School dropout among children aged 5–17 in Türkiye and related factors: a logistic regression analysis

Abstract

Background

School dropout remains a significant concern in Türkiye, especially for children aged 5–17. This study aims to identify key factors associated with dropout risk using a nationally representative dataset.

Methods

Using microdata from the 2019 Child Labour Survey conducted by TurkStat, binary logistic regression analysis with sampling weights was performed. Logistic regression, a non-parametric statistical method, is used when the dependent variable is categorical with exactly two outcomes. Average marginal effects were calculated to interpret the influence of independent variables, including demographic, parental, and household characteristics.

Results

The likelihood of school dropout was higher among girls, older children (ages 15–17), those from large households, working children, and those with caregiving responsibilities. For instance, children working in agriculture were 30.6% more likely to drop out. Conversely, parental employment, maternal education, and participation in household chores reduced dropout probability.

Conclusion

Effective interventions should promote girls’ education, reduce caregiving responsibilities, and improve economic support for families. Policies must be regionally adapted and culturally sensitive to ensure equal educational opportunities for all children in Türkiye.

Background

School dropout among children remains a critical and persistent issue in Türkiye, with significant consequences for individual and societal development. Despite global and national commitments to ensure access to education, many children in Türkiye still drop out before completing primary or secondary education [1]. This challenge is closely linked to the level and accessibility of education, which play a significant role in determining children’s development, participation in society, and future opportunities. Education is a fundamental human right and a high social value from which everyone should benefit [2]. Through the Sustainable Development Goals (SDGs) that took effect in January 2016, the world committed to achieving accessible, equitable, and quality primary and secondary education by 2030 [3]. The UNESCO Institute for Statistics (UIS) and the Global Education Monitoring Report (GEM) estimated in 2017 that approximately 61 million primary school-aged children (ages 6 to 11), 62 million secondary school-aged children (ages 12 to 14), and 141 million high school-aged youth (ages 15 to 17) were out of school worldwide [4]. In 2018, it was estimated that around 258 million youth globally had either never started school or had dropped out at an early age [5].

Students who drop out early face higher unemployment risks, lower income levels, and adverse health outcomes, contributing to broader societal challenges such as increased poverty and reduced economic growth [6]. The impact of education on lifelong health is both direct (educated individuals learn more about health promotion) and indirect (educated individuals generally have higher incomes and can access better housing, food, and high-quality healthcare). Thus, better education translates to better life outcomes for children [7].

The Turkish education system is divided into three levels: primary school (ages 5–10), middle school (ages 10–14), and high school (ages 14–18), with compulsory schooling typically beginning at age six and continuing through the end of high school [8]. However, significant barriers to educational access persist, especially for girls, due to factors like gender inequality, early marriage, financial hardship, and traditional family roles [9, 10]. Parental socioeconomic status and education levels strongly influence children’s school attendance, with children from low-income or less-educated families facing a significantly higher risk of dropping out [11, 12]. Child labour further worsens dropout rates, especially in rural areas where contributing to household income becomes a priority [13, 14]. Additionally, regional disparities, particularly in eastern and southeastern Türkiye, highlight more severe gender inequalities and infrastructure gaps, deepening educational divides [15, 16]. These realities highlight the need for education policies in Türkiye that expand access and address the underlying social and economic inequalities that shape it.

Children’s school attendance may vary according to their circumstances. Depending on these conditions, children may fully attend school, be deprived of education, or drop out [17]. Although every child has the right to a healthy start in life, realising this right largely depends on family conditions and parents’ importance on education. Parents’ educational level and the value of education play an influential role in a child’s schooling process [18]. Additionally, factors such as lack of school access in neighbourhoods like slums, infrastructure issues [19], migration [20], the child’s gender and birth order [21], health status [22], lack of interest in education, low academic achievement, and financial constraints [23] can increase the likelihood of a child dropping out or never starting school. As a result of dropping out or lack of access to education, children and society face various problems. Children who drop out often have fewer job opportunities due to limited education and skills, leading them to join the workforce early [24].

The transition from primary to secondary education is a critical period. Academic failure and early parenthood during this period, for instance, increase the likelihood of a child dropping out [25, 26]. Therefore, the determinants for working children dropping out of school are complex and involve multiple interrelated factors.

The family’s socioeconomic status, income, and parental education are significantly related to the likelihood of children dropping out of school. Studies show that children from low-income, less-educated families have lower continuation rates in education and a higher likelihood of dropping out [27–30]. Stressful changes in the family, challenges faced in early childhood, parental attitudes toward education, and children’s abilities are primary factors that increase the risk of school dropout, independent of sociodemographic factors [31–34].

Compulsory education has reduced the dropout rate by 24% and reduced the likelihood of girls spending long hours on household chores. While this highly correlates with school enrollment in rural areas, it is not as pronounced in urban areas. The effect of compulsory education is decisive in low-income families [11]. Children’s education in Türkiye is linked to economic cycles; school attendance rates rise during economic growth periods and decline in times of crisis, with the most apparent effects seen among children with less-educated parents and those in rural areas [35].

Economic factors also significantly contribute to children’s challenges in staying in school. Individual or general economic conditions, often influenced by disasters, statistically increase the risk of children leaving primary school [36]. It has been demonstrated that socio-economic deprivation is significantly related to school attendance in Türkiye. It was reported that children living below the poverty line were approximately three times more likely to drop out compared to their middle-income peers [37]. In addition, it has been argued that income inequality in Türkiye is reinforced by unequal access to quality education, thereby maintaining intergenerational poverty cycles [38, 39].

