What are the determinants of rural-urban divide in teachers' digital teaching competence? Empirical evidence from a large sample

The paper addresses the growing concern that, despite expanding ICT infrastructures, a second-order digital divide (differences in digital competence and use) persists between rural and urban areas and contributes to educational inequalities. It situates digital teaching competence—teachers’ skills, knowledge, and attitudes to integrate ICT pedagogically—as pivotal for realizing SDG4 and for ensuring that infrastructure investments translate into improved educational outcomes. Prior work shows mixed evidence regarding teachers’ ICT attitude, skills, and data literacy as predictors of digital teaching competence, and most research has focused on students rather than teachers, with limited attention to rural teachers. The study poses three research questions: (1) What are the potential factors that influence teachers' digital teaching competence? (2) Does the divide in digital teaching competence exist between rural and urban teachers? (3) If it does, what is the relative importance of the determinants contributing to the competence divide?

The literature review and hypotheses development focus on five areas. ICT attitude: prior studies are mixed on whether teachers’ ICT attitudes predict digital teaching competence; some find positive effects on technology use and competence, others find non-significant relationships. Differences by school location suggest rural teachers may hold more conservative teaching conceptions and possibly lower ICT attitudes than urban peers. Hypotheses: H1: ICT attitude positively affects digital teaching competence; H2: rural teachers have lower ICT attitude than urban teachers. ICT skills: essential for integrating technology into teaching; skills include proficiency with hardware and software. Prior research shows urban teachers often have stronger ICT skills due to better environments; rural teachers may lag. Hypotheses: H3: ICT skills positively affect digital teaching competence; H4: rural teachers have lower ICT skills than urban teachers. Data literacy: defined as collecting, analyzing, evaluating, and applying data for instruction; generally lacking among teachers but theoretically developable and potentially predictive of digital teaching competence. Limited rural-focused studies suggest neglect of data literacy in rural contexts. Hypotheses: H5: data literacy positively affects digital teaching competence; H6: rural teachers have lower data literacy than urban teachers. Digital teaching competence: evidence on rural-urban gaps is mixed; some studies show rural teachers use digital tech less and have lower competence, while others find no difference in support. Hypothesis: H7: rural teachers show lower digital teaching competence than urban teachers. Digital divide in competence: prior work identifies factors widening competence divides (gender, motivation, location, skills, access, social class), but few studies decompose determinants for teachers. The review motivates analyzing how ICT attitude, ICT skills, and data literacy explain rural-urban divides in teachers’ digital teaching competence.

Design and instruments: A cross-sectional survey was administered via the Wenjuanxing online platform to K–12 teachers across kindergartens, primary, secondary, high schools, and vocational schools in Zhejiang province, China. The instrument comprised demographics (gender, years of teaching experience, geographic location: rural vs urban, and school level) and four validated subscales (from Lin et al., 2022): ICT attitude (3 items), ICT skills (4 items), data literacy (4 items), and digital teaching competence (5 items). Items were rated on 5-point Likert scales (1=strongly disagree to 5=strongly agree). Reliability and validity: Cronbach’s alpha for each subscale exceeded 0.80; AVE values exceeded 0.50, indicating acceptable reliability and convergent validity. Principal component analysis (KMO=0.942; Bartlett’s test p<0.05) extracted four constructs explaining 81.781% of variance; factor loadings exceeded 0.5 on intended constructs and were below 0.5 on others, supporting discriminant validity. Confirmatory factor analysis (AMOS 24) showed acceptable fit (e.g., RMSEA=0.077; NFI=0.963; TLI=0.955; CFI=0.963; IFI=0.963). Sample and data collection: 11,784 valid responses were collected; 43.40% rural (n=5114) and 56.60% urban (n=6670); 79.63% female; teaching experience varied from 1 to 36+ years; school levels spanned kindergarten to vocational. Analytic strategy: Structural equation modeling (SEM, AMOS 24, MLE) tested effects of ICT attitude, ICT skills, and data literacy on digital teaching competence (H1, H3, H5). ANOVA (SPSS) tested rural-urban differences in ICT attitude, ICT skills, data literacy, and digital teaching competence (H2, H4, H6, H7). To quantify determinants of the rural-urban gap in digital teaching competence, Blinder–Oaxaca decomposition (Stata 15, OLS with 500 bootstrap replications) partitioned the mean difference into characteristics (explained) and association (unexplained) effects, with observed factors including ICT attitude, ICT skills, data literacy, gender, and years of teaching experience.

