---
title: "Home learning environment and primary school dropout in Bangladesh: MICS analysis"
id: "plos-one-14-learning-environment-at-home-and-primary-school-dropout-status-in-bangladesh-an"
canonical_url: "https://medichelpline.com/clinical-feed/plos-one-14-learning-environment-at-home-and-primary-school-dropout-status-in-bangladesh-an"
content_type: "clinical_feed_article"
specialty: "Pediatrics"
source_name: "PLOS ONE (Medicine)"
source_url: "https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0360931"
published_at: "2026-10-21T14:00:00.000Z"
evidence_level: "Journal Feed"
license: "CC-BY-NC-4.0 / Informational Use"
---
# Home learning environment and primary school dropout in Bangladesh: MICS analysis
## Provenance & Clinical Metadata
- **Canonical URL:** https://medichelpline.com/clinical-feed/plos-one-14-learning-environment-at-home-and-primary-school-dropout-status-in-bangladesh-an
- **Specialty:** [Pediatrics](https://medichelpline.com/clinical-feed/pediatrics.md)
- **Primary Source:** PLOS ONE (Medicine)
- **Source URL:** [Original Journal Publication](https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0360931)
- **Published At:** 2026-10-21T14:00:00.000Z
- **Evidence Rating:** Journal Feed
## Executive GIST (TL;DR)
- The investigators analyzed UNICEF Multiple Indicator Cluster Survey (MICS) data to examine associations between the **home learning environment** and **primary school dropout** in Bangladesh. The abstract reports a total sample of 14,278 children aged 6–12 years; the Methods section later lists 14,305 — the source text contains both figures. - Overall dropout prevalence reported in the abstract was 1.68% among the sampled children. That figure emerged from cross-sectional survey data and bivariate and multivariable log-binomial modeling described by the authors. - The adjusted risk ratio (ARR) for males compared with females was 2.44 (95% CI: 1.63–3.64), indicating higher dropout risk among boys in this dataset. - Children aged 9–12 years at the start of the school year had higher dropout risk (ARR = 1.83, 95% CI: 1.15–2.93) compared with younger children, according to the adjusted models. - Regional variation appeared substantial; the division of Mymensingh showed an ARR of 3.82 (95% CI: 1.59–9.18) relative to Barishal, which the authors used as reference. - Home reading activity was strongly associated with retention. Children who neither read books nor were read to at home had an ARR of 14.47 (95% CI: 9.30–22.51) for dropout compared with those with home reading exposure. - The investigators applied bivariate analyses and crude and adjusted log-binomial regression models to estimate associations between dropout and covariates; the source reported these methods but did not provide full modeling specifications in the abridged content. - The authors recommended addressing gender inequality, regional disparities, weak home literacy practices, supporting older children within grades, and strengthening early childhood education and targeted regional interventions to reduce primary school dropout. - The source specified data availability (UNICEF MICS), declared no specific funding, and reported no competing interests. The authors noted strengths and limitations sections but the abridged source did not list all specific methodological limitations in full detail.