According to ILO (2021), the primary reasons for school dropout among children in Türkiye include contributing to the household economy, learning a trade, and helping with family finances. Approximately 30.8% of children work in agriculture, 23.7% in industry, and 45.5% in the service sector, representing forms of child labour that prevent school attendance and negatively impact children’s physical, mental, and social well-being, especially for those under 15 [13]. Child labour remains a significant public health issue in low- and middle-income countries and is linked to negative physical and mental health outcomes [40].

Research indicates that factors such as parental education level, household income, family size, child’s age, school-related factors (such as school type and distance), participation in school activities, absenteeism, parental guidance in home education, and engagement in household work are primary determinants of school dropout [41]. Health conditions, particularly mental health issues, significantly increase the risk of school dropout and are associated with childhood health problems among those who drop out [42, 43].

Besides parental education, children’s characteristics also influence their schooling. For instance, students without learning disabilities are less likely to drop out [44–46]. Childhood health conditions impact school dropout, with early mental disorders, in particular, being a determining factor in the risk of leaving school [47]. Attachment to antisocial peers can also influence school dropouts [48]. Additionally, Zengin (2021) emphasizes the role of silent risk types and antisocial behaviours among high school students, highlighting how perceptions of failure can silently increase dropout risk [49]. Although exact dropout reasons are not always clear, research highlights variables such as the child’s gender [50, 51], age [52, 53], caregiving responsibilities [54], duties within the family [33], household size [55–57], and parents’ educational level [12, 58, 59] and employment status [60, 61] as significant.Despite numerous legally binding initiatives to combat and eliminate school dropouts, it remains a significant issue in the industrialized 21 st century [62]. For example, Zorbaz and Özer (2020) applied an ecological framework combining student- and school-level factors to show how environmental risks, academic achievement, and school climate interactively predict dropout risk in Türkiye [63]. In Türkiye, advanced statistical methods have been used in recent studies to examine the causes of school dropout. The systematic review by Eranıl (2024) [64] analyzed dropout causes in Türkiye between 2009 and 2022, identifying parental education, academic failure, and peer pressure as consistent predictors. However, it lacked a focused statistical modelling approach across multiple variables and age categories. These findings suggest that while individual factors have been explored, an integrated theoretical framework incorporating household structure, child labour status, parental background, and caregiving burdens within a national child population (ages 5–17) is still missing.

Furthermore, the child’s gender, father’s education, and child’s age play essential roles in determining dropout likelihood [65]. Social norms and biases also contribute to dropout rates, with research indicating that academically low-achieving girls are more likely to drop out than boys, highlighting gender discrimination in educational opportunities [66]. Similarly, cultural factors, lack of interest in education, absence of schools or facilities, poverty, and disabilities or mental disorders stand out as reasons children drop out [67].

Additionally, especially in situations involving migration or seasonal labour, the need to contribute to family income can lead children to lose interest in lessons and experience declines in academic performance [68, 69]. Working may not wholly eliminate children’s access to formal education but can influence educational outcomes and motivation [70].

Given that staying in school causally reduces risks for diabetes and mortality [71], school closures during the COVID-19 pandemic inflicted substantial harm on children’s and adolescents’ health, manifesting in increased distress, anxiety, and depressive symptoms [72]. Considering the mental health effects of absenteeism, an interdisciplinary approach is essential to achieve success and ensure continued education [73]. Additionally, continuing education helps individuals discover suitable religious truths [74] and develop critical thinking skills [75]. Adolescents who stay in school have the potential to create healthier responses, improve coping skills for stress, and strengthen social bonds, positively impacting mental health [76].

Within the framework of the literature gap mentioned above, this study aims to analyze the main factors affecting school dropout among children aged 5–17 in Turkey using a comprehensive model. In this regard, the study will seek answers to the following questions:

• Research Question 1. Is there a relationship between children’s school dropout rates and parental factors such as education level, employment status, and age?

• Research Question 2: Is there an association between children’s school dropout rates and employment status?

• Research Question 3: Is there a relationship between children’s school dropout rates and factors like gender, household size, and age?

• Research Question 4: Could a link be between children’s school dropout rates and their involvement in household chores or caregiving responsibilities?

In line with these aims, the following hypotheses were formulated to guide the analysis:

• H₁. Older children, girls, and those from larger households are likelier to drop out of school.

• H₂. Children whose parents have lower educational levels are more likely to drop out of school.

• H₃. Children whose parents are unemployed are more likely to drop out of school.

• H₄. Employed children are likelier to drop out of school than those who are not working.

• H₅. Children involved in caregiving responsibilities are likelier to drop out, while those helping with household chores are less likely to drop out.

Methods

Data source

This study used the 2019 Child Labour Survey Micro Data Set provided by the Turkish Statistical Institute (TurkStat). TurkStat last shared data for the year 2019. The Child Labour Survey Micro Data Sets are prepared not only for news bulletins, statistical tables, and databases but also to enable users to conduct other statistical studies. Therefore, the variables in the microdata sets are generally shared for purposes such as econometric modelling studies or data mining analyses rather than for producing detailed cross-tabulations. Based on labour force survey results from the relevant period, it does not include total population figures for age groups outside this range [77].

