Teachers’ well-being, home learning environment and children learning outcomes in emergencies: a case of Afghanistan

Abstract

Teachers’ well-being and a conducive home learning environment play a pivotal role in improving children learning outcomes, but little is known about this in a humanitarian setting. This article assesses the role of teachers’ well-being and home learning environment on children learning outcomes in Afghanistan. A cross section endline survey was conducted among 1407 primary school children, 1407 parents and 315 teachers in seven provinces of Afghanistan. Population-based two stage-cluster sampling was used. Children taught by teacher’s who received professional development support were 3.19 and 4.0 times more likely to attain higher literacy and numeracy levels. Children taught by teachers with adequate teaching and learning materials were 1.60 and 1.83 times more likely to achieve higher literacy and numeracy levels. Children from teachers with ability to voice up and be consulted by school management attained higher social-emotional learning competencies. Children experiencing violent home learning environment had lower odds of achieving higher literacy, numeracy, and social-emotional learning competencies. Children with parents that helped them with learning were 1.85 and 1.94 times more likely to have higher literacy and numeracy levels. There is scope for public sector and development partners to promote a package of interventions aimed at promoting teacher’s professional development, provision of teaching and learning materials, teacher recognition and voice to improve children’s literacy, numeracy, social and emotional learning skills. Parents need to ensure a safe and conducive home learning environment. Parent should be actively involved in children’s learning activities to nurture strong parental-child bond and parental-school partnership.

Keywords:

• Teachers’ well-being
• home learning environment
• literacy
• numeracy
• social and emotional learning
• children
• education in emergencies
• Afghanistan

REVIEWING EDITOR:

• Xin Yun Hu, The Education University of Hong Kong, Hong Kong

SUBJECTS:

• Education
• Teachers & Teacher Education
• Childhood
• Continuing Professional Development
• Education Policy & Politics
• Educational Psychology
• Primary/Elementary Education
• School Leadership, Management & Administration

Introduction

There is a growing body of research around teacher well-being and home learning environment (HLE). The majority of earlier studies on teachers’ well-being narrowly focused on negative constructs including burn out, stress, strain, negative emotions, emotional exhaustion (Capone et al., 2022; Dreer, 2022; Hascher & Waber, 2021; McCallum, 2021; Parker et al., 2012). However, more recently, teacher well-being studies in psychology have started looking at the positive aspects of job satisfaction, positive emotions, professional development, teacher voice and recognition (Billaudeau et al., 2022; Capone et al., 2022; Dreer, 2022; García-Álvarez et al., 2023). Indeed, the teachers well-being model developed by Organization for Economic Cooperation and Development (Billaudeau et al., 2022) demonstrate that well-being is driven by teachers’ working conditions including working hours, earnings, professional development (García-Álvarez et al., 2023; Hohensee & Weber, 2022), quality of relationship with school staff, work autonomy and learning environment (Billaudeau et al., 2022). Another strand of literature considers social-emotional activities as a framework for supporting teacher’s well-being which include equal treatment, school culture, relationship building, professional learning and teacher voice (Cigala et al., 2019; Hascher & Waber, 2021; Oliveira et al., 2021). Teacher well-being is defined as teachers’ feeling and responses to the health, cognitive, emotional, and social conditions concerning their job (Dreer, 2022; Hascher & Waber, 2021).

The increased focus on teacher well-being is driven by its important role in teaching and children learning (Carroll et al., 2021; P.-H. Li et al., 2022; McCallum, 2021; Panadero et al., 2022; Yeh & Barrington, 2023). There is an emerging body of literature that demonstrates that teacher well-being is an important topic in the education sector and society as it is associated with staff commitment, student outcomes, effective teaching and governance (Hascher & Waber, 2021; P.-H. Li et al., 2022; Yeh & Barrington, 2023). There is a growing number of studies that demonstrate the importance of teachers’ well-being on children’s learning outcomes including literacy and numeracy (Benevene et al., 2020; Carroll et al., 2021; Hascher & Waber, 2021; Panadero et al., 2022; Yeh & Barrington, 2023). However, there are very few studies that have systematically explored this issue in a developing country characterized by low-resource, fragile and conflict affected context, especially in Afghanistan (D'Sa et al., 2023).

Children learning outcomes are driven by not only individual and teacher factors but also environmental characteristics (Vanbecelaere et al., 2021) such as home learning environment (Junge et al., 2021; Lehrl et al., 2020a; 2020b; Ndijuye & Benguye, 2023; Shao, 2023; Vanbecelaere et al., 2021). According to literature, a supportive HLE have been seen to compensate for the deficits of poor-quality schools and limited learning materials (Ndijuye & Benguye, 2023). The HLE has been conceptualized to include family’s structural characteristics (e.g. home language), parental beliefs and interest, and interactions between parent and child (e.g. parent-child book reading, storytelling, singing and parent supporting with homework) (Ndijuye & Benguye, 2023), and resources (e.g. books, puzzles) (Junge et al., 2021; Lehrl et al., 2020b; Ndijuye & Basil Tandika, 2022; Vanbecelaere et al., 2021). In a developing country context characterized by shortage of learning materials and teacher shortages (Ndijuye & Benguye, 2023), as in the case of Afghanistan, HLE has been identified as one of the important factors in influencing the level of child’s early learning outcomes. There are a few of studies on HLE and child learning in Asia including China, Iran, India, Turkey, and Philippines (Cheung et al., 2021). In particular, Cheung et al. (2021) found no study that explored the role of HLE on children learning outcomes in Afghanistan.