## Clinical Analysis & Structured Key Points
Learning environment at home and primary school dropout status in Bangladesh: An analysis of multiple indicator cluster survey | PLOS One Browse Subject Areas ? Click through the PLOS taxonomy to find articles in your field. For more information about PLOS Subject Areas, click here . Article Authors Metrics Comments Media Coverage Reader Comments Figures Figures Abstract Primary school dropout is a serious concern worldwide. In Bangladesh, primary education is strongly shaped by the home learning environment. This study aimed to investigate the associations between the dropout of primary school children and the learning environment at home as well as the associated risk factors by analyzing cross-sectional data from the 2019 Multiple Indicator Cluster Survey (MICS), with a total sample size of 14,278 children aged 6–12 years. Bivariate analysis, crude and adjusted log-binomial models were used to assess the relationships between dropout and risk factors. Overall, 1.68% of the children dropped out, with increased rates observed among males (ARR = 2.44, 95% CI: 1.63–3.64) and children aged 9–12 years at the beginning of the school year had a higher risk of dropout (ARR = 1.83, 95% CI: 1.15–2.93). Mymensingh presented the highest dropout rate, 3.82 (ARR = 3.82, 95% CI: 1.59–9.18) times higher than Barishal. Children who did not read books or were not read to at home had 14.47 (ARR = 14.47, 95% CI: 9.30–22.51) times higher risk of primary school dropout. To reduce primary school dropouts in Bangladesh, addressing gender inequalities, regional disparities, and inadequate home reading activity is essential. Strengthening early childhood education initiatives, promoting home-based literacy activities, supporting older children in their grades, and implementing targeted interventions in disadvantaged regions may help reduce primary school dropout in Bangladesh. Citation: Prodhan ME, Alam A, Subha S, Islam F, Al Amin M, Hasan MN, et al. (2026) Learning environment at home and primary school dropout status in Bangladesh: An analysis of multiple indicator cluster survey. PLoS One 21(10): e0360931. https://doi.org/10.1371/journal.pone.0360931 Editor: Jan Vrba, Akademia Jagiellońska w Toruniu: Akademia Jagiellonska, CZECHIA Received: December 10, 2025; Accepted: October 2, 2026; Published: October 21, 2026 Copyright: © 2026 Prodhan et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability: The datasets analyzed during the current study are publicly available from the UNICEF Multiple Indicator Cluster Surveys (MICS) repository at: https://mics.unicef.org/surveys . Funding: The author(s) received no specific funding for this work. Competing interests: The authors have declared that no competing interests exist. Introduction Education is key to economic growth, employment, poverty reduction, economic independence, and self-awareness of happiness [ 1 – 3 ]. Primary education has a substantial positive impact on economic development, allowing citizens to live stable lives and supporting sustainable living [ 4 , 5 ]. The learning environment is the basis of a child’s learning process and decisions regarding their future workplaces. It is considered a crucial element in determining the effectiveness of an efficient curriculum and the academic achievements of students [ 6 ]. Therefore, it is necessary to understand why people decide to drop out of school in developing countries, particularly in low- and middle-income countries [ 7 , 8 ] . A dropout is a student who has left school before completing the prescribed curriculum and has not enrolled in the same or different schools during the following academic session [ 9 ]. Dropping out of school is a major issue that has an impact on individuals, the community, the educational system, and society at large. The primary school dropout for millions of children continues to be a major concern in low- and middle-income nations. Sub-Saharan Africa has the highest dropout rate (42%), followed by South Asia and West Asia (33%) [ 10 ]. The evidence that there are an estimated 20.6 million lower secondary dropout children and 11.3 million elementary school-age dropout children in South Asia is supported by data gathered by the United Nations Development Programme (UNDP) [ 11 ]. An investigation revealed that households with better learning environments are associated with higher child academic outcomes. Children who have supportive parents perform better academically and behaviorally [ 12 ]. Economic status and interrelated social, school, and cultural factors are the root causes of dropout [ 13 ]. The income and educational levels of parents have detrimental effects on dropouts [ 14 ]. Factors such as gender, religion, ethnicity, housework, wage labor, disability, low self-esteem among children, poverty, and profanity in education among children are recognized as major causes of school dropout, with additional correlates such as late entry, grade repetition, mediocre educational objectives, and a 9% increase in the likelihood of dropout linked to the loss of the family head’s job [ 15 – 17 ]. There is a significant relationship among age, race, corporal punishment, and school dropouts, which shows that dropping out of school contributes not only to a cycle of violent victimization but also to educational disparities, long-term