Since 1988, TurkStat has conducted the Household Labour Force Survey, the primary data source for the labour market, by the norms and standards set by the International Labour Organization. Since 2004, it has followed Eurostat standards, making it a significant study. The survey uses a two-stage, stratified cluster sampling method. A rotation pattern ensures a 50% overlap between two consecutive periods and the same periods of consecutive years, with eight sub-samples for each period. The sample size for each period is distributed equally across weeks. As of 2014, the periodic sample size is 44,000 households, based on Eurostat Regulation No. 577/98, and the design considers Türkiye’s 2014 administrative division. The Child Labour Survey was conducted from October to November and December 2019. The survey uses a face-to-face interview and Computer-Assisted Personal Interviewing (CAPI) method conducted by interviewers, and data is directly recorded on laptops during field application [77].

Weighting is performed on the dataset obtained from sampling to arrive at values that represent the population. The study calculated design weights based on the selection criteria to arrive at the final weights; non-response adjustments and external distribution controls were performed. In the weighting process, age group, gender, IBBS Level 2, rural-urban, and household size information were used as external controls [77].

This research reflects the national context and is significant for shedding light on needs, as it enables international and national comparisons.

All residential areas in Türkiye were included in the sample selection. The Household Labour Force Survey sampling unit is the address, covering individuals in the 5–17 age group within households. The data set pertains to children in the 5–17 age group. The selection process of the sample to be included in the study is given in Fig. 1.

Fig. 1.

Selection process of children aged between 5 and 17 years among individuals in the Child Labor Survey

Measures

The dependent variable of the study is the variable related to the age group of 5–17 years and the status of not attending any formal education program (primary school, secondary school or high school and equivalent vocational school), open education school, or not participating in any apprenticeship training organized by the Ministry of National Education. Due to table space limitations, this variable is abbreviated as “dropping out of school.” Therefore, the dependent variable indicates whether the child has dropped out of school. Responses are coded as “yes” and “no” The model’s dependent variable is coded as 0 if the child is still attending school (yes) and one if the child is not (no).

The independent variables in this study include: gender (girl, boy), age of the child (5–11, 12–14, 15–17 years), mother’s age (20–34, 35–54, 55 + years), father’s age (20–34, 35–54, 55 + years), mother’s education (illiterate, less than high school, high school, university), father’s education (illiterate, less than high school, high school, university), sector (agriculture, industry, service, unemployed), participation in household activities like laundry, washing dishes, ironing, cooking, cleaning, or shopping (usually, sometimes, never), contribution to caregiving activities if there is an elderly/disabled/sick household member (usually, sometimes, never), mother’s employment status (employed, unemployed), father’s employment status (employed, unemployed), and household size (3 and below, 4, 5, 6 and above) [77].

Statistical analysis

Survey statistics in Stata 15 (Stata Corporation) were used to account for the complex sampling design and weights, and a weighted analysis was performed. Initially, frequencies and percentages were calculated based on the children’s school dropout status. A chi-square test of independence was conducted to examine the relationship between the dependent and independent variables. Subsequently, a binary logistic regression analysis was used to identify factors associated with the children’s school dropout status.

Non-parametric statistics are used for categorical data (nominal, ordinal). Logistic regression, which is a non-parametric statistical method, is used when the dependent variable is categorical with exactly two outcomes [78].

In social sciences, especially in socio-economic research, some of the variables examined are measured on a sensitive scale, while others consist of dichotomous data such as positive-negative, successful-unsuccessful, and yes-no. Dichotomous data are the most commonly used form of categorical data [79]. When the dependent variable is dichotomous categorical data, logistic regression analysis is used to examine the cause-and-effect relationship between the dependent variable and the independent variable(s) [80].

Logistic regression is a statistical method that allows for classification in accordance with probability rules by calculating the predicted values of the dependent variable as probabilities [81].

The logistic model was initially developed for use in survival analysis. Here, the dependent variable (Y) takes values of 1 or 0, depending on whether the event of interest occurs. The expected value, E(Y), never falls below 0 or above 1. Therefore, the predicted values of y ^ in the logistic model range between 0 and 1 ([82]83).

Logistic model is written as,

1

After dividing the numerator and denominator of Eq. 1 by or exp,

2

In the equation, there is a condition that and X values are qualitative or quantitative independent variables.

The coefficients estimated in the logistic model show the change in the z-score against a unit change in the independent variable. Therefore, the interpretation of the coefficients in the logistic model is not meaningful. The sign of the coefficient gives the direction of the relationship between the independent variable and the probability; the size of the coefficient does not directly give the magnitude of the effect.

The marginal effect is the partial derivative of the regression equation with respect to each independent variable. In the linear regression model, the slope coefficient equals the marginal effect for each variable. It is interpreted as the change in the dependent variable in response to a unit change in the independent variable. In nonlinear regression models such as the logistic model, the slope is not constant at all points but changes according to the observation points; that is, the marginal effect changes according to the observation points. In nonlinear regression models, the average marginal effects are interpreted since the interpretation according to any point will not be meaningful.

Marginal effects can be calculated differently: marginal effects for means (MEMs) and average marginal effects (AMEs). When calculating the marginal effects for the means, the mean of each variable is calculated, and the marginal effects are obtained for these means. Since the mean for dummy variables is statistically invalid, the marginal effects for the means are not appropriate in models with dummy variables among the independent variables. When calculating the average marginal effects, marginal effects are calculated for all units in the data set, that is, for each observation point, and these marginal effects are averaged. Since categorical independent variables are used in the models, average marginal effects were calculated [84].