From the above discussion, it is crystal clear that little is known about the role of teacher well-being and home learning environment on primary children’s learning outcomes in crisis-affected contexts of Afghanistan. In fact, D'Sa et al. (2023) and Cheung et al. (2021) laments that while there is a growing body of teacher well-being and HLE research, respectively, studies examining the role of teachers’ well-being and HLE on children learning outcome in low-resource and fragile contexts are scarce.

This article, therefore, evaluates the role of teachers’ well-being and HLE on children literacy, numeracy, and social-emotional learning in Afghanistan. The novelty of this article is that it contributes to the scarce literature on role of teacher well-being and HLE on children learning outcomes in a fragile and crisis-affected context. In addition, we use multiple indicators of teacher well-being including basic needs fulfilment, social emotional well-being, professional development support, teacher voice, teacher recognition, respect, and prestige and enabling environment. The study findings will help efforts to promote teachers’ well-being, home learning environment and endeavours to enhance children learning outcomes.

The rest of this paper is structured as follows: The next section presents the background on education in Afghanistan, followed by the conceptualization of children learning outcomes. The next section presents materials and methods followed by the estimation strategy. The results are then presented and discussed in the next two sections; and the last section concludes.

Background

Education in Afghanistan

Promoting education ranks very high among the top priorities of all countries and this is understandable given that education is crucial for economic growth, poverty reduction and empowerment (Glewwe et al., 2014; Guimbert et al., 2008). Sustainable Development Goal 4 focus on promoting quality, inclusive and equitable education opportunities (Avelar et al., 2019).

The four decades of war, drought, recurrent natural disasters, widespread displacement, poverty and more recently the prolonged school disruptions due to COVID-19 pandemic have undermined access to quality education for children and adolescents in Afghanistan (Ahmadi et al., 2021; Noori, 2021). Indeed, several studies show that the quality of primary education in Afghanistan is poor (Ahmadi et al., 2021) and a higher proportion of children are out of school (Trani et al., 2019). This is further worsened by the fact that about 48% of the school teachers are not adequately trained and educated (Ahmadi et al., 2021; UNESCO, 2023). The public education system in Afghanistan is fragile and facing even more significant challenges. Development and humanitarian funding that once contributed significantly to improving the education system has been suspended and this compromise its quality and sustainability. This has led to inadequate infrastructure, shortages of learning materials, and decreased teaching quality, all of which have adversely affected the learning environment for children and undermined teacher well-being. To prioritize education of children and improve teachers well-being, donors and development partners started providing community-based classes and teachers training interventions in the country (UNESCO, 2023).

Education cannot wait first emergency response

Following, the protracted period of uncertainties, Afghanistan’s Education in Emergencies Working Group sought the support of Education Cannot Wait (ECW) to preserve access to education and promote and protect the well-being of girls, boys, adolescents, and female teachers. These developments saw the birth of Education Cannot Wait First Emergency Response (ECW-FER) in Afghanistan. The ECW-FER was implemented across seven Afghanistan provinces of Nangarhar, Kandahar, Sar-e Pol, Herat, Helmand, Kunduz, and Uruzgan by Save the Children, War Child UK, Children in Crisis, and International Rescue Committee. The project was implemented from November 2021 to March 2023 and aimed at protecting and promoting the well-being of around 3,8000 Afghanistan crisis-affected girls and boys, including adolescents through access to quality and flexible learning opportunities. Around 1090 teachers were supported to provide better quality education. Over 3300 shura members, communities and education stakeholders were strengthened to protect girls and boys and improve their wellbeing.

Children’s learning outcomes

Literacy and numeracy scores

Various approaches have been used to measure literacy and numeracy learning outcomes. These include the Annual State of Education Report (ASER) reading and math assessment (ASER, 2023; Johnson & Parrado, 2021; Richards & Islam, 2018), Early Grade Mathematics Assessment (EGMA) (Piper et al., 2018b) and early grade reading assessment (EGRA) (Dubeck & Gove, 2015). The ASER reading (literacy) and mathematics (numeracy) assessments (Johnson & Parrado, 2021) are administered orally to assess basic literacy and numeracy skills of children (ASER, 2023; Johnson & Parrado, 2021). The ASER reading and numeracy assessment has been widely used in India, Bangladesh, Afghanistan, Niger, and Kenya among other countries (ASER, 2023; Brown et al., 2022; Johnson & Parrado, 2021; Richards & Islam, 2018). The EGRA and EGMA are individually administered verbal assessments aimed at assessing children’s literacy and elementary math skills in the primary grades, respectively (Dubeck & Gove, 2015; Piper et al., 2018a; 2018b). The EGRA and EGMA have been widely used internationally including Kenya, Uganda, South Sudan, Somalia, Tanzania, India and many international literacy and numeracy programs (Deshpande et al., 2017; Dubeck & Gove, 2015; Piper et al., 2018a; 2018b; Wafula & Mulongo, 2020).

