Enhancing Data Literacy Competencies through the Application of Artificial Intelligence for Primary School Students

Abstract

This study adopted a one-group pretest–posttest quasi-experimental design             (no control group) to develop and evaluate artificial-intelligence (AI)–integrated learning activities that enhance data literacy competencies among lower-secondary students. The objectives were to: (1) design AI-enabled learning activities that foster data-driven thinking; (2) compare students’ data-literacy performance before and after the intervention; and (3) examine learner satisfaction with the AI-based activities. The participants were 30 students from Ban Nong Ta Khong School, under Nakhon Ratchasima Primary Educational Service Area Office 1. A multi-stage sampling procedure was employed: the school was first selected by purposive sampling based on ICT infrastructure readiness; a classroom was then chosen via cluster random sampling; and all 30 students in that class were included. Research instruments were: (1) a 20-item data-literacy competency test scored on a five-point scale (max = 5) with item difficulty 0.25–0.78, discrimination 0.33–0.68, and KR-20 = .81; (2) a 20-item learner-satisfaction questionnaire with Cronbach’s  = .79; and (3) a semi-structured interview protocol with IOC = 0.67–1.00. Data were analyzed using descriptive statistics (mean, standard deviation) and inferential statistics (dependent/paired-samples t-test and Cohen’s d for effect size). Findings showed that: (1) three AI-enhanced learning activities were developed “Future Data Detectives,” “AI Fake-News Detector,” and “Headline-Only News Analyst” all aligned with a five-stage data-thinking process; expert review rated their appropriateness at a high level (= 3.90, S.D. = 0.57). (2) Posttest scores (= 3.88, S.D. = 0.31) were significantly higher than pretest scores (= 2.15, S.D. = 0.35) at the .01 level (t = 16.905, p < .01), indicating a very large effect (reported Cohen’s d = 5.33), evidencing strong practical significance of the intervention. (3) Learners’ satisfaction with the AI-enhanced learning activities was at a high level (= 4.29, S.D. = 0.60).