, within-school randomization was conducted to select a subgroup of 9th grade students within T1 and T2 schools to receive the treatment. 15 students were selected on average per school. The students assigned to the ‘Self-help’ group (T1) were provided audio lessons which could be accessed through the IVR system by calling a registered phone number anytime during the 6-month intervention period. The students assigned to the ‘Assisted’ group (T2) received biweekly phone calls from a tutor to discuss the content covered in the IVR audio lessons in addition to access to the IVR audio lessons. The students in the control group did not receive any intervention.

The experimental design contains three study arms:
a) T1: Self-help group (69 schools, 1058 students) –was provided with information on how to access IVR-based lessons and how this program could result in improved educational outcomes (such as higher marks on the SSC exams). Participants will receive regular text messages from the IVR system.
b) T2: Assisted group (69 schools, 1022 students) – received the same information as students in T1. Additional treatment will be provided in the form of biweekly phone calls from a tutor. Participants will also receive regular text messages from the IVR system.
c) T3: Control (70 schools, 1050 students) - did not receive any intervention.
The focus of our study is to use the existing RCT design to examine the inclination to volunteer among students in the treatment and control groups.

Randomization done in office by a computer using Stata/BE 17.0

A two-stage randomization procedure was employed. In particular, randomly
selected 208 schools were assigned to two treatment groups and one control group, and then within each
school 15 students on average were randomly selected (in a way that ensures almost equal numbers of boys
and girls) from each of the 208 schools (for treatment in the treatment schools, and for data
collection in the control schools).

Yes

208 Schools

3,130 pupils

T1: Self-help group: 69 schools (units of randomization), 1058 students (units of observation)
T2: Assisted group: 69 schools (units of randomization), 1022 students (units of observation)
Control: 70 schools (units of randomization), 1050 students (units of observation)

We use power calculations by the existing RCT(AEARCTR - 0010647). It indicates that, if we choose a 0.24 SD effect size, an α level of 0.05, a 0.2 intra-cluster correlation (ICC), and we have a cluster size of 15 students per school, we will require a minimum sample of 3,105 students from the 208 schools evenly distributed across the three treatment arms to achieve 80% statistical power.