Characteristics of study participants

The frequencies and percentages of independent variables used in the model are presented in Table 1, which shows factors associated with children’s school dropout status and Chi-square test statistics.

Table 1.

Distributions and Chi-square test statistics of factors related to children’s school dropout status

Variables	Dropout n (%)	Attending n (%)	Total n (%)	χ²	p
Gender
Boy	1452 (12.6)	10,028 (87.4)	11,480 (51.0)	149.000	0.702
Girl	1374 (12.5)	9638 (87.5)	11,012 (49.0)
Age group
5–11	1783 (14.9)	10,215 (85.1)	11,998 (53.3)	427.526	0.000
12–14	242 (4.5)	5170 (95.5)	5412 (24.1)
15–17	802 (15.8)	4281 (84.2)	5083 (22.6)
Mother’s age
20–34	1123 (17.4)	5313 (82.6)	6436 (28.6)	253.584	0.000
35–54	1615 (10.3)	14,066 (89.7)	15,681 (69.7)
55+	83 (23.5)	287 (76.5)	370 (1.7)
Father’s age
20–34	639 (25.1)	1904 (74.9)	2543 (11.3)	440.518	0.000
35–54	1997 (10.6)	16,760 (89.4)	18,757 (83.4)
55+	190 (15.9)	1002 (84.1)	1192 (5.3)
Mother’s education
Illiterate	668 (20.5)	2584 (79.5)	3252 (14.5)
Below high school	1536 (11.0)	12,406 (89.0)	13,942 (62.7)	222.423	0.000
High school	354 (11.5)	2726 (88.5)	3080 (13.7)
University	268 (12.1)	1950 (87.9)	2218 (9.9)
Father’s education
Illiterate	171 (28.9)	421 (71.1)	592 (2.6)	183.279	0.000
Below high school	1799 (13.1)	11,884 (86.9)	13,683 (60.8)
High school	473 (10.1)	4212 (89.9)	4685 (20.8)
University	383 (10.8)	3149 (89.2)	3532 (15.7)
Sector
Agriculture	141 (36.8)	242 (63.2)	383 (1.7)	389.277	0.000
Industry	75 (37.9)	123 (62.1)	198 (0.9)
Service	99 (24.6)	303 (75.4)	402 (1.8)
Unemployed	2511 (11.7)	18,998 (88.3)	21,509 (95.6)
Homework involvement
Usually	271 (12.2)	1955 (87.8)	2226 (9.9)	493.672	0.000
Sometimes	421 (5.7)	6913 (94.3)	7334 (32.6)
Never	2134 (16.5)	10,798 (83.5)	12,932 (57.5)
Caregiving activities
Usually	36 (20.1)	143 (79.9)	179 (0.8)	18.881	0.000
Sometimes	92 (9.3)	894 (90.7)	986 (4.4)
Never	2698 (12.7)	18,642 (87.3)	21,340 (94.8)
Mother’s employment
Employed	896 (11.8)	6714 (88.2)	7610 (33.8)	6.542	0.011
Unemployed	1930 (13.0)	12,952 (87.0)	14,882 (66.2)
Father’s employment
Employed	385 (11.2)	3047 (88.8)	3432 (15.2)	23.153	0.000
Unemployed	2815 (13.4)	18,192 (86.6)	21,007 (84.8)
Household size
3 and below	224 (15.1)	1265 (84.9)	1489 (6.6)	109.248	0.000
4–5	794 (11.6)	6449 (88.4)	7243 (32.8)
6 and above	1165 (13.5)	6175 (86.5)	7340 (60.6)

Home-work situation This variable assesses whether the child engages in household activities such as laundry, dishwashing, ironing, cooking, cleaning, or shopping for the household, Maintenance Activities This variable evaluates whether the child contributes to caregiving activities for an elderly, disabled, or sick relative within the household, *Unemployed This refers to individuals who are unemployed or not part of the labor force

^ap<0.01

^bp<0.05

^cp<0.10

According to findings in Tables 1 and 51% of the children in the study are female, with 53.3% aged 5–11, 24.1% aged 12–14, and 22.6% aged 15–17. Regarding the mothers’ ages, 28.6% are between 20 and 34, 69.7% are between 35 and 54, and 1.7% are over 55. Fathers’ ages reveal that 11.3% are between 20 and 34, 83.4% are between 35 and 54, and 5.3% are over 55.

Among mothers’ educational levels, 14.5% are illiterate, 62% have less than a high school education, 13.7% have a high school education, and 9.9% have a university degree. Fathers’ educational backgrounds show that 2.6% are illiterate, 60.8% have less than a high school education, 20.8% have a high school education, and 15.7% have a university degree. Of the children who dropped out of school, 1.7% are employed in agriculture, 0.9% in industry, 1.8% in the service sector, and 95.6% are unemployed.

Regarding household chores, children who are not in school report that 9.9% usually, 32.6% sometimes, and 57.5% never participate in activities such as laundry, dishwashing, ironing, cooking, cleaning, or shopping. Regarding caregiving for elderly, disabled, or sick relatives, 0.8% of non-school-attending children usually, 4.4% sometimes, and 94.8% never contribute. Parents’ employment status shows that 33.8% of mothers and 83.2% of fathers are employed. Household size among school dropouts shows that 8.6% live in households with three or fewer members, 32.2% in four households, 26.5% in five households, and 32.6% in households with six or more members.