Social and emotional learning (SEL) competency

Social and emotional learning (SEL) is defined as a process of acquiring attitudes, knowledge, beliefs and skills to identify and manage emotions; to make good decisions; to care about others; to behave responsibly; to develop positive relationships and to avoid negative behaviours (Anthony et al., 2020; Calhoun et al., 2020; Jukes et al., 2021; Malhotra et al., 2021; Shao, 2023). SEL is concerned about building self-awareness, social awareness, self-management, relationship, responsible decision-making skills and plays an important role in a child’s personal and social wellbeing, successful transition to adulthood, and academic success (Anthony et al., 2020; Calhoun et al., 2020; Jones et al., 2019; Schiepe-Tiska et al., 2021) and has been used extensively in education programming (Thierry et al., 2022). In addition, SEL has been found to help students and teachers mitigate the trauma and intense stress (Jones et al., 2019; Thierry et al., 2022).

Materials and methods

Study design and sampling

This study used endline evaluation data drawn from ECW-FER program implemented in seven provinces of Afghanistan. The survey used population-based two stage-cluster sampling. The first stage determined the number of clusters, i.e. the number of provinces, then the number of districts per province based on the typical number of districts with education interventions. Cluster sample power analysis calculation was used with a 95% confidence interval, 5% margin of error, and a minimum detectable effect of 30%, along with an assumed non-response rate of 10%. Based on this, a total of 1407 primary school children, 1407 parents/caregivers and 315 teachers were randomly selected from villages with ECW-FER program interventions across seven provinces: Nangarhar, Kunduz, Sar-e Pol, Herat, Kandahar, Uruzgan, and Helmand. About 201, 201, and 45 school children, parents and teachers, respectively, were interviewed in each province. The districts covered under each province included: Nangarhar (Achin, Ghani Khil, and Nazyan); Kunduz (Khanabad, Gul Tapa, Aktash, Gul Abad, and Kundoz); Sar-el-Pol (Gosfandi, Kohestanat, Sancharak, and Sayad); Herat (Chisht e Sharif, Obe); Kandahar (Kandahar and Shah walikot); Uruzgan (Chura and Khas Uruzgan); and Helmand (Nowzad, Washeer).

Data collection and analysis

The data was collected from December 2022 to March 2023. Primary data collection was conducted via KoBo survey tools to boys and girls attending the Community Based Education (CBE) activities and their respective parents and teachers to answer the key study questions. All primary data collection tools were child-friendly and translated into local languages – Pashto and Dari. This study collected data on teacher well-being, children literacy, numeracy, disability, and home learning environment among others. For CBE teachers, the study collected data on teacher well-being. Household socio-economic data including household head’s disability status were collected from parents and or caregivers. Descriptive and multivariate logistic regression analysis were done using STATA version 18.

Measurement

Teacher well-being

This study assessed teacher’s well-being by asking teachers to answer 17 questions that were mapped into six domains of basic needs fulfillment, social emotional wellbeing, professional development, teacher voice, teacher recognition, and enabling environment (Benevene et al., 2020; Dreer, 2022; Jian et al., 2022; P.-H. Li et al., 2022; Panadero et al., 2022). The components of each of the six domains included:

• Perceived basic needs fulfilment: Included receiving a sufficient salary, getting paid regularly and on time, feeling safe living, working, and commuting to work, and accessing proper toilets and facilities.

• Social emotional well-being: The questions focused on enjoying my role as a teacher, rarely stressed due to work, handling any stressful situations at work, and accessing the necessary support when feeling stress.

• Professional development support: Receiving support from the school leader, having regular opportunities to share ideas with other teachers, and having opportunities to be observed and receive feedback.

• Teacher voice: Included being consulted by the leadership and management of the school and being confident to share ideas with school leadership and colleagues.

• Teacher recognition, respect, and prestige: Being respected as a teacher within the community or by the other staff in my school and having a positive relationship with students.

• Enabling environment: Asked questions around having sufficient teaching and learning materials.

We used factor analysis to compute the index score for each domain and aggregated them into teacher’s well-being index score. Teachers with a well-being index score in the median value or above were classified as having relatively high well-being status and those with a well-being index score below the median value had relatively low well-being status.

Home learning environment

Children living in violent abusive home environment were coded as one and zero otherwise. For safety issues, children who felt unsafe at home or on way to and from school were coded one and zero otherwise. Children who sought help from family members on education work where coded one and zero otherwise. These three binary variables were used as proxy measure of home learning environment.

