The importance of agency in the learning process and institutional strategies to increase a sense of agency to enhance academic outcomes was noted already a few decades ago (Thomas, 1980). Student agency and supportive educational practices for it are, however, scarcely studied in the higher education field. This is surprising as recent work-related learning research (e.g., Eteläpelto et al., 2013; Goller & Paloniemi, 2017) has broadly analysed the concept, and argued for the central role of agency in experts’ work. Agency relates to expert work demanding creativity, collaboration, joint knowledge construction, and development of work practices (e.g., Eteläpelto et al., 2013; Hökkä et al. 2017). It has a crucial role in (lifelong) learning and in coping with uncertainty and changes in working life (Su, 2011). Existing literature (e.g., Trede, Macklin & Bridges, 2012) claims that universities typically focus on content-based knowledge construction of individual learners, instead of preparing students for taking agentic stances on work and professionalism. To support students in their agency construction, we need research-based tools for analysing agency in learning situations.

The growing capacity of current technologies has made it possible to collect evidence of learning progressions in different learning environments. As a result, a new emergent field, learning analytics (LA), has been gaining interest in the recent decade (Bond et al., 2018). The purpose of LA is to collect and analyze educational data by creating models and patterns to understand and improve learning and arrangements within learning environments (Ferguson, 2012; Siemens, 2013). However, recent reviews have revealed that there is still little evidence of the effect of LA on learning outcomes, or on the support of learning and teaching in higher education (Viberg et al., 2018). To improve learning practice using learning analytics, Viberg et al. (2018) suggest to critically consider the choice of data and purpose of its use. They call for taking into account the discussion in the learning sciences. We see the possibilities with the concept of agency in a) providing a holistic perspective to understand the constituents of intentional, purposeful and meaningful learning, b) grasping the call for studying the complexity and dynamic interaction related to learning situations in higher education (e.g. Haggis, 2009), and c) advancing the linkage between LA and learning.

The study is based on the existing conceptualisation of agency in higher education context and the developed multidimensional quantitative assessment instrument, the Agency of University Student (AUS) Scale (Jääskelä, Poikkeus et al., 2017). University students’ agency involves access to (and use of) different resources. Personal resources capture aspects of efficacy and competence beliefs. Relational factors comprise, in particular, power relations between the teacher and students, and experiences of trust and support in the learning situations. Participatory resources refer to student-perceived utility of the course, opportunities for participation, making choices and influencing, as well as utilising peer support. Within a project on the development of higher education, we have developed a students’ self-assessment tool that bases on the AUS Scale and includes standardised feedback for the respondents. 

In this study, we analyse the possibilities of LA and the developed self-assessment tool to support student in their agency construction in higher education. We apply LA for the concept of agency, and call this method as agency analytics. The following research questions were set: 

1.    What kind of course-specific knowledge of student agency can be obtained with agency analytics?

2.    How do the students perceive the utility of the self-assessment tool developed in the project to support students in their agency development?

To answer the first research question, the data from students were collected using the AUS Scale (Jääskelä, Poikkeus et al., 2017). The present AUS scale constitutes 11 dimensions and 58 items. Each dimension included three to seven items rated using a five-point Likert scale (1 = fully disagree; 5 = fully agree). In addition, students were asked to respond to two open questions concerning their experiences of which factors in the course have a) fostered them as students and b) constrained their learning. A total of 208 university students (two university courses from a different field: information technology, n = 130; teacher education, n = 78) filled in the questionnaires at the end part of their course. The respondents of both courses were at the beginning stage of their studies. In terms of quantitative data, underlying factor pattern matrix of the validated AUS questionnaire is used to obtain the individual student’s agency factors. Missing values are imputated using nearest neighbor (NN), and computed factors were rescaled to represent the original Likert scale values from 1 to 5. Robust nonparametric techniques (Kärkkäinen & Heikola 2004) were used to obtain the general agency profile of all student in the course, which is provided in comparison to the student’s agency profile. The teacher of the course receives the clustered results of all students, which in our analysis reveals four different interpretable agency profiles. A robust k-SpatialMedians++ (Hämäläinen, Jauhiainen & Kärkkäinen 2017) algorithm was used in clustering. To answer the second research question, 16 students (8 from both courses) were interviewed a few weeks after the course. They were asked to reflect a) the authenticity of the self-assessment they had done in the course and b) the utility value of the summary constructed on their agency ratings and given them beforehand. As a part of the summary, the students got also information of the concept of agency and guidance of how to interpret the summary results. Thus, the students were also asked to reflect their understanding concerning the concept of agency. Data from the interviews and the open questions in the questionnaire were analysed according to questions with using qualitative content analysis method (Patton, 2015).