The analysis in Table 1 indicates significant differences between school dropout status and factors such as the child’s age, parental age and education levels, parental employment status, the sector in which the child is employed, and household size. Additionally, significant differences are found between school dropout status and participation in household chores and caregiving for a relative.

Results

A binary logistic regression model was employed to determine the factors associated with school dropout among children included in the study.

This study used the Hosmer-Lemeshow test to determine the model’s goodness of fit (H0 = Observed data and regression model fit well). A p-value exceeding the test level (p > 0.05) indicated that the information in the current data was sufficiently extracted, and the model fit was good [85]. The Pseudo R² value for the binary logistics model was calculated as 0.22. The estimated model results are presented in Table 2.

Table 2.

Estimated model results and marginal effects of factors preventing children’s school attendance

Variables	β	Std. Error	95%CI		ME	Std. Error	VIF
			Lower	Upper
Gender (reference: girl)
Boy	−0.134a	0.050	−0.233	−0.036	−0.014a	0.005	1.04
Age of the child (reference: 15–17)
5–11	−0.329a	0.070	−0.466	−0.192	−0.040a	0.009	2.12
12–14	−1.166a	0.098	−1.359	−0.974	−0.111a	0.009	1.69
Mother’s age (reference: 20–34)
35–54	0.444a	0.063	−0.567	−0.320	−0.049a	0.007	1.46
55 + years	0.050	0.206	− 0.453	0.354	−0.006	0.025	1.33
Father’s age (reference: 55 + years)
20–34	0.440a	0.147	0.151	0.729	0.055a	0.017	3.80
35–54	−0.135	0.129	−0.388	0.119	−0.014	0.014	3.51
Mother’s education (reference: illiterate)
Education below high school	−0.432a	0.074	−0.577	−0.287	−0.048a	0.009	2.45
High school	−0.181c	0.103	−0.383	0.020	−0.021c	0.012	2.30
University	0.059	0.125	−0.186	0.305	0.008	0.016	2.63
Father’s education (reference: university)
Illiterate	0.784a	0.156	0.479	1.090	0.098a	0.022	1.44
Education below high school	0.145c	0.088	−0.027	0.318	0.015c	0.009	3.19
High school	−0.010	0.092	−0.190	0.170	−0.001	0.009	2.33
Sector (reference: unemployed)
Agriculture	1.869a	0.149	1.576	2.161	0.306a	0.031	1.06
Industry	1.492a	0.196	1.108	1.876	0.228a	0.039	1.04
Service	0.918a	0.153	0.617	1.218	0.123a	0.025	1.06
Home-work situation (reference: never)
Usually	−0.345a	0.086	−0.514	−0.177	−0.041a	0.009	1.18
Sometimes	−1.147a	0.070	−1.284	−1.010	−0.105a	0.005	1.21
Maintenance activities (reference: never)
Usually	0.631a	0.236	0.168	1.094	0.079b	0.035	1.04
Sometimes	0.326b	0.141	0.049	0.603	0.038b	0.018	1.07
Mother’s employment status (reference: unemployed)
Employed	−0.147a	0.057	−0.258	−0.036	−0.015a	0.006	1.11
Father’s employment status (reference: unemployed)
Employed	−0.140b	0.067	−0.271	−0.009	−0.015b	0.007	1.11
Household size (reference: 6 and above)
3 and below	−0.210b	0.097	−0.401	−0.019	−0.023b	0.010	1.31
4	−0.229a	0.067	−0.359	−0.099	−0.025a	0.007	1.64
5	−0.281a	0.067	−0.411	−0.150	−0.030a	0.007	1.44
Constant	−0.444b	0.188	−0.812	−0.076	-	-	-

House-work status Participation in household activities such as doing laundry, washing dishes, ironing, cooking, cleaning, or shopping for the household, Maintenance activities This variable assesses whether the child contributes to caregiving activities for an elderly, disabled, or sick relative within the household

^ap<0.01

^bp<0.05

^cp<0.10

The model also examined the presence of multicollinearity among independent variables. Variance Inflation Factor (VIF) values were used to assess multicollinearity, with VIF values of 5 and above indicating moderate multicollinearity and 10 and above suggesting high multicollinearity [86]. According to the VIF results shown in Table 2, no variable was found to cause multicollinearity issues in the model.

An examination of Table 2 reveals that factors such as the child’s gender, age, parental age and education, the sector in which the child works, household chore support, caregiving duties for dependents, parental employment status, and household size influence a child’s likelihood of dropping out of school.

Compared to children in households with six or more members, the probability of dropping out is 2.3% points lower for those in households of three or fewer, 2.5% points lower for those in households of four, and 3.0% points lower for those in households of five. The results show that the probability of dropping out of school for male children is 1.4% points lower than for female children. Compared to children aged 15–17, the probability of school dropout is 4.0% points lower for children aged 5–11 and 11.1% points lower for those aged 12–14. Compared to children with mothers aged 20–34, the probability of dropout for those with mothers aged 35–54 is 4.9% points lower. In contrast, compared to children whose fathers are over 55, the probability of school dropout for those with fathers aged 20–34 is 5.5% points higher.

Compared to children with illiterate mothers, the probability of dropping out is 4.8% points lower for those whose mothers have less than a high school education and 2.1% points lower for those with a high school education. Compared to children with university-educated fathers, the probability of dropping out is 9.8% points higher for those whose fathers are illiterate and 1.5% points higher for those with less than a high school education.