unemployment, poverty, poor health outcomes, continued reliance on public assistance, political and social apathy, and juvenile crime [ 18 – 23 ]. UNESCO has reported that continuing school dropout and educational gaps impose substantial economic costs globally, with losses projected to reach US$10 trillion annually by 2030, disproportionately affecting low- and middle-income countries [ 24 ]. In Bangladesh, the primary school dropout rate was 14.15% in 2021 [ 25 ]. Additionally, evidence suggests that the proportion of rich children who complete school is 22 percent greater than that of poor children [ 26 ]. Households in the richer and richest wealth groups have lower school dropouts, whereas children living in a larger family or having siblings have higher school dropouts [ 27 ]. The risk that a daughter will drop out of school increases as fathers participate more in home decision-making [ 28 ]. Age, gender, marriage, and financial constraints are the top predictors of dropout in Bangladesh. Dropout continues to be strongly correlated with parental disengagement from their children’s academics and absences from school [ 29 – 31 ]. With respect to low-income families, the children (41%) had to take care of their homes, which included helping parents with household tasks (cooking or agricultural work) or raising a sibling or younger child [ 32 ]. However, a further study explored the factors contributing to school dropout in Bangladesh, including persistent poverty, parents’ unwillingness, financial obstacles, biased social practices, poor school infrastructure, lack of quality education, geographic isolation, unequal access to education, and security issues for girls [ 33 ]. Although several studies have been conducted on dropout, no study has examined the association between primary school dropout and the home learning environment in Bangladesh. This study aims to investigate the association between home learning environment and primary school dropout in Bangladesh. This study uniquely examines household-level stimulation conditions, assessing whether children growing up in environments that encourage reading are more likely to remain enrolled in school. The findings can provide policymakers and development stakeholders who wish to reduce dropout rates in low- and middle-income countries. Household-targeted interventions, such as initiatives for educating parents, low-cost learning materials, and community-led tutoring support, may help children from poor families overcome barriers to learning. National education monitoring systems may enhance their effectiveness by incorporating home learning environment indicators, especially in relation to Sustainable Development Goal 4, which aims for inclusive and equitable quality education. Methods Data source We used the latest dataset of the Multiple Indicator Cluster Survey (MICS) conducted in 2019 by UNICEF, a large-scale multidimensional household survey that is nationally representative. Standardized questionnaires are used to address child labor, violent discipline, parental involvement, education, child development, and socioeconomic and household characteristics [ 34 ]. Sampling design and sample size The MICS survey applies a two-stage cluster sampling technique to randomly select children between the ages of 5 and 17. The 2019 MICS is based on a sample of 40,617 mothers/caretakers who were surveyed, with 97.0% of them responding. On the basis of the expected sizes of the enumeration regions from the most recent population census, primary sampling units (PSUs) for the census were developed and chosen from each sample domain using systematic probability proportional to size (PPS) sampling processes. Among the eight administrative divisions (Dhaka, Mymensingh, Chittagong, Sylhet, Rajshahi, Rangpur, Barishal, and Khulna) of Bangladesh, MICS offers a thorough picture of children’s education. The main sampling strata were urban and rural areas within each district, and the sample of households was drawn in two stages [ 34 ]. A total of 31,196 children admitted to school last year were included in the study. Of these, 15,639 were enrolled in primary school, while those not enrolled were excluded. A total of 14,305 children aged 6–12 years were included, and the remaining children who were not aged 6–12 years were excluded [ 35 ]. The final sample of 14,278 children who started school at age 6 or older was included, whereas those who started school before 6 years were excluded because the official primary school entrance age in Bangladesh is 6 years [ 25 ]. Fig 1 shows a schematic diagram of the study sample. Download: PNG larger image TIFF original image Fig 1. Schematic diagram of the study sample. https://doi.org/10.1371/journal.pone.0360931.g001 Outcome variable To create our outcome variable (dropout status), we assigned a value of 1 to each child on the basis of not attending school or an early childhood program during the current school year but was admitted last year, indicating a “yes” response; otherwise, 0 indicated “no”. Among the observed sample, the weighted prevalence of dropout was 1.68% (95% CI: 1.44%−1.95%, unweighted n = 236), while 98.32% remained in school (95% CI: 98.05%−98.56%, unweighted n = 14,042) ( Fig 1 ). Risk factors To determine the possible risk factors affecting dropout status, a set of factors was selected: child’s age (6–8 years and 9–12 years), sex, child’s age at the beginning of the school year (6–8 years and 9–12 years) [ 36 ], children’s functional difficulties, involvement in economic activities, needed to be physically punished to be brought up properly, severe physical punishment, mother’s education (preprimary or none, primary, secondary and higher secondary or above), mother’s functional difficulties and whether she had another child under 5, father’s education (preprimary or none, primary, secondary and higher secondary or above), area (urban or rural), division, gender of the household head, ethnicity of the household head (Bengali or others), religion (Islam, Hinduism, Christianity, Buddhism), household size (up to 4 or more than 4) [ 37 ], household wealth index [ 38 ], household had electricity, household had internet, types of toilet facility (improved or unimproved), salt iodization, sources of water (improved or unimproved), child had 3 or more books to read at home, child read books or was read to at home, and child spoke the same language at home and school. Age at the beginning of the school year was obtained from the MICS variable schage and categorized into 6–8 years and 9–12 years. Child’s involvement in economic activities was categorized as “yes” if he/she worked or helped in the garden, helped in the family business, produced or sold articles, and engaged in other activities for income; otherwise “no”. A child faced severe physical discipline if he/she was hit or slapped on the face, head, or ears or beaten up as hard as possible; otherwise, “no”. The types of toilet facilities were categorized into improved (flush toilet, flush to piped sewer system, flush to septic tank, flush to pit latrine, pit latrine with slab and ventilated improved pit latrine) and unimproved (flash to open drain, pit latrine without slab, hanging toilet, bush/field) [ 39 ]. Salt iodization was categorized as “yes” if the iodine level was between 0 and 15 ppm or above and “no” if the iodine level was 0 ppm or no salt in the household [ 40 ]. Sources of water were categorized into improved (piped water, tubewell, rainwater, water kiosk) and unimproved (unprotected well, spring, river/pond) [ 41 ]. Children read books or were read to at home was measured using FL6A (whether the child reads books at home), and FL6B (whether someone reads books to the child at home). These two variables were combined and categorized as “Yes” when either activity was reported and “No” when neither activity was reported. Statistical analysis We used the svyset (declare data as survey data) command in Stata to account for the complex survey design [ 42 ]. The primary sampling unit was cluster number ( HH1 ), the sampling weight was children 5–17’s sample weight ( fsweight ), and stratum was specified as the stratification variable. The survey design was incorporated into the estimation of percentages, design-based tests of association, and regression analyses, including the calculation of p-values and 95% confidence intervals. The eligible study population was defined as a survey subpopulation using the subpop option so that the survey design information was retained while restricting estimation to the target population. Missing observations were retained in the dataset during descriptive and bivariate analyses. For the multivariable analysis, a complete-case approach was used. Bivariate analysis using two-way cross-tabulation with Design-based F tests (Rao-Scott adjusted chi-square test) was performed to evaluate the associations between dropout and each of the 26 variables. Seventeen variables with P-value ≤ 0.20 were retained for subsequent crude regression analyses, following a staged variable selection approach used in previous epidemiological studies [ 43 ]. We used logistic, robust Poisson, and log-binomial regression models to assess and evaluate the associations among dropout, learning environment at home, and associated risk factors. Logistic regression model is widely used for binary outcomes and provides odds ratios (ORs) that are useful for assessing risk factors [ 44 ]. The log-binomial regression model directly estimates prevalence or risk ratios (RRs), making it more suitable for cross-sectional studies since the risk ratio is simpler for nonspecialists to comprehend and explain than the odds ratio is [ 45 ]. On the other hand, robust Poisson regression model is an alternative approach that uses robust error variance to provide risk ratio estimates while addressing issues of overestimated errors in binary outcomes [ 46 ]. Crude models were employed for each of the mentioned models, and we included the risk factors in the adjusted models on the basis of an arbitrary p-value threshold of ≤ 0.20. A liberal screening threshold was used to avoid prematurely excluding potentially important predictors or confounding variables that may become significant after adjustment, consistent with established regression model-building recommendations [ 47 – 49 ]. Additionally, with a cutoff value of 5.00, we used the variance inflation factor (VIF) value to analyze multicollinearity in the final model [ 50 ]. All variables, except for the child’s age and age at the beginning of the school year, had VIF values less than 5.00. Therefore, the child’s age was dropped from the adjusted model to mitigate multicollinearity