Literacy and numeracy learning outcomes

The study created two binary outcome variables to measure literacy and numeracy learning outcomes. The research assistants administered the ASER reading or literacy and math assessment to measure literacy and numeracy scores to all children ages 5 − 13. The ASER literacy test assigned each respondent one of five literacy levels (Johnson & Parrado, 2021): can’t identify letters, can identify letters but not words, can read words but not a paragraph, can read a short paragraph but not story, and can read a longer story (Johnson & Parrado, 2021). On the same note, the ASER mathematics test assigned each respondent one of five numeracy levels (Johnson & Parrado, 2021): can’t identify numbers 1–9, can identify numbers 1–9, can identify numbers 10–99, can perform summation, and subtraction (Johnson & Parrado, 2021). In this study, children’s literacy scores were ordered from level: 1 (Beginner), 2 (letter), 3 (Word), 4 (Sentence), and 5 (Story with comprehension). Children with level 4 and 5 were denoted as having higher literacy and zero otherwise. Children’s numeracy scores were ordered from 1 (Beginner-unable to read the numbers), 2 (Number recognition (1–9), 3 (Number recognition (10–99), 4 (Summation), and 5 (Subtraction: 4 out of 6 questions correct). Children with level 4 and 5 were denoted as having higher numeracy and zero otherwise.

Social and emotional learning (SEL) competency

In this study, children’s social and emotional learning (SEL) skills were assessed using the SEL questions anchored on the Collaborative for Academic, Social, and Emotional Learning (CASEL) framework (Calhoun et al., 2020; Schiepe-Tiska et al., 2021; Thierry et al., 2022), widely used by development partners and donors to improve education and other outcomes in developing and humanitarian contexts (Jukes et al., 2021; SCI, 2019). It consists of 32 indicators contextualized to Afghanistan and focus on the measures of five competency domains (Schiepe-Tiska et al., 2021; SCI, 2019), as follows:

• Self-Awareness competency (emotional awareness, self-concept)

• Self-Management competency (emotional regulation, stress management/resilience)

• Social Awareness competency (empathy, tolerance, and diversity)

• Relationship Management competency (conflict resolution, social engagement)

• Responsible Decision-making competency (problem solving, critical thinking, assessing consequences

We used factor analysis to compute the SEL index score in each competency domain. The SEL index score was then used to create a binary outcome variable. Children with a SEL index score in the median value or above were classified having higher SEL competency and zero otherwise.

Children’s disability status

The study used the Washington Group Short Set (WG-SS) questions to measure children’s functional difficulties (Eide & Loeb, 2016). The tool has been extensively tested and focuses on six functional difficulties domains, i.e. hearing, seeing, mobility, remembering, or concentrating, communication, and self-care (Baart et al., 2023; Eide & Loeb, 2016). A child is classified as a person with a disability if they reported "A lot of difficulty" or "cannot do at all" in at least one of the six functional domains. On the other hand, a child is classified as a person without disabilities if they reported "No difficulty" or "some difficulty" across all domains (Baart et al., 2023).

Household poverty status

The study employed the poverty score card assessment tool (Skoufias et al., 2020) that uses several low-cost indicators, adapted from Afghanistan 2007/2008 National Risk and Vulnerability Assessment, to estimate the likelihood of the household having expenditures below a given poverty line. The poverty score card has been found to be as good as any of the regression-based estimators (Skoufias et al., 2020). The poverty scores were constructed from six indicators of (1) household dwelling type; (2) the number of rooms the household occupy; (3) the type of main toilet facility; (4) the main source of cooking oil; (5) the ownership of motorcycle or car; and (6) having access to irrigated land. The poverty scores were used to create a binary construct, with households scoring below the median poverty score within the sample classified as relatively poor and households on or above the median poverty score classified as non- poor.

Estimation strategy

A multivariate logistic regression model is employed to understand how teacher’s well-being and home learning environment are associated with children learning outcomes. The regression model is specified as: (1) $$\mathit{\text{CLO}}=\alpha +{\beta }_1\mathit{\text{TWB}}+{\beta }_2\mathit{\text{HLE}}+{\beta }_3\mathit{Y}+\epsilon$$(1) where CLO is the learning outcomes (literacy, numeracy, and social and emotional learning competency score) for children; Teacher well-being is captured by TWB; HLE represents home learning environment (safe learning environment and parental involvement in child education); Y capture the vector of other children, teacher and household characteristics; and $\beta$ are coefficients for estimation, and $\epsilon$ represents all variables not included in the model and $\alpha$ represent fixed-effects. ${\beta }_1$ and ${\beta }_2$ measure the role of TWB and HLE on children learning outcomes. The other independent variables utilised are age and gender of child, disability status, school length, household, and teacher factors. Given that we do not control for endogeneity, this study focuses on the correlation and not the causal relationship. The outcome variables (literacy, numeracy and SEL) are captured as dummy variables and as such multivariate logistic regression model is used for estimating Equation (1)(1) $$\mathit{\text{CLO}}=\alpha +{\beta }_1\mathit{\text{TWB}}+{\beta }_2\mathit{\text{HLE}}+{\beta }_3\mathit{Y}+\epsilon$$(1) (Cameron & Trivedi, 2022; Wooldridge, 2010). The logistic regression technique has been used extensively in estimating learning outcomes (Bollom et al., 2023; Rodriguez et al., 2021).

