Dropping out of studies is a large issue for the university, society, and often for the individual. While dropping out could be a positive issue for students who find a better-suited profession or study field, it admittedly also has negative effects, such as losing financial aid and time, and experiencing more unemployment and lower incomes than persisters (Davies & Elias, 2003; OECD, 2019). From an institutional perspective, universities’ funding is usually dependent on the number of graduates, so aside from the wasted resources, dropouts have a direct effect on universities’ funding. The drop-out issue is greatest in STEM fields, where the drop-out rates are the highest and where it is especially important to obtain more workforce to answer the needs of the quickly developing technology industry and to solve global issues such as climate change.

Drop-out is a complex phenomenon where both an individual’s internal factors and external factors interact with each other, eventually leading to the decision to drop out. In Heublein’s model of drop-out, the internal factors include aspects such as study behavior and motivation, and external factors for example study conditions and guidance (Heublein, 2014). Heublein argues that for the study programme to be successful, these factors should align and alter respectively. Though plenty of empirical research has been done and theoretical models built, existing empirical evidence still has limitations. Previous research often approaches the issue from a variable-centered perspective, which may prevent the identification of the smaller at-risk subpopulations and understanding the complex interrelations behind drop-out. Existing research also lacks a multi-variable perspective which is vital in a multi-faceted process of dropout. As well, attention should be paid to differentiating between types of dropouts and gaining information from the context of different countries’ education systems. A better understanding of the phenomenon could help the work of reducing dropout rates.

We approach this issue using a person-centered approach to examine the study profiles of first-year university students. We aim to identify distinct patterns of students’ study orientations across dimensions of motivation, learning approach, and experienced stress. In this study, we explore what type of study profiles can be identified from first-year science and mathematics students and whether the profile membership is related to first-year grade point average (GPA). The variables included are interest, self-efficacy, surface learning approach, and academic stress. (Korhonen, 2014; Korhonen & Rautopuro, 2019; Lastusaari, 2018; Widlund et al., 2023). All variables are related to students’ study processes and recognized as being connected to drop-out, and they are malleable variables that the universities have a chance to affect (Condren & Greenglass, 2011; Haarala-Muhonen et al., 2017; Heublein, 2014; Jesús et al., 2022; Kehm et al., 2020; Lastusaari et al., 2016; Parpala et al., 2010). Possible at-risk profiles are observed and discussed.

Identifying plausible profiles helps institutions get a picture of the new students and their support, information, and teaching needs. Intervening with the risk elements at an early stage could prevent dropouts. It also adds important information on the large and yet unclear phenomenon of drop-out, especially from the perspective of the crucial STEM fields and the first study year, and from person-oriented and multi-variable perspectives, also including both self-reported and student register-based variables. 

The data consisted of 177 first-year Finnish university science and mathematics students’ survey answers and grade point averages (GPAs), collected in spring 2023. The self-reported items, interest, self-efficacy, surface learning approach, and academic stress, were used to explore the study profiles, and GPA was used as a direct measure to validate the profile memberships. Interest and self-efficacy were measured with an instrument, originally designed to measure mathematical motivation (Widlund et al., 2023) as the expectancy-value theory’s beliefs and values (Eccles, 1983), and then developed to fit university science and mathematics students. Both interest (α=0.903) and self-efficacy (α=0.858) were measured with three questions, measuring students’ interest in their major and their beliefs about their abilities to perform in their studies. The surface learning approach was measured with a modified version of the ChemApproach -questionnaire (Lastusaari, 2018), originally designed to measure chemistry students’ four different learning approaches, now developed further, ending up with four questions measuring the surface learning approach (α=0.841). Academic stress was measured with an instrument developed in the Campus Conexus -project (Korhonen & Rautopuro, 2019). One question was removed to increase the internal consistency of the measurement, ending up with a four-question solution (α=0.839). All questions were answered on a Likert scale of 1 (Completely disagree) – 5 (Completely agree). Confirmatory factor analysis confirmed the structures of the constructs. All measures were formed by calculating the means of the questions. Latent profile analysis with a three-step method was conducted with the variables interest, self-efficacy, surface learning approach, and academic stress, and finally grade point average (GPA) as an auxiliary variable. First, the number of profiles was obtained by fitting latent profile models iteratively to the data, starting with two and continuing up to six profiles. The best-fitting model was identified by interpreting fit indices. The analysis was conducted four additional times to check robustness. Second, the students were assigned profiles based on the class membership probabilities. Finally, logistic regression analysis and ANOVA were conducted to observe the connection between the profile membership and GPA.

