Learning Analytics is benefitted in education in different levels for different stakeholders; such as governmental level for policymakers, institutional level for curriculum developers, pedagogical level for teachers and learners . I have developed my own definition for LA in one of my assessments based on my experiences which I gained from the past research studies.
“Learning Analytics is a technique of revealing patterns based on the data obtained from the access logs, surveys, feedbacks, interviews, observations, historical facts, students’ marks and results in educational environments which can be used to formulate descriptive, predictive, diagnostic or prescriptive information models for proposing innovative resolutions to address the teaching and pedagogical concerns. LA is more targeted at changing teaching and learning techniques at the individual level.”
Attaining the LA expectations of enhancing and emphasizing the learning process which best matches with teacher and learner requirements to achieve the learning outcomes effectively is a very challenging task which requires a lot of data collection and extensive data analysis .
The features of LA depend on various data; concepts or theories collect for different research purposes which were used in LA research studies as listed below ;
- Descriptive research describes the phenomenon and available status
- Philosophical research which reflects a phenomenon without data or a theory
- Theoretical research which reflects on a phenomenon based on a theory without test data
- Theory application research which applies a theory or models to a collected data set
- Theory generation research which uses a data set to propose a new theoretical model in pedagogy
- Theory testing research which tests a theory using collected data
I would like to share some thoughts from a research paper which I submitted to IEEE journal (in the review phase) to exemplify the LA usage for “theory generation research which uses a data set to propose a new theoretical model in pedagogy” (No 5).
In my research I developed the following activity access behavioural model; e-Activity Access Behavioural Model (eAABM) (figure 1) under the supervision my supervisors Prof Roshan Ragel and Mr Sampath Deegalla and tested using different Data mining classification techniques.
Access time and access hits of each activity were calculated for each student using their access logs stored in the log table of the Moodle database. Data mining discretization technique  was used to calculate the access group of each student in each activity according to the rules available in the above picture.
As per my understanding, students’ psychological ownership  level towards the online activities gives a clear picture of students willingness to access the intended activities. Therefore from the following paragraph, I have argued how this behavioural model represents the educational ownership level of a student in the given e-activity.
“Psychological ownership, the sense of possession is considered as a reliable measure to capture the students’ ownership towards technology-enabled learning like online self-learning . The sense of ownership towards the considered online activity has a positive impact on the retention time and learner involvement . Accordingly, it can claim that the groupings in the eAABG model can reflect the different level of psychological ownership since those who possess high ownership typically get frequent visits and high retention in a learning environment and vice versa. Therefore, the characteristics of groupings can form different ownership levels which can describe different access scenarios as stated in the table. ”
Apart from using eAABG model to represent educational ownership, it can be used for online adaptive learning solutions since it represents the learning behaviour simply from a single parameter.
 C. Schumacher and D. Ifenthaler, “Features Students Really Expect From Learning Analytics,” 2016.
 D. Ifenthaler and C. Widanapathirana, “Development and Validation of a Learning Analytics Framework: Two Case Studies Using Support Vector Machines,” Tech Know Learn, vol. 19, p. 221–240, 2014.
 O. Viberg, M. Hatakka, O. Bälter and A. Mavroudi, “The current landscape of learning analytics in higher education,” Computers in Human Behaviour, vol. 89, pp. 98-110, 2018.
 A. Gupta, K. G. Mehrotra and C. Mohan, “A clustering-based discretization for supervised learning,” Statistics and Probability Letters, pp. 816-824, 2010.
 I. Jussila, A. Tarkiainen, M. Sarstedt and J. F. Hair, “Individual Psychological Ownership: Concepts, Evidence, and Implications for Research in Marketing,” Journal of Marketing Theory and Practice, vol. 23, no. 2, pp. 121-139, 2015.
 I. Buchem, “Psychological Ownership and Personal Learning Environments: Do sense of ownership and control really matter?,” in The third PLE conference, Aveiro, Portugal., 2012.
 V. S. Asatryan, L. Slevitch, R. Larzelere, C. Morosan and D. J. Kwun, “Effects of Psychological Ownership on Students’ Commitment and Satisfaction,” Journal of Hospitality & Tourism Education, vol. 25, no. 4, pp. 169-179, 2013.