Abstract
Mobile applications have become a central tool for learning English as a foreign language. Two designs now dominate the market, gamification and artificial-intelligence-based adaptive learning, yet they are seldom compared directly, especially in Kazakhstan and among the future foreign-language teachers who will shape its national language education. This study compared a gamified application, Duolingo, with an artificial-intelligence-based adaptive application developed in Kazakhstan, Oki/Aleem, over a four-week intervention. A two-group quasi-experimental design was used. In total, 120 undergraduate students in a foreign-language teacher-training programme worked with one of the two applications, and 115 were matched across a pre-test and a post-test. Achievement was measured with parallel twenty-item forms administered before and after the intervention. A weekly form tracked how the applications were used, and three open-ended questions were coded thematically. Scores rose markedly in both groups over the month. Once pre-test scores were entered as a covariate, the analysis of covariance showed no advantage for either application. Weekly practice was sustained in both groups and increased toward the end of the intervention. The open-ended responses revealed a clear contrast. Duolingo was valued for its game-like motivation, whereas Oki/Aleem was valued for its adaptivity, feedback and local relevance. The two applications therefore produced similar learning gains through different mechanisms, which reframes the choice between them as a question of purpose rather than of effectiveness. The findings support building direct, comparative experience of gamified and adaptive applications into foreign-language teacher preparation in Kazakhstan