Play-Based Learning in Crisis Contexts: A Randomized Controlled Trial on Educational and Psychosocial Outcomes in Host Community Children

BRAC-IED (Institute of Educational Development) initiated a 12-week afterschool intervention program named “Creative Club Program,” which promotes physical and mental well-being through creative expressions and pro-environmental behavior. The program includes a safe space for adolescent youth where they can engage in various creative activities, such as art, writing, playful games, and group work. The intervention was given after school hours within the school premises, where the Creative Club constitutes a safe space for children where they can play with each other.

The intervention will be conducted with 150 Class Six students from 10 schools across Teknaf and Ukhiya districts in Cox’s Bazar over a period of 12 weeks. The sessions will take place after their regular school hours. Through this intervention, the clubs will promote the well-being of adolescents by helping them manage their emotions positively. In other words, the “Creative Club Program” will function as an inclusive and playful environment for adolescents, thereby helping them embrace their change and celebrate their turning point in life.

The play activities are moderated by 'Bondhu Apa' in the clubs, where they help students with playing instruments.

1. Co-creation of a permanent environment along with a created environment: In 5 schools, the club is developed within a school setting that occupies a permanent space in the schools.

2. Co-creation of a pop-up environment along with the created environment: In 5 schools, the club is developed within a school setting that occupies a temporary space in the schools.

3. 10 schools are in control; no such club is created.

The intervention will take place in those spaces. The play activities include painting, drawing, writing stories, and handcrafting using bamboo, paper, and locally available resources, etc. The key idea is to determine whether creating a Creative Club will enhance students' psychosocial outcomes, including self-esteem, well-being, resilience, knowledge, and practice. By utilizing various settings, such as permanent and temporary environments, we aim to determine whether these settings lead to different outcomes.

BRAC-IED has designed the Creative Club Program intervention and will implement it in the host community for evaluation. The intervention will investigate effective ways to integrate creative club interventions for children between the ages of 11 and 18. The intervention program follows Dr. Martin Seligman’s Positive Youth Development (PYD) theoretical framework, also known as the PERMA model. The model focuses on positive psychological elements, including positive emotion, relationships, and well-being that directly complement the intervention's expected outcomes. The hypothesis is that integrating authentic and culturally relevant creative clubs with a safe space will lead to improved emotional state, academic performance, and psychosocial outcome indicators such as emotional management, community engagement, and leadership skills.

The rationale behind selecting specific art activities and playful games is that it directly connects to the core objectives of the intervention, including socio-emotional development, creativity enhancement, and environmental awareness. For example, art activities are directly linked to creative expression and creativity development. Playful games are directly linked to enhancing cognitive and social skills and improving physical and mental health. These also create a scope of inclusivity and engagement. Furthermore, the selection of activities takes into account their availability and adaptability within a resource-constrained setting. For instance, the use of bamboo or mats for art zones and the emphasis on simple, low-cost games serve as examples. This initiative also addresses the challenges of adolescents in Bangladesh, such as limited scopes of creative and recreational expression.

To measure the results, we used five trusted tools:

1. The Rosenberg Self-Esteem Scale (RSES): To measure self-esteem outcomes;
2. The Stirling Children’s Wellbeing Scale (SCWBS): To measure well-being outcomes;
3. The KAP Questionnaire (Knowledge, Attitude, and Practice): To measure knowledge and practice outcomes;
4. The Child and Youth Resilience Measure (CYRM-28): To measure youth resilience;
5. Foundational learning tool from the Multiple Indicator Cluster Survey (MICS): To measure foundational learning outcomes.

A pilot study on 80 students from those districts through BRAC-IED validated the instruments. BIGD also validated the instruments, which translated into a questionnaire, through a second pilot study on 55 students in those districts.

2024-08-15

2025-09-30

Students' self-esteem, well-being, resilience, and knowledge, practice, and attitude (KAP)

To assess primary outcome variables, we have employed four validated instruments: the Rosenberg Self-Esteem Scale (RSES), the Stirling Children’s Well-being Scale (SCWBS), the Knowledge, Attitude, and Practice (KAP) Questionnaire, and the Child and Youth Resilience Measure (CYRM-28).

1. Self-Esteem Measurement: We use the Rosenberg Self-Esteem Scale (RSES), a self-reflective instrument consisting of 10 Likert-scale items, with scores ranging from 1 to 4. The cutoff score is determined by the mean, where scores above the mean indicate high self-esteem, and scores at or below the mean indicate low self-esteem.

2. Well-being Measurement: We use the Stirling Children's Well-being Scale (SCWBS), a self-reflective questionnaire consisting of 15 Likert-scale items, with scores ranging from 1 to 5. The cutoff score is determined by the mean, where scores above the mean indicate a high level of well-being, and scores at or below the mean indicate a low level of well-being.

3. Resilience Measurement: We use the Child and Youth Resilience Measure (CYRM-28), a self-reflective instrument consisting of 28 items, each scored from 1 to 5. Resilience is measured in two dimensions: Caregiver/Relational or Family-Based Resilience is assessed using the total scores of items 4, 5, 6, 11, 15, 17, and 18. Personal/Individual-Level Resilience is assessed using the total scores of items 1, 2, 3, 7, 9, 10, 12, 13, and 14. The resilience levels are categorized as follows: One standard deviation above the mean (Value 1) indicates high resilience. Within one standard deviation of the mean (Value 2) indicates moderate resilience. One standard deviation below the mean (Value 3) indicates low resilience.