The model with five different study profiles was identified as the best fit. The profiles were named respectively: “well-performing, interested” (55.8%), “lower-performing, interested” (19.8%), “high-performing, interested” (11.5%), “lower-performing, uninterested” (7.4%), and “well-performing, uninterested” (5.4%). The “well-performing, interested” and “high-performing, interested” profiles seemed to not have any major issues in their studies, as they had high interest, mediocre-to-high self-efficacy, low surface learning approach, mediocre-to-low stress, and mediocre-to-good GPAs (M=3.55, SD=0.92 and M=3.09, SD=0.70). The “lower-performing, interested” profile seemed to struggle with all aspects other than interest, having low self-efficacy, and high surface learning approach and stress, and a lower GPA than most of the profiles (M=2.92, SD=0.73), indicating that this profile would benefit from support offered by the university. The two smallest profiles came across as at-risk groups, as both “lower-performing, uninterested”, and “well-performing, uninterested” had low interest, indicating they are not interested in the field they are currently studying. In addition, the former had low self-efficacy, and high surface learning approach and stress, and the lowest GPA of the profiles (M=2.62, SD=0.75), indicating that also their learning habits would need some improvement. These students will most probably end up dropping out if not intervened by the university. The latter, however, didn’t seem to have other challenges than the low interest, as they had high self-efficacy, low surface approach, and high GPA (M=3.46, SD=0.50), indicating that these students may eventually transfer to another study field. The at-risk groups could benefit from the university actively communicating about possible specialization fields and professions, and positive environmental and societal impacts offered by the current study field, helping the students find the motivation towards the study field.

Condren, M., & Greenglass, E. R. (2011). OPTIMISM, EMOTIONAL SUPPORT, AND DEPRESSION AMONG FIRST-YEAR UNIVERSITY STUDENTS Implications For Psychological Functioning Within The Educational Setting [Book]. In G. Reevy & E. Frydenberg (Eds.), Personality, stress, and coping implications for education (p. 133). Information Age Pub. Davies, R., & Elias, P. (2003). Dropping Out: A Study of Early Leavers From Higher Education. Research Report RR386. Institute For Employment Research (IER). Eccles, J. (1983). Expectancies, values, and academic behaviors. In J. T. Spence (Ed.), Achievement and achievement motivation (pp. 75–146). W. H. Freeman. Haarala-Muhonen, A., Ruohoniemi, M., Parpala, A., Komulainen, E., & Lindblom-Ylänne, S. (2017). How do the different study profiles of first-year students predict their study success, study progress and the completion of degrees? Higher Education, 74, 949–962. https://doi.org/10.1007/s10734-016-0087-8 Heublein, U. (2014). Student Drop-out from German Higher Education Institutions. European Journal of Education, 49(4). https://doi.org/10.1111/ejed.12097 Jesús, E., Simón, L., & Gijón Puerta, J. (2022). Prediction of early dropout in higher education using the SCPQ. Cogent Psychology, 9. https://doi.org/10.1080/23311908.2022.2123588 Kehm, B. M., Larsen, M. R., & Sommersel, H. B. (2020). Student dropout from universities in Europe: A review of empirical literature. Hungarian Educational Research Journal, 9(2), 147–164. https://doi.org/10.1556/063.9.2019.1.18 Korhonen, V. (2014). Opintoihin kiinnittymisen arviointia kehittämässä - Nexus-itsearviointikyselyn teoreettista taustaa ja empiiristä kehittelyä: Vol. B:3. University of Tampere. Korhonen, V., & Rautopuro, J. (2019). Identifying problematic study progression and “at-risk” students in higher education in Finland. Scandinavian Journal of Educational Research, 63(7), 1056–1069. https://doi.org/10.1080/00313831.2018.1476407 Lastusaari, M. (2018). Persistence in Major in Related to Learning Approaches - Development of a questionnaire for university chemistry students [Doctoral thesis]. University of Turku. Lastusaari, M., Laakkonen, E., & Murtonen, M. (2016). ChemApproach: validation of a questionnaire to assess the learning approaches of chemistry students. Chemistry Education Research and Practice, 17(4), 723–730. https://doi.org/10.1039/C5RP00216H Organisation for Economic Co-operation and Development. (2019). Education at a glance 2019 : OECD indicators (p. 493). OECD Publishing. https://doi.org/https://doi.org/10.1787/f8d7880d-en. Parpala, A., Lindblom-Ylänne, S., Komulainen, E., Litmanen, T., & Hirsto, L. (2010). Students’ approaches to learning and their experiences of the teaching-learning environment in different disciplines. British Journal of Educational Psychology, 80(2), 269–282. https://doi.org/10.1348/000709909X476946 Widlund, A., Tuominen, H., & Korhonen, J. (2023). Motivational Profiles in Mathematics - Stability and Links with Educational and Emotional Outcomes [Manuscript submitted for publication]. https://doi.org/10.31234/osf.io/ugrpy