Compared to children who do not work, the probability of dropping out is 30.6% points higher for those working in agriculture, 22.8% points higher for those in industry, and 12.3% points higher for those in services. Compared to the reference group (never helping), the probability of dropping out is 4.1% points lower for children who “usually” help with household chores and 10.5% points lower for those who “sometimes” help. Compared to the reference group (never providing care), the probability of dropping out is 7.9% points higher for children who “usually” provide care and 3.8% points higher for those who “sometimes” provide care. The probability of dropping out is 1.5% points lower for children with employed parents.

Discussion

This study aimed to identify risk factors associated with school dropout among children in Türkiye. Overall, the results reveal that factors such as gender, age, parental age and education level, parental employment status, sector of child employment, frequency of household chores, frequency of caregiving activities, and household size are significantly associated with a child’s likelihood of leaving school.

Previous research has shown that girls are more likely to leave school due to gender-based discrimination, early marriage, and domestic roles [9, 10, 87–91]. This study confirmed that, within the Turkish context, girls remain in a disadvantaged position and that gender inequalities persist. According to the study, girls are slightly more likely to drop out than boys. While the observed gender difference is statistically significant, the effect size is relatively small. Nonetheless, this finding aligns with previous literature suggesting that girls may be more susceptible to irregular attendance, gender-based expectations, and environmental challenges that influence school continuity [9, 87–89]. Although the schooling rate in Türkiye has increased over the years, girls are still disadvantaged. When the distribution of literacy by gender in Türkiye for ages 6 and above and ages 15 and above 2008–2023 is analyzed, the number of illiterate women is higher than that of illiterate men in all years [90]. Furthermore, In Türkiye, a young woman is three times more likely to be out of education and employment than a young man [91]. It is thought that factors such as girls assuming adult roles at an early age, being forced into marriage under the pretext of early puberty, and being limited to basic literacy—with the assumption that their husbands will support them after marriage—along with conservative family concerns about coeducation, may contribute to girls dropping out of school against their will [9]. This suggests that, beyond individual preference, dropout among girls may stem from a range of socio-cultural issues related to gender imbalance, patriarchal values, and educational inequality [10].

It is well-established in the literature that the risk of school dropout increases with age [92–95]. It is also stated that age-related biological and psychological changes during adolescence and socioeconomic pressures increase the risk of school dropout [92–96]. According to the research results, older children are likelier to drop out of school. In support of this studies show that dropout rates increase among children in higher age groups [92–94]. Literature suggests that each additional year of age raises dropout risk by 2.1 times [95].

Türkiye’s education system is divided into three levels based on age: primary, middle, and high school. Primary school covers ages 5–10, middle school includes ages 10–14, and high school is for ages 14–18. Compulsory education in Türkiye typically begins around age 6 and continues through high school [97]. Notably, the transition from primary to middle school coincides with adolescence, a period marked by biological, physical, and psychological changes [96]. Children face numerous developmental changes during this time, which can be negatively associated with their school life [98]. For example, factors such as academic failure [99], peer bullying [100], and family trauma [101] can make it difficult for children to maintain regular school attendance. The transition from middle to high school, covering ages 14–18, represents a critical phase for adolescents [102]. During this period, there is an observed increase in school absenteeism [103], heightened interest in the opposite sex [104], and a greater risk of substance use [105]. In particular, for girls, these risks may be compounded by outcomes such as adolescent pregnancy and early marriage. Due to unsafe sexual relationships, many young girls may leave school and feel compelled to marry, risks that become more common during this period [106, 107]. Economic poverty is also a significant factor in this age group [108]. As a result, Children from economically disadvantaged families may be forced to enter the workforce to help support their household [109]. This situation may be associated with detachment from school and early dropout [110]. Family conflicts are another key factor associated with school dropout [88]. Children may decide to leave school in defiance of their parents. During these transitional periods, factors such as academic achievement, psychosocial support, and family structure play a critical role in whether children continue their education.

In the literature, parental age is generally associated with economic stability and life experience [111–115]. This study found that in the context of Türkiye, higher parental age is statistically associated with a greater likelihood of children remaining in school. The study found that higher parental age is associated with a lower likelihood of school dropout. This finding aligns with previous studies indicating that children of older parents are generally more advantaged [111–113]. Additionally, older parents are usually at a higher level in their careers and have a more stable financial situation [114]. This economic security may be related to an increased ability to provide educational opportunities for their children and invest in additional resources such as tutoring or educational materials. Moreover, through their life experiences and academic background, these parents can better understand the importance of education [115]. Thus, they can powerfully convey the importance of education to their children and motivate them to succeed in school. The maturity and experience gained with age enable parents to participate more actively in their children’s educational processes [12]. They can support their children’s school success more effectively as they are more competent in problem-solving, communication and support. In addition, the family structures of older parents are generally more stable, which helps children focus more on their education, thus reducing the risk of dropping out [116]. In some societies, older parents place a higher value on traditional principles, which may further reinforce their support for children’s academic completion [117, 118]. In short, having older parents may be positively associated with a child’s education through multiple channels, including economic stability, life experience, and family consistency.

While prior studies [12, 59, 119] have acknowledged the influence of parental education on children’s academic performance, this study quantitatively demonstrated how both the mother’s and father’s education levels are directly associated with the risk of school dropout in the Turkish context. The results indicate that higher parental education levels and employment status are associated with higher school attendance. Specifically, children with working parents are more likely to continue their education. The literature shows that parental education positively influences children’s academic performance, with the father’s education often having a more pronounced effect than the mother’s [59]. The findings supported Hypothesis 3, as children of unemployed parents showed significantly higher probabilities of leaving school early. Parents’ education levels shape their professions and income [12]. Family dynamics, such as socioeconomic status, parental education, and family structure, are significantly associated with students’ likelihood of continuing their education [119]. The relationship between education level and social welfare may be linked to influenced this situation. The results supported Hypothesis 2, indicating that children whose parents had lower levels of education—especially those with illiterate mothers or fathers—were more likely to drop out.

