Digital Support of P. Nation’s Three-Stage Model in Teaching Construction Terminology in English: a Quasi-Experiment

Abstract

Purpose. To assess the effectiveness of integrating digital tools into the three-stage model of professional vocabulary acquisition (noticing – retrieval – creative use) for engineering students, and to determine whether the systematic matching of each tool to the cognitive task of a stage offers an advantage over traditional instruction. Professional foreign-language competence is critical for engineers, yet a large share of terminology remains in the passive vocabulary, and the fragmentary use of digital tools does not solve this problem. Materials and methods. A quasi-experiment with non-equivalent groups was conducted with the second-year students of construction specialties (71 persons: control group – 35, experimental group – 36) as part of an English course for professional purposes. The material was the “Building materials” module of an author-developed textbook: 50 key terms across 7 texts, 32 academic hours. The amount of study time and content were identical across groups; only the way of organizing activities differed. In the experimental group each stage of Nation’s model was supported by digital tools: noticing – corpus services (SKELL, Ludwig.guru) and artificial-intelligence platforms (Qwen, GigaChat); retrieval – spaced repetition and gamification (Quizlet, Wordwall, JeopardyLabs); creative use – the textbook’s productive tasks supported by Perplexity and Twee, as well as final situational games (digital quest in PowerPoint and role-playing game in Genially). Receptive vocabulary gain, productive use (lexical density and collocation accuracy), four-week retention, and motivation were measured. Data were analyzed with a normality check (Shapiro–Wilk test), Student’s t-test, and effect size; qualitative data were processed through thematic analysis; instrument reliability was confirmed by Cronbach’s alpha and Cohen’s kappa coefficient. Results. The experimental group showed a receptive vocabulary gain 1.4 times higher than the control group (15.3 vs 11.3 points; p < 0.01; effect size 0.84). On the four-week delayed test, knowledge loss was 14.3% versus 22.8% in the control group, indicating more durable retention. Productive use of terminology was significantly higher: lexical density by 47%, collocation accuracy - 69.5% versus 58.6%. Experimental-group students rated the course positively (3.8 out of 5.0), with the highest score given to the situational games of the creative-use stage (4.1) and the lowest to the autonomy scale (3.5). Conclusion. Integrating digital tools into Nation’s three-stage model improves professional vocabulary acquisition across all measured parameters – receptive knowledge, retention, and productive use. The key success factor is the systematic matching of each tool to the cognitive task of a specific stage, rather than the fragmentary use of technology. The model is reproducible, relies on tools available in Russia, and is transferable to other subject areas of professional foreign-language education.