A Quasi-Experimental Comparison of Paper-Based and Computer-Based Vocabulary Layering Techniques on EFL University Students’ Vocabulary Development

Abstract This study has aimed to compare paper-based and computer-based vocabulary layering, and to examine vocabulary outcomes and learner attitudes under controlled instructional conditions. Vocabulary learning remains a major challenge for university EFL learners, particularly in developing collocational knowledge and achieving long-term retention. Although corpus-informed Data-Driven Learning (DDL) has been associated with improved vocabulary acquisition, previous studies have primarily compared DDL with traditional instruction rather than examining paper-based and computer-based implementations of the same approach. This study compared these two delivery modes and investigated their effects on vocabulary learning and learner attitudes. A quasi-experimental pre-test–post-test design was conducted with 90 EFL students (aged 19–24) from three intact classes in Northern Cyprus. Participants were assigned to a control group, a paper-based DDL group, and a computer-based DDL group (30 students each). Baseline equivalence was confirmed (p = .590). Over 14 weeks, both experimental groups received identical vocabulary instruction, differing only in the mode of delivery. A 40-item vocabulary test and a 12-item attitude questionnaire were administered, and the data were analysed using ANOVA, paired and independent t-tests, effect sizes, and reliability analysis. Both experimental groups demonstrated significant vocabulary gains from pre-test to post-test. The paper-based group improved from M = 38.93 to 58.60 (p = .0001, d = -1.42), while the computer-based group increased from M = 38.86 to 71.66 (p = .0001, d = -2.55). Both groups significantly outperformed the control group, with the computer-based group demonstrating the larger effect. The computer-based group also achieved significantly higher scores than the paper-based group (p = .0005). Both DDL groups reported positive attitudes toward corpus-based DDL, with no significant difference between them (p = .93). These findings indicate that vocabulary layering effectively enhances lexical development in both instructional modes. At the same time, computer-based DDL produces greater vocabulary gains. The results support integrating technology-enhanced, corpus-informed vocabulary instruction in higher education to promote more effective, evidence-based language learning.