Research on the Value added Evaluation of Digital Competence of University Teachers from the Perspective of Education DigitalizationMotion

Minjia Yu, Xianya Wu, Tong Yang, Ming ZhouJournal of Education Humanities and Social Sciences 2026 · DOI

TL;DR

This paper introduces a novel framework for evaluating the "value-added" digital competence of university teachers, crucial for modernizing higher education. It combines expert opinions (Fuzzy Delphi) with data-driven methods (Entropy Weight) to create a robust weighting system, then models the dynamic evolution of digital competence across universities, considering their classification and regional location.

What Problem Does This Paper Address?

The rapid digital transformation of education, fueled by AI and big data, necessitates a corresponding evolution in university teachers' digital competence. However, existing evaluation methods often lack a comprehensive, dynamic, and nuanced approach. This paper addresses the need for a robust system that can:

  • Quantify digital competence: Move beyond qualitative assessments to a measurable evaluation.
  • Integrate diverse perspectives: Combine subjective expert knowledge with objective data.
  • Account for university context: Recognize that digital competence development might differ based on university type (e.g., "Double First Class" vs. local) and geographical region (eastern, central, western).
  • Track progress over time: Understand the "value-added" aspect, not just a static snapshot.

Methods

The paper proposes a multi-stage methodology:

  1. Indicator Refinement: The broad "Teacher Digital Literacy" standard is broken down into four primary indicators (Digital Basic Capability, Digital Teaching Ability, Digital Analysis Ability, Digital Innovation Capability) and further into 12 specific secondary indicators.
  2. Weight Determination (Hybrid Approach):
    • Fuzzy Delphi Method: Expert opinions on indicator importance are collected using fuzzy linguistic variables (e.g., "very high," "moderate") which are then converted into triangular fuzzy numbers. This method aggregates subjective judgments while accounting for uncertainty.
  3. Analytic Hierarchy Process (AHP): A pairwise comparison matrix is constructed based on expert judgments to determine the relative importance of indicators. Consistency testing (CR < 0.1) ensures the rationality of these subjective weights.
  4. Entropy Weight Method: This objective method is used to incorporate data variability and adjust weights based on actual indicator performance.
  5. Comprehensive Weighting: A weighted average of subjective (AHP) and objective (Entropy) weights is calculated using a parameter $\alpha$ (set to 0.7), balancing expert experience with data-driven insights.
  6. Value-Added Evaluation Model:
    • Static Weighting: A "static competence score" (SS) is calculated for universities using the derived comprehensive weights and normalized teacher scores.
  7. Classification and Regional Influence Model: The model incorporates university classification (e.g., Double First Class) and regional location (eastern, central, western) as impact coefficients, adjusting the static score.
  8. Dynamic Evolution Model: A differential equation model is used to describe the "value-added speed" of digital competence over time, considering endogenous growth, external impacts, and policy factors. This allows for tracking the cumulative growth.

Datasets

Not explicitly detailed as a distinct "dataset" in the paper. The methodology relies on:

  • Expert Judgments: Collected from K experts (15-20) in the field for the Fuzzy Delphi and AHP methods.
  • Teacher Scores: Normalized scores ($x_{ij}$) of teachers on the 12 secondary indicators, collected from $m$ teachers.
  • University Classification and Regional Location Data: Used to define matrices ($C$ and $R$) representing university attributes.

Key Insights and Ideas

  • Bridging Subjectivity and Objectivity: The core innovation lies in the hybrid weighting system that combines expert intuition with data variability, leading to more reliable and practical evaluations.
  • Context Matters: Digital competence isn't uniform; university classification and regional location significantly influence its development and value-added potential.
  • Dynamic Evaluation is Crucial: A static snapshot is insufficient. Understanding the growth trajectory of digital competence over time provides a more accurate picture of progress and effectiveness.
  • Policy Impact: The model can quantify the marginal effect of policies (like digital training or subsidies) on digital competence, offering actionable insights for educational administrators.

Key Results

  • Empirical Evidence: The model shows that "Double First Class" universities ($Y_1=0.82$) and universities in the eastern region ($n_1=0.75$) have experienced significantly increased teacher digital competence.
  • Policy Impact: Policies like digital training center coverage and intelligent teaching subsidies demonstrated a marginal effect of 48% on western universities, highlighting their potential to boost digital competence in less developed regions.
  • Visualizations: The paper presents dynamic weight heatmaps, 3D variation graphs of AHP weights, and time-series plots of digital competence evolution, visually supporting the model's capabilities.

Limitations and Future Work

  • Data Granularity: The paper mentions using normalized scores for teachers, but the specific source and scale of this data are not detailed.
  • Constant Coefficients: The dynamic evolution model assumes constant coefficients for growth rate, external impacts, and policy factors. In reality, these might fluctuate.
  • Generalizability: While the model is applied to university teachers, its adaptability to other educational levels or professional groups is not explored.

Future work could involve:

  • Real-world Data Application: Applying the model to a larger, diverse dataset of universities and teachers to further validate its findings.
  • Dynamic Coefficient Estimation: Investigating methods to estimate time-varying coefficients in the dynamic evolution model.
  • Intervention Strategies: Developing more specific, data-driven intervention strategies based on the model's insights into promoting factors.

Summary updated

Abstract

In the context of AI, Internet and big data technology accelerating the digital transformation of education, the digital competency of college teachers has become the core element to promote the modernization of higher education. Based on the industry standard of "Teacher Digital Literacy", the four primary indicators of digital basic ability and digital teaching ability are refined into 12 secondary indicators (such as hardware and network access ability, proficiency in online teaching platform operation, etc.). Using the fuzzy Delphi method to aggregate expert fuzzy judgments (five level language variables and triangular fuzzy number conversion), combined with the Analytic Hierarchy Process (AHP) to construct a fuzzy pairwise comparison matrix, and ensuring the rationality of weights through consistency testing (CR&lt;0.1), ultimately forming a comprehensive weight system that integrates subjective and objective factors ( =0.7), achieving scientific quantification of indicator importance. Constructing a "static weighting+dynamic evolution" model, calculating the static competence score SS of universities through the weight vector W. Empirical evidence shows that the teacher competence of Double First Class universities ( 1=0.82) and the eastern region ( 1=0.75) has significantly increased, and policies such as digital training center coverage and intelligent teaching subsidies have a marginal effect of 48% on western universities.
BibTeX
@article{yu2026research,
  author = {Minjia Yu and Xianya Wu and Tong Yang and Ming Zhou},
  title = {{Research on the Value added Evaluation of Digital Competence of University Teachers from the Perspective of Education DigitalizationMotion}},
  journal = {Journal of Education Humanities and Social Sciences},
  year = {2026},
  doi = {10.54097/exncpj79},
}
Figure