Results

Demographic characteristics

About 68% of the children had higher literacy implying they were able to read and understand paragraphs thus Level 4: Story and 5: Story with Comprehension. Sixty three percent of the children were classified as having higher numeracy as they attained levels 4 (Summation), and 5 (Subtraction: 4 out of 6 questions correct) (Table 1). The results show that 60% of the children had high SEL competencies. Overall, a third (34%) of the teachers perceived they had basic needs for teaching. The rest of teachers well-being factors were denoted to be met by about half of the interviewed teachers. This ranged from 52% of the teachers on social-emotional well-being to 56% on professional development support. Available socio-demographic data indicated that 53% of the children were girls. About a third of the children were in the 5-to 7, 8- and 9-years category and 6% of the children had disability. With reference to home learning environment, about 25% and 15% of the children highlighted that they were living in violent home environments and felt unsafe on way to and from school, respectively. About half of the children acknowledged that they sought help from parents and or family members on educational activities and or assignments. Almost all the children (92%) attended class for 10 weeks or more. Most household heads had low education with an average of 2 years spent in school. About a third of the teachers were females, which is relatively low.

Table 1. Descriptive statistics.

Variable	Description	Mean	Std. Dev	Min	Max
Outcome variables
Literacy	Higher literacy score (%)	68	47	0.0	100
Numeracy	Higher numeracy score (%)	63	48	0.0	100
SEL	Higher social emotional competency (%)	60	49	0.0	100
Teacher well-being
Basic need	Fulfillment of Basic needs (%)	34	47	0.0	100
Social emotional	Socio-emotional well-being (%)	52	50	0.0	100
Professional development	Professional development support (%)	56	50	0.0	100
Teacher voice	Teachers Voice (%)	53	50	0.0	100
Recognition	Teachers’ recognition, respect, and prestige (%)	53	50	0.0	100
Environment	Availability of teaching and learning materials (%)	55	50	0.0	100
Children home learning environments
Abuse home	Children living in violent home environment (%)	25	43	0.0	100
Safety issue	Children feel unsafe at home or on way to school (%)	15	35	0.0	100
Children seek help	Children seek help from parents and family members on education (%)	52	50	0.0	100
Children gender
Boys	Proportion of boys (%)	47	50	0.0	100
Girls	Proportion of girls (%)	53	50	0.0	100
Children age and disability
5 to 7-year-old	Proportion (%)	27	44	0.0	100
8-year-old	Proportion (%)	30	46	0.0	100
9-year-old	Proportion (%)	30	46	0.0	100
10 to 13-year-old	Proportion (%)	13	34	0.0	100
Disability	Children with a disability (%)	6	24	0.0	100
Children class length
Less than 10 weeks	Children’s attend school for less than 10 weeks (%)	8	27	0.0	100
10 weeks and above	Children’s attend school for 10 weeks or more (%)	92	27	0.0	100
Household and teacher factors
Head education	Years spent in school by household head	2.4	4.1	0	16
Female household head	Household head is female (%)	18	80	0.0	100
Poverty status	Non-poor households (%)	49	50	0.0	100
Female teacher	Female teacher (%)	33	47	0.0	100
Teacher age	Teacher’s age in years	27.5	6.84	18	60

Role of teachers’ well-being and home learning environment on children literacy and numeracy

The results of logistic regression analysis predicting children’s likelihood of having higher literacy and numeracy levels are shown in Table 2. The log-likelihood ratio tests (${\mathrm{\chi }}^2,$ 129, p = 0.000) and (${\mathrm{\chi }}^2,$ 151, p = 0.000) are significant denoting the goodness of fit of the literacy and numeracy logistic models, respectively.

Table 2. Regression results predicting literacy and numeracy of the children.

	Literacy^i		Numeracy^ii
	Odds ratio	95% CI	Odds ratio	95% CI
Teacher well-being
Fulfillment of basic needs	1.351	0.84–2.17	1.397	0.87–2.23
Socio-emotional well-being	0.519**	0.28–0.97	0.380***	0.20–0.72
Professional development support	3.189***	1.78–5.71	4.034***	2.26–7.21
Teachers voice	1.285	0.76–2.18	1.487	0.89–2.49
Teachers’ recognition, respect, and prestige	0.443**	0.23–0.84	0.442***	0.24–0.82
Availability of teaching and learning materials	1.602*	0.93–2.75	1.827**	1.17–3.12
Children home learning environments
Children living in a violent home environment	0.479***	0.30–0.77	0.486***	0.30- 0.78
Children feeling unsafe at home or on the way to school	0.503*	0.24–1.05	0.649	0.30–1.38
Children able to seek help from parents or family members	1.852***	1.22–2.82	1.943***	1.28–2.95
Children wellbeing
High positive SEL	0.820	0.52–1.30	1.636**	1.05–2.54
Children learning outcome
Literacy score (baseline level)	1.199	0.93–1.55
Numeracy score (baseline level)			1.032	0.81–1.32
Child characteristics
Girls (ref-boys)	0.705	0.45–1.11	0.660*	0.42–1.03
Child’s age (ref-5 to 7 years old)				.
8-year-old	2.084***	1.21–3.58	1.281	0.75–2.19
9-year-old	1.691*	0.99–2.89	1.146	0.67–1.97
10- to 13-year-old	1.303	0.64–2.64	1.192	0.58–2.44
Children with a disability (ref-children without a disability)	0.353**	0.14–0.87	0.301**	0.12–0.77
Household characteristics
Household’s education	1.009	0.96–1.06	1.047*	0.99–1.10
Female-headed household (ref-male)	0.628*	0.38–1.04	0.842	0.51–1.40
Non-poor households (ref-poor)	2.224***	1.41–3.50	1.717**	1.10–2.67
School attendance (ref-below 10 weeks)				.
Children’s school attendance: 10 weeks /above	2.440**	1.08–5.49	2.460**	1.10–5.51
Teacher characteristics
Female teachers (ref-male)	2.159***	1.32–3.54	2.318***	1.42–3.79
Teacher’s age	0.999	0.96–1.04	0.983	0.95–1.02
Number of observations	581		581
Pseudo R^2	0.18		0.20
Log-likelihood	−290***		−295***
Log-likelihood ratio test: LR chi^2 (22)	129***		151***

