This study aims to identify underlying subgroups regarding the activities with ICT resources for their learning purposes among 15-year-old students in Finland and to examine the relationships between the latent classes and students’ background (gender, socio-economic status [ESCS], and immigrant status) and their ICT efficacy.  

In Finnish schools, students’ activities with ICT for their learning purposes vary because of teachers’ pedagogical autonomy and their ICT use (Oinas et al., 2023). In addition, students’ background is related to their ICT use (Oinas et al., 2023). In international and European countries, the relationships between ICT use and educational outcomes have gathered attention (OECD, 2023), while studies have provided conflicting results (Oinas et al., 2013). For example, Li and Peterson (2022) indicate that ICT use enhances engagement in ICT and is related to educational outcomes, while Oinas et al. (2023) state ICT use seems slightly negatively correlated with learning outcomes. Some researchers claim that the relationship is not linear (Li & Zhu, 2023). Kim and Kim (2023) conducted a latent class analysis to identify how students with specific ICT use could be situated in their learning process. They found that the class with the highest ICT use efficacy and low use of ICT at school exhibited the highest reading literacy. Conversely, the class with the highest ICT familiarity had the lowest reading literacy. However, previous studies have still focused on the quantity of ICT use (how often to use). This suggests that there is a lack of discussion on the types of activities with ICT resources (what to do with ICT resources) for learning purposes. 

Moreover, the relationships between the activities with ICT resources for learning purposes and students’ background characteristics are still unclear. Alvarez-Garcia et al. (2024) identified seven clusters with PISA2022 data. Their findings indicated that the cluster of students with high ESCS, moderate academic ICT use, and strong academic support achieved the highest academic performance. Kim and Kim (2023) have pointed out the students’ profile of ICT use with their gender and ESCS. They found that the latent classes with high ICT familiarity tended to have a higher proportion of male students, while ESCS of the class with the lowest ICT familiarity was significantly lower than that of other classes. Furthermore, a report by Oinas et al. (2023) found that advanced digital technology for learning purpose was targeted at students with an immigrant background to receive support more than the majority population. In addition, previous studies have suggested a link between ICT self-efficacy and academic outcomes (Hu & Yu, 2021; Kim & Kim, 2023), however, researchers have captured ICT self-efficacy as a comprehensive concept. Competence in ICT use is often discussed in terms of foundational use (FU) and advanced use (AU) (International Association for the Evaluation of Educational Achievement [IEA], 2023). Foundational use is related skills to investigate, create, and communicate with ICT, while advanced use is related computational thinking to recognise aspects of real-world problems including programming (IEA, 2023). Thus, ICT efficacy related to FU and AU should be explored separately.  

To fill the gaps mentioned above, we address the following research questions:  

1. What kinds of latent classes emerge regarding students’ activities with ICT for learning purposes? 

2. What are the relationships between the identified latent classes and the variables including gender, ESCS, immigrant status, and efficacy in FU and AU? 

We address our research questions by analysing the Finnish PISA2022 data (N = 10239) and employing latent class analysis (LCA) to identify distinct subgroups within the participants. Since the PISA questionnaire consists of variety of questions, we selected the questions about ICT use for enquiry-based learning (IC174Q01-10), including creating multimedia presentation (IC174Q01), writing texts (IC174Q02), online search about a problem (IC174Q03), data collection (IC174Q04), data analysis (IC174Q05), reporting and sharing results (IC174Q06), planning and manging projects (IC174Q07), tracking process of the work (IC174Q08), collaboration to create digital contents (IC174Q09), and playing learning games (IC174Q010). In addition, we selected gender, social economic background (ESCS), immigrant status, and ICT efficacy (IC183Q01-16) as covariates. Concerning ICT efficacy, we conducted an exploratory factor analysis (EFA) because the items are theoretically divided into different dimensions (IC183Q1–13 and IC183Q14–16). EFA showed two factors, namely FU and AU. We used the factor scores for FU and AU. Data were analysed using Mplus version 8.4. The LCA was conducted to determine the optimal number of latent classes. Some models were tested, ranging from one to six classes. Model fit was evaluated using the Bayesian Information Criterion (BIC), the Akaike Information Criterion (AIC), entropy, and the Lo-Mendell-Rubin adjusted likelihood ratio test (LMR-LRT). In addition, the classification probabilities for the most likely latent class membership were assessed. The criterion of the probabilities should be more than .80 between all identical latent class memberships and latent classes. Finally, the model with the lowest BIC and AIC, the entropy (acceptable values above .70; Jung & Wickrama, 2008), and a significant LMR-LRT was selected as the best-fitting model. After identifying the latent classes, we extended the model by inputting the covariates that may predict the students’ class membership. The relationships between the classes and the covariates were analysed with multinomial logistic regression. We employed the R3STEP method, an indirect auxiliary-variables approach, to calculate the regression and odds ratio coefficients (Asparouhov & Muthén, 2013).

