Games are powerful educational tools that require teachers to possess knowledge of game mechanics, content, and pedagogy to integrate them effectively into learning environments (Hsu et al., 2015). To address this, the Technological Pedagogical Content Knowledge (TPACK) framework, introduced by Mishra and Koehler (2006), highlights the interplay of technological, pedagogical, and content knowledge as critical predictors of effective teaching. Building on this, Hsu et al. (2013) expanded the framework by introducing TPACK for games (TPACK-G), which focuses on the unique demands of teaching with games and offers a tailored approach to understanding teacher competencies in game-based learning environments. While TPACK has received extensive attention in educational research, recent efforts have focused on identifying and exploring TPACK profiles and technology integration from intra-individual and inter-individual perspectives. For instance, profiling teachers based on their TPACK has resulted in low, medium, and high profiles (Yang et al., 2023), low and high profiles (Tondeur et al., 2017) among pre-service teachers, and low, medium, high, and mixed profiles (Howard et al., 2021) among secondary school teachers. Similarly, Li et al. (2024) identified six more granular profiles, including low-all, modest-all, low content and pedagogical content knowledge + modest-others, moderate-all, high pedagogical and pedagogical content knowledge + moderate-others, and high-all, among pre-service foreign language teachers. However, little is known about TPACK-G profiles among teachers, as existing research has not specifically examined teacher profiles focused on TPACK-G. Thus, this study aimed to address this gap by identifying latent subgroups of TPACK-G among teachers using Factor Mixture Modeling (FMM; Lubke & Muthén, 2005) to determine whether distinct “classes” existed based on their TPACK-G dimensions. This approach recognizes that teachers with varying levels of game knowledge may conceptualize TPACK-G dimensions differently, whereas traditional statistical methods often assume participant homogeneity, potentially overlooking meaningful variations in how subgroups develop and apply game-based teaching competencies. To do this, data were collected from 347 pre-service teachers (74.3% female, 25.7% male) using the TPACK-G measure.

The analysis followed three steps:
(1) confirmatory factor analysis (CFA) to validate factor structures,
(2) latent profile analysis (LPA) to identify distinct profiles, and
(3) FMM to integrate categorical and continuous latent variables.

Results revealed the best fit for a three-class model with a four-factor structure, identifying three distinct TPACK-G profiles: High Knowledge: Highest scores across all four dimensions; Medium Knowledge: Moderate scores overall, except for low scores in game pedagogical knowledge (GPK); Low Knowledge: Low scores across most dimensions, with moderate scores in GPK. These findings suggest significant variability in pre-service teachers' TPACK-G competencies, with GPK emerging as a key determinant of profile membership. This aligns with prior research (Kuo & Kuo, 2024), which emphasizes the critical role of pedagogical knowledge in integrating games effectively into teaching. These insights enhance understanding of TPACK-G heterogeneity and have implications for designing targeted interventions to strengthen game-based teaching competencies in teacher education programs.

Keywords: Game knowledge, game pedagogical knowledge, teachers, latent profiles, factor mixture modeling.