Statistically significant (p-value): *p < 0.05, **p < 0.01, ***p < 0.001. CI means Confidence interval.

Notes: 

ⁱChildren’s literacy scores at the endline survey were ordered from 1 (Beginner), 2 (letter), 3 (Word), 4 (Sentence), and 5 (Story with comprehension). Children with level 3 and 4 were denoted as having higher literacy and zero otherwise (Johnson & Parrado, 2021).

ⁱⁱChildren’s numeracy scores at the endline survey were ordered from 1 (Beginner-unable to read the numbers), 2 (Number recognition (1–9)), 3 (Number recognition (10–99)), 4 (Summation), and 5 (Subtraction (4 out of 6 questions correct)). Children with level 4 and 5 were denoted as having higher numeracy and zero otherwise (Johnson & Parrado, 2021).

Among the six domains of teachers’ well-being, the teacher’s professional development support and availability of teaching and learning material had a significant and positive relationship with children’s learning outcomes. Children taught by teacher’s who received professional development support were 3.19 and 4.0 times more likely to attain higher literacy and numeracy levels. In addition, children taught by teachers with adequate teaching and learning materials were 1.60 and 1.83 times more likely to achieve higher literacy and numeracy levels, respectively.

We now turn to home learning environment. Children living in homes where verbal or physical violence is prevalent had odds ratios that were significant and less than one for both literacy and numeracy outcomes. This implies that children experiencing violent home learning environment had lower likelihood of achieving higher literacy and numeracy levels. Children who sought help with learning from parents or family members had higher literacy and numeracy levels. Children with parents and family members that helped them with learning were 1.85 and 1.94 times more likely to have higher literacy and numeracy levels, respectively. Children with relatively high social and emotional learning competencies were 1.64 times more likely to have higher numeracy levels.

Eight- and 9-year old children were more likely to have higher literacy compared to the reference group of 5 to 7 years olds. Disability had a negative and significant association with children’s learning outcomes. Disabled children had lower odds of attaining higher literacy and numeracy levels than non-disabled children. The regression results showed that children living in households with higher socio-economic status were more likely to have higher literacy and numeracy. Children living in non-poor households were 2.22 and 1.72 times more likely to achieve highest literacy and numeracy levels, respectively. This implies that household socio-economic status is an important determinant of children’s learning outcomes.

Children’s regular attendance at school was associated with higher literacy and numeracy. Those who attended more than ten weeks during the school class period had 2.44 times 2.46 times the odds of having the highest literacy and numeracy scores. Parents and schools should ensure that children regularly attended school to ensure the achievement of improved learning outcomes. Children who were taught by female teachers were associated with higher literacy and numeracy levels. It suggests that having qualified female teachers is essential. Female teachers may not only help keep girls in schools, but they can also mentor and act as role models subsequently improving children’s learning outcomes. However, it is worth noting that only around one-third of the teachers in our sample were female. This suggests that increasing the recruitment of female teachers could enhance children’s learning outcomes.

Role of teachers’ well-being and home learning environment on children social emotional learning

The study used logistic regression analysis to predict children’s likelihood of having a high social and emotional learning competency at endline (Table 3). The log-likelihood ratio test is significant (${\mathrm{\chi }}^2,$ 116, p = 0.000), denoting the goodness of fit of the logistic model.

Table 3. Logistic regression results predicting the likelihood of children having a high SEL competency.

	Odds ratio	95% Confidence interval
Teacher well-being
Fulfillment of basic needs	1.254	0.86–1.83
Socio-emotional well-being	0.823	0.50–1.35
Professional development support	0.810	0.51–1.29
Teachers voice	2.501***	1.64–3.81
Teachers’ recognition, respect, and prestige	1.072	0.66–1.74
Availability of teaching and learning materials	1.283	0.83–1.99
Children’s home learning environments
Children living in a violent home environment	0.416***	0.29–0.60
Child’s feeling unsafe at home or on the way to school (ref-safe)	0.415***	0.25–0.70
Children able to seek help from parents or family members	0.860	0.62–1.19
School attendance
Children’s school attendance: 10 weeks and above (ref-below 10 weeks)	0.888	0.50–1.58
Child characteristic
Girls (ref-boys)	0.810	0.57–1.16
Child’s age (ref-5 to 7 years old)
8-year-old	0.717	0.47–1.09
9-year-old	1.249	0.81–1.92
10- to 13-year-old	1.151	0.69–1.92
Children with a disability (ref-children without a disability)	0.300***	0.13–0.67
Household characteristics
Household’s education	1.015	0.98–1.06
Female-headed household (ref-male household head)	1.926***	1.26–2.93
Non-poor households (ref-poor households)	0.820	0.59–1.15
Teacher characteristics
Female teachers (ref-male)	0.846	0.58–1.23
Teacher’s age	1.015	0.99–1.05
Observations	781
Pseudo R^2	0.11
Log likelihood	−468***
Log likelihood ratio test LR chi^2(20)	116***