The negative associations of child labour on education are widely documented in the literature [14, 120]. The findings of this study clarified that, regardless of the sector, working children in Türkiye face a significantly higher risk of leaving school. The findings show a significant association between child employment and increased likelihood of school dropout. Previous studies support this, indicating that employment outside of school can negatively affect academic performance and increase dropout risk [14, 120]. Although its prevalence has decreased, hard work is still a risk factor for low grades and school dropout [94]. The data supported Hypothesis 4, showing that employed children—regardless of sector—had a notably increased risk of dropping out of school.

Literature indicates that children from large families risk educational disruption due to poverty, sibling competition, and caregiving burdens [14, 95, 121]. This study provided numerical evidence that households with six or more members pose a heightened risk of school dropout, reinforcing this relationship. Children in crowded families are more likely to drop out of school. The findings partially confirmed Hypothesis 5: caregiving responsibilities significantly increased dropout risk, while regular involvement in household chores showed only a modest protective association. According to the literature, household size is a significant predictor of school dropout [95]. Studies also indicate that larger family sizes negatively impact children’s education [57]. According to the literature, having an extra sibling increases the risk of children dropping out of school, which is said to be more pronounced in crowded families [14]. Consequently, children in large families often face higher poverty risks, which contribute to early labour and school dropout [121]. High poverty and housing problems are thought to be associated with children working at an early age and may contribute to school dropout. Hypothesis 1 was confirmed, as the analysis showed that older children, girls, and those living in households with six or more members had significantly higher school dropout rates.

Although this study provides critical insights into the Turkish context, it is important to acknowledge cross-country differences. For example, while early marriage, child labour, and conservative family values are prominent dropout drivers in Türkiye, in sub-Saharan Africa or South Asia, factors such as extreme poverty, long distances to school, and lack of infrastructure play a larger role [122]. In contrast, in high-income countries like the United States or the United Kingdom, dropout risks are often more closely tied to academic failure, learning disabilities, mental health issues, and school disengagement rather than economic necessity [123–125]. Furthermore, Türkiye’s unique socio-cultural dynamics, including conservative attitudes toward girls’ education and coeducation, set it apart from many middle-income countries where economic factors might dominate [15, 16]. Recognizing these contextual differences is essential to ensure that educational policies and interventions are culturally sensitive and effectively address the specific barriers faced by students in Türkiye.

In Türkiye, these policies should be adapted to each region’s specific economic, cultural, and educational needs. In the eastern and southeastern regions, the focus should be on promoting girls’ education and reducing child labour; in major cities, supporting migrant children and providing academic assistance should take priority. In rural areas, solutions should address transportation barriers and caregiving burdens. Ministry, local governments, and non-governmental organizations (NGOs) should collaborate to establish monitoring and evaluation systems. In short, rather than a one-size-fits-all approach, regionally tailored solutions that actively involve local stakeholders are essential.

In summary, this study aligns with prior research by confirming several known risk factors while expanding the discussion by offering new insights specific to Türkiye’s unique socioeconomic and cultural context. It makes an important contribution to the literature by providing new evidence on the combined effects of gender, parental education and employment, household size, child labour, and caregiving burdens on school dropout among children aged 5–17 in Türkiye, using a large national sample and multivariable logistic regression analysis. Unlike earlier studies focusing on narrow age groups or limited factors, this research integrates individual, household, and sociocultural dynamics, addressing gaps in theory and empirical evidence. Methodologically, it strengthens analytical rigour by accounting for Türkiye’s complex survey design and sampling weights, thereby enhancing the generalizability of the findings. Together, these contributions offer valuable insights for future research and policy efforts to reduce dropout risks in Türkiye and other middle-income countries facing similar socioeconomic and cultural challenges.

Limitations

This study has several limitations that should be acknowledged. First, the analysis relied solely on secondary data from the 2019 Child Labour Survey, limiting the scope of variables to those available in the dataset. Since the dependent and independent variables are based solely on the questions asked by TurkStat, other potential factors associated with school dropout — such as mental health status, peer relationships, school environment factors, and detailed family dynamics — could not be analyzed. The exclusion of these variables could lead to omitted variable bias, where the observed effects of our included variables might partially capture the influence of these unmeasured factors. Second, the cross-sectional nature of the data means causality cannot be inferred; the identified associations reflect correlations, not necessarily causal relationships. For example, while child labour is associated with a higher probability of dropout, it is also possible that children who have already decided to drop out for other reasons are more likely to enter the workforce. Third, while the sample is nationally representative, it may not capture local or regional differences in school dropout patterns, particularly in marginalized or underrepresented communities. The exclusion of children not living with both parents (n = 2,698) may limit the generalizability of our findings concerning parental influence. The dynamics of school dropout in single-parent or orphan households could differ significantly, and this specific population warrants separate investigation. Finally, the study’s findings are based on data from 2019, which may not fully reflect the impacts of more recent events, such as the COVID-19 pandemic, on children’s education and dropout risks. Additionally, the COVID-19 pandemic may have exacerbated school dropout risks by increasing household economic pressures, limiting access to remote learning, and amplifying mental health challenges among children. Future research should incorporate longitudinal data, qualitative perspectives, and additional variables to deepen understanding and inform targeted interventions.

