Learning Analytics Research Questions

learning with insight

LA – Issues for Research

Contributor: Prof Vivekanandan Kumar

  1. “Learning is a change in the mental representations of the world” (Rob Kopper, ICSLE 2015). How does LA influence changes in students’ mental representations in select domains and activities?
  2. How to reduce barriers to LA adoption by Universities’ administration and by students?
  3. How much of course outcomes have been met by this student so far in the course?
  4. To what extent formative/continuously arriving data arising from the study episodes of this student contribute to measuring competencies related to course outcomes? PS: A study episode includes data from the LMS as well as data from other tools outside the LMS (e.g., IDE, reading of an online book, RStudio, MATLAB, writing in Microsoft Word, video of students working on the Science lab kit, …)
  5. How to measure student sentiments from their discussion forum postings and other such communication media (e.g., emails, LMS-feedback channels)? What is the rate of change of sentiments of this student at different points in time in the course? Does the change correlate with specific/targeted competency growth in subject matter?
  6. Can observational data (ethical, formative and continuous) from student’s assignments and projects be used to automatically mark/assess competencies and accomplishment of course outcomes?
  7. Does the introduction of self-regulation, co-regulation and grit modules in task tool inside and outside LMS make an impact on the regulation and grit competences of students? Regulation has little to do with measured intelligence and anyone can develop it (R. J. Davidson, 2003; Tinnesz, Ahuna, & Kiener, 2006). Regulation is rarely introduced as part of education! Self-control, self-discipline, perseverance, and determination in pursuing long-term goals outweigh IQ as predictors of postsecondary academic success and, among children, performance on a spelling bee (Duckworth, Peterson, Matthews, & Kelly, 2007; Tough, 2012).
  8. Typically, LA systems offer functionalities to raise questions about “explain what happened in the past”, “diagnose why it happened”, and “predict what will happen in the future”. If the explain-diagnose-predict outcomes are available, can we venture further into “prescribing activities that, if accomplished by students to a specified extent, will guarantee fruitful outcomes to students”? 
  9. Can we classify the types of pathways/groupings and the associated timeline students pursue to complete a course? (e.g., read LMS, watch youtube, discuss with friends/classmates, contribute to social media, consult tutor/coordinator, read books, work-on assignments). Is there a particular pathway/grouping that is proving to be successful for a majority of students? 
  10. Do students trust LA solutions? How to improve the level of trust between students and cognified LA tools? 
  11. Does content adaptation (e.g., based on knowledge-space theory) enhance learning performance of students? 
  12. Should a reading analytics system be included within the LMS to help students better comprehend the text matter? Ref: Atsushi Shimada, Shin’ichi Konomi, Hiroaki Ogata Real-Time Learning Analytics System for Improvement of On-Site Lectures Interactive Technology and Smart Education, Vol.15, No.4, pp.314-331, 2018. 
  13. What is the impact of learner performance and student satisfaction if a set of metacognitive tools (e.g., self-regulation and co-regulation enhancers, grit enhancers) are introduced at the LMS environment rather than at the task environment (e.g., IDE for coding tasks, Editor for writing tasks)?
  14. Since most of the data formatively/continually collected during the execution of a task by the student (e.g., writing essays, solving programming problems, creating test cases for a software application, designing software engineering solutions, arguing for a research design) is deemed ‘intrusive’ in nature, what type of ethically-valid data-procurement mechanism should be in place to defend the use of LA systems? What other data ethics should be in place for LA systems?
  15. How to measure the operational quality of a course being conducted at AU based on both static metrics (e.g., quality of video clips) and dynamic metrics (e.g., timeliness of response postings by tutors/coordinators)? PS: About 200 course design and course delivery metrics have been identified and populated in a Dynamic Bayesian Network. 
  16. How to detect students’ ADD/ADHD afflictions, measure the impact of these afflictions on the writing processes of students, and offer ‘cognified scaffolds’ to assist them, in real time, in their writing process? This link has a snap shot of a draft BBN that lists both static and dynamic elements – https://drive.google.com/file/d/1EI_FXdbuSg5TL8TJDBsO9Yat94IFhoIQ/view?usp=sharing 
  17. What is the ‘degree of communication’ between students and learning resources (LMS resources, Tutors/Instructors, other students, non-LMS resources)? 
  18. Is Observational Research Methods more conducive to studies in LA rather than RCT methods? 
  19. Beyond SCORM, what is the impact of standards such as IMS Global’s Caliper (http://www.imsglobal.org/activity/caliper) or xAPI (https://xapi.com/scorm-vs-the-experience-api-xapi/) on data transparency on students’ learning traces? 
  20. What is the level of mastery in English expressed by students in course-level communication? 
  21. How to facilitate the access to textbooks and improve the reading process? (from AIED2019 intelligent textbook workshop proposal)  
  22. What can be extracted from textbook content and data-mined from the logs of students interacting with it? (from AIED2019 intelligent textbook workshop proposal) 
  23. LMS courses typically offer static-guided or even adapted-guided pathways to instruction. In these cases, guidance is designed by course designers. How would these study pathways compare against self-directed pathways where students themselves choose their study pathways in a course? For example, do student satisfaction and student performance improve under ‘agile study pathways’ as opposed to LMS-like pathways? 
  24. How to store, share and protect LA datasets? 
  25. How to improve AU’s individualized study model to include mastery learning and one-to-one tutoring with a view to achieve 2-sigma effect? (http://web.mit.edu/5.95/www/readings/bloom-two-sigma.pdf). 
  26. How to apply mastery + one-to-one tutoring in the context of adult learners, given that 2-sigma improvements have been experimentally shown only in high school levels (grades 4, 5 and 8)? 
  27. How to detect and support students on specific Written Language Disorders (WLD: https://www.asha.org/Practice-Portal/Clinical-Topics/Written-Language-Disorders/Disorders-of-Reading-and-Writing/) and Reading Disorders?