Understanding who are the students choosing STEM careers has become vital (Wang, Ye, & Degol, 2017) coupled with the necessity to foster higher student numbers into the science track (Guo, Parker, Marsh, & Morin, 2015). Gender equity has been at the fore (Nagy et al., 2010) alongside with the idea of supporting low SES students to choose the STEM path (Turner, Joeng, Sims, Dade, & Reid, 2017).

Parallel to that, relationship between science performance (Jansen, Scherer, & Schroeders, 2015), students’ motivation for science (Wang, Chow, Degol, & Eccels, 2017) and various science-related dispositions such as self-efficacy (Sahin, Ekmekci, & Waxman, 2017) and epistemological beliefs (Kampa, Neumann, Heitmann, & Kremer, 2016) have been explored, providing evidence on their strong association. Conversely, despite the high number of ongoing investigations, one of the key features across the field lies in its use of the variable-centred approach. Regardless of the fact, the approach itself has great advantages; its sole use across the educational field may disguise certain subgroups both within and across samples. On the other hand, a person-centred approach assumes the existence of such subgroups and allows patterns to be observed through different subcategories within a population or across diverse cultural settings (Bergman & Trost, 2006). While studies using the person-centred approach are on the rise, the field is still very much lacking attempts, which address different phenomena, in connection to how students may experience science from the person-centred stance. Moreover, studies from diverse cultures settings are also sparse.

Grounded in the affordances of the person-centred approach and the idea on the necessity to understand a very diverse student body in numerous corners of the world,  this paper focuses on investigating self-related dispositions and motivation in science using the 2015 PISA dataset for Italy. Compared to the other OECD countries Italy scores somewhat lower in science than the set OECD average (489 compared to 495 score points), gender differences are significant (17 points in favour of boys), coupled with the fact students in Italy participate in PISA at the point when they have already been enrolled in different programmes at upper secondary level (i.e. grammar school or VET programmes).  

With respect to the above, in this paper, we focus on (a) possible student profiles relative to their enjoyment in science, interest in science, instrumental motivation in science, science self-efficacy, involvement in different science activities and science epistemological beliefs; (b) the relationship between distinguished student profiles and different aspects of science competence; and (c) to what extent distinguished student profiles differ by gender, socio-economic background, and programmes students may be enrolled in.

Student profiles were investigated using the sample of 11 583 fifteen-year-old-students in Italy by means of the latent profile analysis (LPA). From the total number of students, 5792 are girls (50%), 8% are first and second generation students and 99.1% of students were attending upper-secondary programmes. Likert type composite scores: ‘interest in broad science topics’, ‘enjoyment of science’, ‘instrumental motivation’, ‘science self-efficacy’, ‘epistemological beliefs about science’ and ‘students´ science activities’ were used in the LPA. ‘Interest in broad science topics’ includes themes such as the biosphere, motion and forces, energy and its transformation, while ‘students´ science activities’ captures participation in science content in leisure time. The measure ‘epistemological beliefs about science’ includes items describing science as an evolving and changing subject and who holds authority in creating knowledge in the domain. ‘Enjoyment of science’ captures the extent students like being involved in different aspects of activities related to science, ‘instrumental motivation’ relates to their perception of the usefulness of science for future paths, and ‘science self-efficacy’ the extent they perceive themselves competent addressing different problems in science. All measures were used in line with the proposed PISA framework. The LPA is a latent variable mixture modelling technique, which allows the identification of groups of individuals with similar values on the clustering variables used in the analyses (Geiser, 2013). To uncover the number of profiles that emerge from the data, models with 2 through 7 latent classes (k=2 to 7) were tested. Mplus 8.2 was used in the analyses (Muthén & Muthén, 2018). Each model was set with an instruction to use 1000 random sets of starting values. After 50 iterations, the 100 best sets of starting values that were identified by the highest likelihood values were then selected for final optimization. Following this, the chosen latent class solution was validated. Using the AUXILIARY (e) function in Mplus, each of the groups was observed against their competence in science and index of economic, social and cultural status (ESCS). The ESCS is also a composite score constructed from the indicators parental education, highest parental occupation, and home possessions. AUXILIARY (r) function was employed to explore how the distinguished student profiles differ by gender and student programmes. Prior to main analyses, descriptives for each variable were inspected and multiple imputation procedures using Mplus for multilevel data were performed (Enders, Mistler, & Keller, 2016).