Conclusion

This study provides nationally representative insights into the multifaceted risk factors associated with school dropout among children aged 5–17 in Türkiye. It found that gender, age, parental education and employment status, household size, child labour participation, and caregiving burdens are significantly associated with school retention. Based on the observed associations, girls, older adolescents, children from large families, and those involved in employment or caregiving were more likely to be out of school. In contrast, parental employment and maternal education were inversely associated with school dropout risk. Policy efforts should consider the observed risk patterns alongside broader access strategies to address these inequalities. Interventions should target and remove the identified social and structural barriers that at-risk groups face. Recommendations include promoting girls’ education, providing families with financial and employment support, enforcing stronger regulations against child labour, and increasing school-based social support for children with caregiving responsibilities. Given Türkiye’s unique socio-cultural dynamics, proposed solutions must be culturally sensitive and adapted to local realities rather than imported wholesale from other global contexts. Although this study focuses on Türkiye, the findings may be relevant to other middle-income countries with similar socio-cultural and economic challenges, offering insights for comparative policy development [126].

Policy implications and future directions

The findings of this study point to several multidimensional policy recommendations aligned with the Sustainable Development Goals (SDGs) to reduce school dropout. Under SDG 4 (Quality Education), initiatives should focus on increasing girls’ access to education through scholarships, family awareness programs, and flexible learning models. In line with SDG 1 (No Poverty) and SDG 8 (Decent Work and Economic Growth), conditional cash transfers for low-income households and employment support for parents should be prioritized. To address SDG 5 (Gender Equality), efforts to combat early marriage and implement school-based gender equality awareness programs are essential. Additionally, under SDG 10 (Reduced Inequalities), it is crucial to reduce rural-urban education gaps and develop mobile or travelling educational support systems for migrant and seasonal worker children. These recommendations emphasize a holistic approach that targets the social, economic, and cultural barriers preventing children from staying in school.

Authors’ contributions

ÖA conceived and led the design and development of the study proposal. ÖA and EB supervised data collection, led the data analysis and drafting the manuscript. MÇ and EB made substantial contributions to the conceptualization and design of the study, data interpretations and writing the manuscript. All authors read and approved the final version of the manuscript.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

The data underlying this study is subject to third-party restrictions by the Turkish Statistical Institute. Data are available from the Turkish Statistical Institute (bilgi@tuik.gov.tr) for researchers who meet the criteria for access to confidential data. The authors of the study did not receive any special privileges in accessing the data.

Declarations

Ethics approval and consent to participate

The data were obtained through the joint teamwork of both the Turkish Statistical Institute (TurkStat) and the European Union Statistical Office (SOEU). Data were obtained this data from TurkStat in return for a contract without needing an ethics committee document and used it in this study.

TurkStat is an institution that compiles, evaluates, and presents statistical information to decision-makers to prepare development plans and programs, make economic decisions, and address all other issues needed. TurkStat carries out internationally comparable statistical production activities according to the standards of organizations such as the European Union Statistical Office, the United Nations, OECD, ILO, etc. TurkStat collects data within the scope of the Official Statistics Program. The Official Statistics Program is prepared for five-year periods based on the Turkish Statistics Law No. 5429 to determine the basic principles and standards regarding the production and publication of official statistics and to ensure the production of up-to-date, reliable, timely, transparent and impartial data in areas of need at national and international levels. TurkStat also conducts the Child Labour Survey within the scope of the Official Statistics Program put into effect by law. Since the Child Labour Survey is conducted within the scope of legal responsibility by the state, ethical approval is not required.

For this study, secondary data were employed. Official approval was received from the Turkish Statistical Institute to use the microdata set from the Child Labour Survey. The Child Labour Survey provides many indicators in the field of health, including the utilization of health services by individuals aged 15 and over, the degree of difficulty they experience in performing their daily activities, and their smoking and alcohol use habits. The Turkish Statistical Institute also received a “Letter of Undertaking” authorizing it to use the study’s data.

The letter of undertaking for the use of micro data without restrictions in dissemination:

Article 1- This letter of undertaking determines the rules, principles and obligations of the use of micro data, which are safe to disclose apart from the Presidency.

Article 2-This letter of undertaking regulates the use of micro data sets of Child Labour Survey in 2019, within the framework of the Directive on Access and Use of Micro Data in line with the purpose specified in Article 1.

Article 3- The following provisions apply for the use of micro data:

1. Findings obtained by the researcher as a result of incorrect calculation only bind the researcher.

2. The researcher refers to the micro data of the Institution that he uses while disclosing the results obtained from the study.

3. The researcher is obliged to send a copy of the published report, article, publication etc. to the Institution Library within three months at the latest. Subsequent micro data usage requests of the researcher who is found not to fulfill this obligation are not covered.

4. The researcher cannot reproduce, give to third parties, sell or transfer the micro data set he obtained.

Article 4-The researcher, by taking into account the principles of confidentiality defined in 13. and 14. articles of Turkish Statistical Institution numbered 5429 and Regulation on Procedures and Principles Regarding Data Confidentiality and Confidential Data Security in Official Statistics, is deemed to guarantee hereby that he shall not disclose the information, table, etc. violating this principle and shall only use micro data for statistical purposes.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.