issues. Each of the adjusted models was then employed. All models converged successfully. The log-binomial and robust Poisson models estimated adjusted risk ratios (ARRs) with 95% confidence intervals, whereas logistic regression estimated adjusted odds ratios (AORs) with 95% confidence intervals. Following the complete-case approach, the final analytic sample for the multivariable analysis comprised 10,813 children. Two further sensitivity analyses were performed to test the robustness of the main findings. First, because missing observations were present for the key exposure variable “Child read books or are read to at home”, missing values were alternatively classified as “No” and “Yes,” and the survey-weighted log-binomial regression model was re-estimated under each assumption. Second, a restricted survey-weighted log-binomial regression model was fitted using a theory-based set of covariates comprising child’s sex, age at the beginning of the school year, functional difficulties, involvement in economic activities, severe physical discipline, maternal education, paternal education, household wealth, and home reading activity. The results of these sensitivity analyses were compared with those of the primary complete-case model. We used evaluation metrics such as log-likelihood, Akaike information criterion (AIC), and Bayesian information criterion (BIC) values to compare all the models to find the best model. The higher value of log-likelihood as well as the lowest values of AIC and BIC indicate a better fit of the data after accounting for model complexity (i.e., the number of model parameters) [ 51 ]. Ethics statement This study relied on secondary analysis of publicly available data from Bangladesh Multiple Indicator Cluster Survey (MICS). The Bangladesh Bureau of Statistics (BBS) and UNICEF have acquired ethical clearance for the original survey following national and international ethical standards. Informed consent was obtained from all participants or parents/guardians prior to data collection. The MICS dataset was de-identified and was made available to the researcher without further ethical approval. Results From bivariate analysis using two-way cross-tabulation with Design-based F tests (Rao-Scott adjusted chi-square test), seventeen variables were included in the crude model at a p-value ≤ 0.20 threshold (child’s age, sex, age at the beginning of the school year, child’s functional difficulties, child faces severe physical discipline, involvement in economic activities, mother’s education, father’s education, division, sex of household head, household wealth index, household has internet, type of toilet facility, salt iodization, sources of water, child has 3 or more books at home, child reads books or are read to at home), as shown in Table 1 . Download: PNG larger image TIFF original image Table 1. Sample characteristics of children by dropout status. https://doi.org/10.1371/journal.pone.0360931.t001 Table 1 shows that among the 14,278 children, 236 (1.68%) were primary school dropouts and 14,042 (98.32%) remained in school. Dropout proportion was more common among male children than female children (2.54% vs. 0.85%) and among children aged 9–12 years than those aged 6–8 years (1.97% vs. 1.03%). Similarly, dropout was higher among children aged 9–12 years at the beginning of the school year than among those aged 6–8 years (1.97% vs. 1.11%). Children with functional difficulties had a higher dropout proportion than those without such difficulties (3.00% vs. 1.56%). Dropout was also more common among children involved in economic activities than among those who were not (3.00% vs. 1.54%)
## Related Clinical Research

- [Movement Behaviors Linked to Social-Emotional Outcomes in Early Childhood](https://medichelpline.com/clinical-feed/plos-one-8-24-hour-movement-behaviour-composition-associated-with-desired-social-emotional.md)
- [Cross-national validation of the DASS-Y in Lebanese and Chinese adolescents](https://medichelpline.com/clinical-feed/plos-one-20-cross-national-validation-of-the-depression-anxiety-and-stress-scale-for-youth.md)
- [Longitudinal links between pain and anxiety/depression in young people: systematic review and meta](https://medichelpline.com/clinical-feed/medrxiv-10-temporal-associations-between-pain-and-anxiety-depression-in-young-people-a.md)
- [Human Bocavirus Identified in Stool During Pediatric Acute Gastroenteritis Outbreak in Salvador, B](https://medichelpline.com/clinical-feed/medrxiv-16-detection-of-human-bocavirus-in-stool-samples-collected-during-an-outbreak-of.md)
- [Spatial profiling links STING to Kawasaki disease cardiac inflammation](https://medichelpline.com/clinical-feed/biorxiv-2-spatial-profiling-of-kawasaki-disease-cardiac-inflammation-identifies-sting-as.md)

## Navigation
- [← Back to Pediatrics Feed](https://medichelpline.com/clinical-feed/pediatrics.md)
- [← All Clinical Specialties](https://medichelpline.com/clinical-feed.md)
## Medical & Regulatory Disclaimer

> [!CAUTION]
> MedicHelpline content is structured for research, educational, and professional discovery purposes. It does not constitute individual medical advice, clinical diagnosis, or treatment recommendations.
> Always verify dosing, contraindications, and regulatory alerts against official product labeling and primary regulatory sources before clinical decision-making.