4. Knowledge, Attitude, and Practice (KAP) Measurement: The KAP questionnaire consists of 30 items designed to assess students' attitudes, knowledge, and practices on various themes such as gender equality, environmental consciousness, social aptitude, and recreational activities. The first 12 items assess students' attitudes, with a total possible score of 24. The last 18 items assess students' knowledge and practices, with a total possible score of 40.

Foundational Learning Skills

For the Foundational Learning (FL) skills, we have used the foundational learning skills methodology of the standard Bangladesh Multiple Indicator Cluster Survey 2019 (MICS-2019) questionnaire. The FL module measures one composite indicator by the percentage of children with foundational reading and number skills. Since, the denominator of children’s ages is 7-14 years, the age range is applicable to our survey.

The numerator consists of a reading element and a numeracy element with the following specifications:
1. Reading: Students who successfully completed three foundational tasks:
a) 90% of words read correctly in the reading passage
b) Answers correctly 3 literal questions about the reading passage
c) Answers correctly 2 inferential questions about the reading passage

2. Numeracy: Students who successfully completed four foundational number
tasks:
a) 6 simple number recognition tasks
b) 5 simple number discrimination tasks
c) 5 simple addition tasks
d) 5 simple pattern recognition & completion tasks

3. Composite: Applying to the students having both foundational reading and
number skills, this indicator consists of the following specifications:
Numerator: Number of children 7-14 years who successfully complete
a) three foundational reading tasks (listed above in “Reading”)
b) four foundational number tasks (listed above in “Numeracy”)
Denominator: Total number of children age 7-14 years

The study uses a school cluster-based randomized control trial (RCT). The school intervention and research location is Cox’s Bazar, particularly in the Teknaf and Ukhiya upazilas. There are

1. 10 treatment schools (each school has 15 students selected randomly) and
2. 10 control schools (each school has 15 students selected randomly).

The total sample size is 300. The allocation of intervention to treatment and control schools has been randomized proportionately. For example, Teknaf has four treatment schools, and Ukhiya has six treatment schools and vice versa. Ukhiya has four control schools, and Teknaf has six control schools. To minimize potential imbalances, we controlled for these factors by using stratified randomization. Schools were categorized based on distance, resources, and teacher qualifications, and random assignment was performed within each stratum, which helped minimize variation and reduce selection bias.

The treatment group has two arms:

1. The randomly selected five schools will have both co-creation of a permanent environment along with a created environment,
2. The rest of the selected five schools will have both co-creation of a pop-up environment along with a created environment.

Note: The term permanent environment refers to a permanent space within the school. The pop-up environment refers to a temporary space within the school.

A proportionate randomization was done by using STATA.

School

Yes

20 schools

300 adolescent students

10 schools control (10*15=150 students), 10 schools treatment (10*15=150 students, with five schools permanent environment, and five schools pop-up environment).

We identify the minimum detectable effect size for our main outcomes as follows: 1. Rosenberg Self-Esteem Scale (RSES): Wong et al. (2012), with regard to their leadership program intervention in school, found that 6-month leadership training and services identified a 2.5 percentage point self-esteem increase, assuming a standard deviation of 4.0 in both groups. Based on similar statistics, we found similar results with a sample size of 300 students from 20 schools (10 in treatment and 10 in control). This sample size is sufficient to detect a 2.5 percentage point increase in the well-being scale score between the intervention and control groups with a power of at least 90%. This holds true for intra-cluster correlation coefficients (ICCs) ranging from 0.05 to 0.15. 2. Stirling Children’s Wellbeing Scale (SCWBS): Schoeps et al. (2019), with regard to their emotional development intervention on subjective well-being of the university students, found that 14-hour sessions during two months, identified the effect size for emotional intelligence, cognitive empathy, happiness, positive affect, and life satisfaction ranged from 0.20 to 0.44, indicating modest impact assuming a standard deviation of 5.00 in both groups. Based on similar statistics, we found similar results with a sample size of 300 students from 20 schools (10 in treatment and 10 in control). This sample size is sufficient to detect a 3.0 to 4.0 percentage point increase in the well-being scale score between the intervention and control groups with a power of at least 90%. This holds true for intra-cluster correlation coefficients (ICCs) ranging from 0.05 to 0.15. 3. Child and Youth Resilience Measure (CYRM-28): Yamamoto et al. (2019), with regard to their effects of the cognitive-behavioral intervention program on the resilience of Japanese elementary schools, found that eight program lessons, identified the effect size of 0.10 for anxiety measurement, social support, and resilience scale, assuming a standard deviation of 20.00 in both groups. Based on similar statistics, we found similar results with a sample size of 300 students from 20 schools (10 in treatment and 10 in control). This sample size is sufficient to detect a 10 percentage point increase in the well-being scale score between the intervention and control groups with a power of at least 80%. This holds true for intra-cluster correlation coefficients (ICCs) ranging from 0.05 to 0.15.