Statistically significant (p-value): ∼ ***p < 0.01, **p < 0.05, *p < 0.1.

Teachers’ well-being was associated with children’s SEL competencies. Children taught by teachers with ability to voice up and be consulted by the school’s leadership and management were 2.5 times more likely to have higher SEL competencies. Home learning environment was found to be an important factor contributing to higher children’s SEL competency. Children living in households with incidents of violent situations or children experiencing a safety concern on the way to school had a lower likelihood of having a high SEL competency. Violent home environments are associated with a reduction in the likelihood of children having high SEL competencies. In addition, poor safety at home and on way to school were also associated with the lower likelihood of children having higher SEL competencies.

Children with disability were less likely than children without a disability to show a high SEL competency. It suggests that children with disability may be potentially facing challenges associated with lack of appropriate disability friendly equipment and infrastructure and negative stigma including discrimination in society. This subsequently restricts them in their involvement in learning activities and negatively affects their social-emotional learning skills. Children living in women headed households were 1.93 times more likely to have higher SEL competencies than those in a male headed household. It suggests that children living in a household headed by women were likely to be more resilient and develop better skills to adapt to a more challenging situation.

Discussion

This study used multivariate logistic regression to analyze the role of teacher’s well-being and home learning environment on children learning outcomes (literacy, numeracy and SEL competencies).

Teachers’ well-being and children learning outcomes

Teacher’s professional development and availability of teaching and learning material had a significant and positive relationship with children’s literacy and numeracy levels. Children taught by teachers with ability to voice up and be consulted by the school’s leadership had higher SEL competencies. Therefore, results demonstrate that teachers’ well-being improved children literacy, numeracy and SEL competencies. The result that teacher professional development improved children literacy, numeracy and social-emotional outcomes resonates with earlier studies (Fischer et al., 2018; Piper et al., 2018b; Shen et al., 2020; Siraj et al., 2022; X. Wang & Chen, 2022; M. Wang & Zhang, 2023; Wijaya et al., 2022). For example, Fischer et al. (2018) found small positive effects of teachers’ professional development on students’ academic achievement. In another study, Wijaya et al. (2022) highlight that teacher-parent support positively influenced students mathematics achievements.

The availability of teaching and learning materials is crucial as it makes access to knowledge and information easier and subsequently improves learning outcomes. Busingye and Najjuma (2015) found the distribution of learning and teaching resources to learners and teachers in Uganda had minimal influence on learners mathematics and english learning outcomes. However, simply providing textbooks and instructional materials was found to have small to medium effects in Kenya (Piper et al., 2018a). Instead, Piper et al. (2018a) found that interventions that combined teacher instructional materials, student books, and relevant teacher professional development produced strong results. Furthermore, teachers with ability to voice up and be consulted by the school’s leadership and management were found to improve children’s SEL competencies, further emphasizing the crucial role of teacher professional development (Piper et al., 2018a). Our results that children with higher social and emotional learning competencies are likely to have higher numeracy tally with Ingram et al. (2021) who noted that SEL improved student learning outcomes. Hence, children SEL competencies are important drivers of improved children learning outcomes.

Home learning environment and children learning outcomes

Home learning environment was found to be an important factor contributing to higher children’s learning outcomes. Children living in homes where verbal or physical violence is prevalent had lower literacy and numeracy levels as well as lower social emotional learning competencies. In addition, poor safety on way to and from school reduced children social emotional learning competencies. The importance of conducive home learning environment and parental involvement on childrens educational activities on children literacy, numeracy and social–emotional competencies is also emphasized by previous studies (Fry et al., 2018; Lehrl et al., 2020b; S. Li et al., 2023; Lloyd, 2018; Tamis-LeMonda et al., 2019). For instance, a meta-analysis conducted in 21 countries by Fry et al. (2018) show that violence on children have negative impact on educational outcomes. Our results that children who sought learning support from parents or family members had higher literacy and numeracy resonates with recent studies (S. Li et al., 2023; Ndijuye & Basil Tandika, 2022). For example, S. Li et al. (2023) notes that parental involvement in children’s educational activities enhances children’s social–emotional competencies. Ndijuye and Basil Tandika (2022) found that children demonstrated higher learning attainments when their fathers were closely involved in their education work.

The study found that children taught by female teachers had higher literacy and numeracy scores than from male teachers. Boys and children with disabilities had lower social and emotional learning competencies compared to their opposite counterparts.