The model identified five distinct classes: Class 1 (C1; n = 1036, 12.2%) characterised by very high levels of all kinds of ICT use; Class 2 (C2; n = 3074, 36.2%) characterised by high levels of all kinds of ICT use; Class 3 (C3; n = 1149, 13.5%) characterised by high levels of ICT use without the items including data collection (IC174Q04), data analysis (IC174Q05), reporting and sharing results (IC174Q06), and playing learning games (IC174Q10); Class 4 (C4; n = 2059, 29.6%) characterised by moderate levels of all kinds of ICT use; and Class 5 (C5; n = 719, 12.2%) characterised by low levels of all kinds of ICT use. Interestingly, the relationships between the classes and gender were found to be non-linear. The girls were more likely to be in C3, while boys were more likely to be in C1, C2, and C5 (PostHoc result – boys: C1, C2, C5 > C4 > C3; p < .05). Immigrant status was significantly linearly associated with the classes. The higher immigrant status, the higher ICT use (PostHoc result: C1 > C2, C3 > C4 > C5; p < .05). ESCS also showed a significant linear relationship, showing that students with higher ESCS more likely to use ICT often (C1 > C2 > C3 > C4, C5; p < .05). Additionally, the significant relationships between the classes and FU and AU efficacy were found. For the relationship between the classes and FU, it was almost linear (C3 > C1, C2 > C4, C5; p < .05). However, the relationship with AU was found to be non-linear (C1 > C2 > C5 > C4 > C3; p < .05). This suggests that the class profile of ICT use varies depending on gender, immigrant status, ESCS, and FU and AU efficacy.

Alvarez-Garcia, M., Arenas-Parra, M., & Ibar-Alonso, R. (2024). Uncovering student profiles. An explainable cluster analysis approach to PISA 2022. Computers & Education, 223, 105166. https://doi.org/https://doi.org/10.1016/j.compedu.2024.105166 European Commission. (2019). 2nd survey of schools – ICT in education – Objective 1 – Benchmark progress in ICT in schools, final report. Directorate-General for Communications Networks, Content and Technology. Publications Office. https://data.europa.eu/doi/10.2759/23401 Hu, J., & Yu, R. (2021). The effects of ICT-based social media on adolescents’ digital reading performance: A longitudinal study of PISA 2009, PISA 2012, PISA 2015 and PISA 2018. Computers & Education, 175, 104342. https://doi.org/10.1016/j.compedu.2021.104342 IEA. (2023). ICILS 2023 Framework. https://ilsa-gateway.org/studies/frameworks/1661#Assessment%20or%20survey%20framework Kim, M., & Kim, H. (2023). Profiles of students’ ICT use in high-performing countries in PISA 2018. Computers in the Schools, 40(3), 262-281. https://doi.org/10.1080/07380569.2023.2180338 Li, S. C., & Petersen, K. B. (2022). Does ICT matter? Unfolding the complex multilevel structural relationship between technology use and academic achievements in PISA 2015. Educational Technology & Society, 25(4), 43-55. https://doi.org/10.30191/ETS.202210_25(4).0004 Li, S., & Zhu, J. (2023). Cognitive-motivational engagement in ICT mediates the effect of ICT use on academic achievements: Evidence from 52 countries. Computers & Education, 204, 104871. https://doi.org/10.1016/j.compedu.2023.104871 OECD. (2023). PISA 2022 ICT Framework. In OECD, PISA 2022 Assessment and Analytical Framework. Paris: OECD Publishing. https://doi.org/10.1787/9bd299c1-en Oinas, S., Vainikainen, M.-P., Asikainen, M., Gustavson, N., Halinen, J., Hienonen, N., Kiili, C., Kilpi, N., Koivuhovi, S., Kortesoja, L., Kupiainen, R., Lintuvuori, M., Mergianian, C., Merikanto, I., Mäkihonko, M., Nazeri, F., Nyman, L., Polso, K.-M., Schöning, O., Svedholm-Häkkinen, A. M., . . . Hotulainen, R. (2023). The impact of digitalisation on learning situations, learning and learning outcomes in lower secondary schools: Initial results and recommendations of a national research project. Tampere University and University of Helsinki.