Uneven Technology Landscapes: Digital Divide Effects on PISA Achievement, 2000-2018



Author Information

Joseph Lavallee, Ming Chuan University, Taiwan
Pin-Ju Chen, Ming Chuan University, Taiwan
Bruno Di Giusto, Ming Chuan University, Taiwan

Abstract

Understanding technology-education relationships becomes crucial as education systems approach the generative AI revolution. This study examines technology's relationship with student achievement using PISA reading, math and science scores from 2000-2018 (excluding 2022 to avoid COVID-19 distortions) across multiple countries. We test two digital divide levels: Level 1 (access to computers, smartphones and the internet) and Level 2 (skills/usage patterns). Hypothesis 1 examines whether "connected homes" (computer+internet) enhanced learning, while Hypotheses 2-3 explore smartphone effects across and within countries respectively. We estimate country fixed-effects panel models with year dummies and precision weights to isolate technology's impact from stable country characteristics. For H1, we analyze the pre-smartphone era (2000-2012) with year controls to capture global trends in connected home benefits. For H2, we test whether smartphone effects differ between high and low-income countries using GNI per capita and interaction terms. For H3, we examine smartphone vulnerability across ESCS quartiles (PISA’s measure of socioeconomic status) within countries, testing whether SES moderates technology’s impact. Three findings emerge. First, we find no clear evidence that connected homes boosted achievement—suggesting Level 1 access alone does not drive performance. Second, smartphone access shows strongly negative associations in high-income countries, particularly for science (-204 points, p<0.01) and mathematics (-143 points, p<0.1), suggesting greater vulnerability in these countries. Third, within countries, both the lowest (Q1) and upper-middle (Q3) ESCS quartiles show stronger negative associations than highest-quartile (Q4) peers in mathematics (-14.7 points each), indicating Level 2 digital divide effects where affluent students better protect themselves against digital distractions.


Paper Information

Conference: ACE2025
Stream: Design

Published: Tuesday August 26, 2025

The full paper is not available for this title


Virtual Presentation


Comments & Feedback

Place a comment using your LinkedIn profile

Comments

Share on activity feed

Powered by WP LinkPress

Share this Research

Posted by The International Academic Forum