In the light of theory and results of previous studies, characteristics of each profile, and their interpretability a 5-class model solution was chosen with the following parameters: Log likelihood=90036.66, number of free parameters=40, BIC=180447.612, pLMR=0.0001, entropy=.820, and smallest class frequency 1.5%. Interpretation-wise the largest student group comprises students that mainly fulfil an interest in science outside the school context (out-of-the-school group - 59.5%), followed by the group that shows no interest in science within or outside the school context (the non-interested - 23.8%). Students that genuinely enjoy and are interested in science in school and own leisure time, perceive themselves as competent in the domain and see the science field as important for their future striving, amount to 12% of all students (the scientists). Following, a small group of students that is not interested, does not enjoy science, does not involve in science-related activities or perceives it as instrumental to their future success is distinguished (3.1% - no use of science group). Finally, 1.5% of students perceive themselves as competent in the field, possess no interest in the science within the school context, yet involve themselves in out-of-school activities connected to science and considers it instrumental to future success (out-of-school explorer). The ‘scientists’ group outperforms others (535.425 point score), while the ‘out-of-school explorers’ lag behind with over one hundred point difference (425.537). The ‘scientists’ group gathers students from higher ESCS families and boys have more chances to fall into this profile. Across the distinguished groups, students attending the vocational education programmes have two to four time less of a chance to be found in the ‘scientist’ group when compared to other profiles. Overall results will be discussed relative to the wider cross-cultural perspective and issue of equity within the education system in Italy.

Bergman, L. R., & Trost, K. (2006). The person-oriented versus the variable-oriented approach: are they complementary, opposites, or exploring different worlds? Merrill-Palmer Quarterly, 52, 601–632. Enders, C. K., Mistler, S. A., & Keller, B. T. (2016). Multilevel multiple imputation: A review and evaluation of joint modeling and chained equations imputation. Psychological Methods, 21, 222e240. Geiser, C. (2013). Data Analyses with Mplus. Guilford Press. Guo J., Parker P. D., Marsh H. W., Morin A. J. S. (2015). Achievement, motivation, and educational choices: a longitudinal study of expectancy and value using a multiplicative perspective. Dev. Psychol. 51, 1163–1176. Jansen, M., Scherer, R., & Schroeders, U. (2015). Students' self-concept and self-efficacy in the sciences: Differential relations to antecedents and educational outcomes. Contemporary Educational Psychology, 41, 13-24. Kampa, N., Neumann, I., Heitmann, P., & Kremer, K. (2016). Epistemological beliefs in science—a person-centered approach to investigate high school students' profiles. Contemporary Educational Psychology, 46, 81-93. Muthén, L. K., & Muthén, B. O. (1998-2018). Mplus User’s Guide: Statistical Analysis with Latent Variables (8th ed.). Los Angeles, CA: Muthén & Muthén. Nagy, G., Watt, H. M. G., Eccles, J. S., Trautwein, U., Lüdtke, O. & Baumert, J. (2010). The development of students’ mathematics self-concept in relation to gender: Different countries, different trajectories? Journal of Research on Adolescence, 20(2), 482–506. Sahin, A., Ekmekci, A., & Waxman, H. C. (2017). The relationships among high school STEM learning experiences, expectations, and mathematics and science efficacy and the likelihood of majoring in STEM in college, International Journal of Science Education, 39:11, 1549-1572 Turner, S. L., Joeng, J. R., Sims, M. D., Dade, S. N., & Reid, M. F. (2017). SES, Gender, and STEM Career Interests, Goals, and Actions: A Test of SCCT. Journal of Career Assessment. https://doi.org/10.1177/1069072717748665 Wang, M. T., Chow, A., Degol, J. L., & Eccles, J. S. (2017). Does Everyone’s Motivational Beliefs about Physical Science Decline in Secondary School? Heterogeneity of Adolescents’ Achievement Motivation Trajectories in Physics and Chemistry. Journal of youth and adolescence, 46(8), 1821-1838. Wang, MT., Ye, F. & Degol, J.L. J (2017). Who Chooses STEM Careers? Using A Relative Cognitive Strength and Interest Model to Predict Careers in Science, Technology, Engineering, and Mathematics. Journal of Youth and Adolescence, 46(8). 1805–1820.