In conditions of extreme poverty, what factors drive educational attainment of village girls in western China?

• In Event: Education at the Margins: Education for Social Equity for Children from Marginalized Communities in China

Proposal

This proposal addresses the theme of the conference, Envisioning Education in a Digital Society, from the perspective of the digital divide and energy poverty. Our focus on China is appropriate for the East Asia Special Interest Group (SIG). In China the digital divide maps on the urban - rural divide and coastal – western regional inequality. 81% of the urban population, and 65% of the rural population are internet users (CNNIC. 2024). Rural internet connectivity, speed, and access to devices remain persistent challenges (Yang, Y., Jiang, M., Li, X., and Ya Chen 2024).

Social Context of the Study: Despite China’s “upper middle income” status (WB), in 2020, nationally 21.9% of children (UNICEF, 2020, 7) lived in families with an average income below the poverty line of $5.50 per day (Castaneda, Fujs, Gerszon, 2020). From comparison with independent studies (Wang, Zhao, Bai, Zhang, and Yu, 2020; 2016 China Rural Development Survey), we conclude that our sample falls into a lower extreme poverty rate of 1.7% nationally (National Bureau of Statistics, 2018). Our sample also is part of the 54.5% of China’s children who are affected by migration (UNICEF & NBS, 2021).
Purpose:
This study explores the educational attainment of the Guanlan Sisters (GS) from the village we call Anjinggou, in the remote Qinling Mountains of Shaanxi in the years between 2000-2020. Given the surprisingly high educational attainment that we found among the GS in previous research (Seeberg, Sun, Wang, 2024), the study turned to find out what might cause it. The study investigates the primary drivers of the educational attainment of the GS, seeking to identify family and contextual factors that may cause the high and variation in educational attainment.
The framework for this quantitative study is informed by Ecological Systems Theory (Bronfenbrenner, 1979) as well as Social Stratification Theory focused on structural resources.

The first research question asked what family background factors may predict the variation in educational attainment among the Guanlan Sisters. One sub-research question asks about contextual factors that may independently or in interaction with family factors produce varying educational attainment outcomes. These include time as a factor, described as birth year, and location of the school. Time is a factor that expresses macro-environmental changes during the 22 years of the study, and location of the school expresses the advantage of urban location of the school and its environment.

Method:
Data collection: We sourced observational data from four primary qualitative source channels from 2000-2022: annual reports provided by our partner NGO in China, annual letters from the total 139 Guanlan Sisters, seven rounds of field research of sound-recorded interviews with GS and informants, and scattershot telephone interviews with individual GS.
Preparing for statistical modeling, we converted qualitative into quantitative data, conducted extensive exploratory factor analysis using several statistical methods to select effective individual and interaction variables. We constructed a parsimonious model to answer the main research and its sub questions, explaining the variation in educational attainment.

Findings, detailed statistics will be provided at the conference:

Our parsimonious logistic regression model based on research question 1one did not identify any significant family involvement factors as causing variation in educational attainment. Turning to the sub-research question, we entered one exogenous factor, time defined as birthyear, into the model and found significant prediction for the educational attainment measured by the location of the JS completion of compulsory education. With the inclusion of the exogenous factor the expanded parsimonious logistic regression model was 84% accurate. It showed significant results for time and number of earners in determining the location of compulsory education school, answering the sub research question for compulsory education attainment as step 1.

In step 2, when comparing urban and rural school settings over time using marginal treatment effects, we found that the number of earners in the family, with birthyear as a condition, was a strong predictor of years of educational attainment, answering the sub research question.

Descriptive data, using 5-year birth year cohorts, demonstrated a clear trend toward urban education, with urban primary and junior secondary attainment rising from 0% in the 1986-1990 cohort to 82% for the 2011-2015 birth cohort. In sum, this significant shift from rural to urban settings for primary through junior secondary schooling aligns with the national trend. According to the National Bureau of Statistics, in 2010, 10% of rural children were attending urban primary and jr. secondary schools (NBS, 2013, p.5) and by 2020, 76% of rural children were doing so (unicef, NBS, UNFPA, 2021, p. 7). Even though our Guanlan Sisters count among the most marginalized rural population, the extremely poor, they were outperforming national statistics 82% to 76%.

Discussion & Conclusion:
This research provides a broad understanding of a significant shift in educational attainment from rural to urban settings over time and illustrates the mechanics of urbanization in the nation.

In areas experiencing extreme poverty, such as Anjinggou in western China, demographic data often show limited variation due to uniform economic conditions and hardship. Extreme poverty tends to level differences in family resources, as families across a community face similar challenges in meeting basic needs. This explains why research question 1 on family effect on educational attainment was inconclusive. Finding significant effects of a contextual factor highlights the importance of structural conditions.

In China the rapidity of changes in the past 40 years is well recognized. Our research shows that it also affected highly marginalized, poverty-stricken areas and the educational attainment of individual youth. It shows that structural factors, suggesting schooling provisions, infrastructure and job markets, were of greater importance than family factors in determining the school life trajectory of the Guanlan Sisters.

Given the unexpectedly high educational attainment of this population, we conclude that unknown factors beyond the structural and possibly family conditions played a significant role. While we anticipate identifying these factors in future research, we hypothesize that this group of young women may continue to make remarkable progress.

Authors

• Vilma Seeberg, Kent State University

• Ya Na, Kent State University

• Yixuan Chen, Guanlan Scholarship Foundation

• Kan Sun, Guanlan Scholarship Foundation

• Shuyue Luo