- SEM results: The model explained R2=60.3% of variance in digital teaching competence. All three predictors had significant positive effects: ICT attitude b=0.152, p<0.001; ICT skills b=0.378, p<0.001; data literacy b=0.374, p<0.001 (supporting H1, H3, H5). - Rural–urban differences (ANOVA): Urban teachers scored higher than rural teachers on all constructs (supporting H2, H4, H6, H7): • ICT attitude: Urban M=4.542 vs Rural M=4.509; F=8.158; p=0.004. • ICT skills: Urban M=3.803 vs Rural M=3.692; F=61.847; p<0.001. • Data literacy: Urban M=3.716 vs Rural M=3.601; F=50.082; p<0.001. • Digital teaching competence: Urban M=3.896 vs Rural M=3.779; F=65.528; p<0.001. - Blinder–Oaxaca decomposition of digital teaching competence gap: Mean difference Urban−Rural=0.116 (p<0.001). • Characteristics (explained) effect: 0.092 (79.3% of gap; p<0.001). • Association (unexplained) effect: 0.024 (20.7% of gap; p=0.011). Contribution of observed factors to characteristics effect: • Data literacy: 0.050 (43.10%; p<0.001) — largest contributor. • ICT skills: 0.031 (25.86%; p<0.001). • ICT attitude: 0.004 (3.45%; p=0.009). • Gender: 0.007 (6.03%; p=0.001). • Years of teaching experience: 0.001 (0.86%; p=0.011). Overall, data literacy and ICT skills are the most important determinants of the rural–urban divide in teachers’ digital teaching competence.

Findings show that teachers’ ICT attitude, ICT skills, and data literacy each positively associate with digital teaching competence, aligning with strands of prior literature and underscoring the importance of both affective and capability dimensions, including the often underexplored role of data literacy. The existence of rural–urban gaps across all four constructs verifies a second-order digital divide in the teacher workforce: despite generally high ICT attitudes in both groups, rural teachers report lower ICT skills and data literacy and, consequently, lower digital teaching competence. The decomposition clarifies that most of the competence gap (≈79%) is attributable to observable characteristics, particularly deficits in data literacy and ICT skills among rural teachers. These insights suggest that equalizing competence—especially via data literacy and ICT skills—would substantially narrow the rural–urban competence gap. As education systems become more data-driven, improving teachers’ capacity to collect, analyze, evaluate, and apply data is critical. Similarly, sustained, practice-focused ICT training and support are necessary to translate infrastructure into effective classroom integration, with special emphasis on rural contexts to avoid idle infrastructure and a reinforcing cycle of underuse and low competence.

This study demonstrates that ICT attitude, ICT skills, and data literacy are significant predictors of teachers’ digital teaching competence and that rural teachers lag behind urban teachers in all four areas. Using Blinder–Oaxaca decomposition, the study identifies data literacy and ICT skills as the primary determinants of the rural–urban competence divide. Contributions include large-scale empirical evidence on teachers (rather than students), explicit quantification of determinants of the divide, and highlighting data literacy as a pivotal yet under-addressed factor. Implications: Policymakers and school leaders should recognize the competence-based nature of the digital divide and design targeted interventions for rural teachers—policies and programs for sustained training, peer support, and rural–urban mutual aid—to build ICT skills and data literacy. Teacher professional development frameworks should systematically integrate both ICT skills and data literacy with pedagogical application. Teachers should develop proficiency with digital tools and platforms, troubleshoot classroom technologies, and use end-to-end data practices to inform instruction. Future research should expand beyond a single province, examine contexts with varying infrastructure levels, and explore downstream third-order divides (digital outcomes) once competence gaps are reduced.

- Generalizability: The sample is from a single province in China; findings may not generalize nationally or internationally. - Context dependence: Results are most applicable where ICT infrastructures are relatively developed; in low-connectivity contexts, addressing first-order access gaps may be more urgent. - Scope: While documenting a persistent second-order divide, the study does not analyze third-order (outcome) divides; future work should explore how closing competence gaps affects educational outcomes.