This study deepens understanding of the critical elements students experience in terms of agency in the course context and contributes to the development of measurement tools for assessing agency in higher education context. Concretely, the agency analytics we have developed provides information on the agency for the individual student including comparison to peers in the same course. The teacher of the course also receives an aggregated data in the form of agency profiles. The students perceived the meaning of the assessment tool differently: For the one group of the students, the assessment tool offered useful knowledge of agency and opportunities to analyse one’s own studying from the new perspective that inspired them. As for, the other part of the students did not find the utility of this type of assessment in terms of their own development. The students’ attitudes to the summary results based on their own self-assessment varied. Some of the students considered the summary as a diagnosis about their agency made by an outsider. Whereas, the other counterparts took a self-reflective role to the summary results, and activated to analyse the temporal and contextual factors that might influence their assessment moment. In the presentation, we will demonstrate the set of options the agency analytics can offer. We will also display the developed assessment tool. Finally, we will discuss what kinds of opportunities and challenges are included in agency analytics and the self-assessment tool in terms of supporting students in their personal development and developing university pedagogy.

Bond, M., Zawacki-Richter, O., & Nichols, M. (2019). Revisiting five decades of educational technology research: A content and authorship analysis of the British journal of educational technology. British journal of educational technology, 50(1), 12–63. Eteläpelto, A., Vähäsantanen, K., Hökkä, P., & Paloniemi, S. (2013). What is agency? Conceptualizing professional agency at work. Educational Research Review, 10, 45–65. Ferguson, R. (2012). Learning analytics: drivers, developments and challenges. International Journal of Technology Enhanced Learning, 4, 304–317. Goller, M., & Paloniemi, S. (Eds., 2017). Agency at work. Agentic perspective on professional learning and development. Professional and Practice-based Learning series. Cham, Switzerland: Springer. Haggis, T. (2009). What have we been thinking of? A critical overview of 40 years of student learning research in higher education. Studies in Higher Education, 34(4), 377–390. Hämäläinen, J., Jauhiainen, S., & Kärkkäinen, T. (2017). Comparison of Internal Clustering Validation Indices for Prototype-Based Clustering. Algorithms, 10(3), 105. Jääskelä, P., Poikkeus, A-M., Vasalampi, K., Valleala, U-M., & Rasku-Puttonen, H. (2017). Assessing agency of university students: Validation of the Agency of University Students Scale. Studies in Higher Education, 42(1), 2061–2079. Kärkkäinen, T., & Heikkola, E. (2004). Robust formulations for training multilayer perceptrons. Neural Computation, 16(4), 837–862. Patton, M. Q. (2015). Qualitative methods & evaluation methods (4th ed.). Thousand Oaks, CA. Sage Publications. Redecker, C., & Johannessen, Ø. (2013). Changing Assessment - Towards a New Assessment Paradigm Using ICT. European Journal of Education, 48, 79–96. Siemens, G., & Baker, R.S.J. d. (2012). Learning analytics and educational data mining. Proceedings of the 2nd International Conference on Learning Analytics and Knowledge - LAK ’12 (pp. 252). ACM Press: New York, USA. Siemens, G. (2013). Learning Analytics: The Emergence of a Discipline. The American Behavioral Scientist, 57(10), 1380–1400. Su, Y-H. (2011). The Constitution of Agency in Developing Lifelong Learning Ability: The ‘Being’ mode. Higher Education, 62, 399–412. Thomas, J. (1980). Agency and achievement: Self-management and self-regard. Review of Educational Research, 50(2), 213–240. Trede, F., Macklin, R., & Bridges, D. (2012). Professional identity development: a review of the higher education literature. Studies in Higher Education, 37(3), 365–384. Viberg, O., Hatakka, M., Bälter, O., & Mavroudi, A. (2018). The current landscape of learning analytics in higher education. Computers in Human Behavior, 89, 98–110.