Conclusion and policy implications

The main objective of this article was to examine how teacher’s well-being and home learning environment were associated with children learning outcomes in a conflict affected and fragile country. We use data from 1407 primary school children, 1407 parents or caregivers and 315 teachers randomly selected from villages in seven provinces of Afghanistan.

Teacher’s well-being improved children learning outcomes and SEL competencies. Teacher’s professional development and availability of teaching and learning material improved children’s literacy and numeracy. In addition, teacher’s recognition, and voice up improved children’s SEL competencies. From a policy perspective, school authorities need to roll-out regular teacher professional development and recognition interventions. This could take the form of refresher courses, receiving mentorship support and recognition from the school leader, having regular internal and external platforms for teachers to share and exchange ideas with their peers. The school authorities complemented by development partners should ensure adequate provision of teacher instructional materials and student books. Overall, there is scope for promoting a combined package that includes distribution of teacher instructional materials and student books complemented with relevant teachers’ professional development, recognition, and consultation.

This study found that unsafe and violent home learning environment lowered children learning outcomes. These results call for parents, communities, and all stakeholders to ensure conducive and safe home learning environments for children to enhance learning outcomes. These findings call for strengthening child protection and safeguarding at home, way to and from school and around school environments. Like previous studies, we found that greater parental involvement on children’s educational activities enhanced children learning outcomes. To maximize children’s learning outcomes in a conflict and fragile country with limited educational resources, involvement of parents is critically important. Understanding the importance of parents’ involvement in children’s education could potentially help in nurturing strong parental-child bond as well as strong parental-school partnership and engagement. As such, this study recommends policymakers in Afghanistan to consider drafting practical child development and education policies which encourage parents to be involved in children’s learning.

The study recommends that the public sector and development partners involved in education system need to promote the recruitment and professional training of teachers to improve children’s learning outcomes. The education delivery methods aimed at improving children learning outcomes should be inclusive of gender and disability.

Study limitations

This article has some noted shortcomings. First, the cross-section data used for analysis is based on recall and is not immune to recall bias. Second, the cross-section data allowed us to estimate associations and not causality given that we could not find valid instruments to run instrumental variables regression. Third, teacher’s well-being was assessed on six domains of basic needs fulfillment, social emotional wellbeing, professional development, teacher voice, teacher recognition, and enabling environment (Benevene et al., 2020; Dreer, 2022; Jian et al., 2022; P.-H. Li et al., 2022; Panadero et al., 2022). It should be noted that there are various approaches to assessment of teacher’s well-being beyond the one outlined in this study. For example Jian et al. (2022) divided teacher wellbeing into 3 factors of affective, cognitive and behavioral. It is recommended that future research should collect panel data to systematically estimate robust causal effects.

Ethics statement

The study was submitted to Save the Children’s Ethics Review Committee and approved (SCUS-ERC-FY2022-116) in September 2022, prior to data collection. The following important ethical considerations were taken. Permission was obtained from the local authorities. The survey was conducted in a safe space for children, parents and teachers and voluntary participation was ensured. The parents, teachers and children gave verbal informed consent and assent to participate in the study. Issues of privacy, anonymity, confidentiality, and do no harm were all ensured. A referral mechanism was put in place in case any child safeguarding/protection concerns arose.

Acknowledgements

The authors are grateful to all teachers, school children and caregivers for their participation in the study. The research assistants who collected the data are acknowledged.

Data availability statement

The data that support the findings of this study are available from the corresponding author, upon reasonable request.

Disclosure statement

This publication does not necessarily reflect the policy position of Save the Children International, any Save the Children Member organization and Education Cannot Wait. No responsibility is accepted by Save the Children International, any Save the Children Member organization and Education Cannot Wait for any errors or omissions contained within this article.

Additional information

Funding

The program implementation and endline evaluation of Education Cannot Wait-First Emergency Response in Afghanistan were funded by Education Cannot Wait. We are thankful to them for this funding.

Notes on contributors

Conrad Murendo

Conrad Murendo is a Research, Evidence and Learning Manager with Save the Children, Afghanistan. All the co-authors are working with Save the Children, Afghanistan, or Asia Regional Office.

Zahra Azemi

Zahra Azemi is the Education Technical Advisor.

Than Zaw Oo

Than Zaw Oo is a Regional Senior Research and Evaluation Officer.

Silvia M. Arlini

Silvia M. Arlini is a Regional Research and Evaluation Lead.

Men Chanbona

Men Chanbona is a Regional Research and Evaluation Officer.

Rachel Fermin

Rachel Fermin is the Education in Emergencies Humanitarian Technical Advisor.

Nishtha Shrestha

Nishtha Shrestha is a Regional Research and Evaluation Officer.

Sayed Maher Sadat

Sayed Maher Sadat is a Monitoring, Evaluation and Learning Manager.

Edwin Kimani

Edwin Kimani is the Head of Research, Monitoring, Evaluation and Learning.

Kasundu Bosco

Kasundu Bosco is a Monitoring, Evaluation and Learning Expert with Food and Agriculture Organization, Afghanistan. The authors research expertise lies in child education, child protection and mental health, poverty, health and nutrition, food security and livelihoods, climate change and resilience.