13-­‐05-­‐06 Learning Analy2cs and Learner Profiling for Early Alert and Help Provision Jim Greer, Director University Learning Centre Brad Wuetherick, Program Director Gwenna Moss Centre University of Saskatchewan 2013 Learning Analy2cs A plethora of data about learners + Tools to analyse, cluster, model and predict Deeper personalized informa2on about learners Which can be used for many purposes (beneficial and poten2ally harmful) 1 13-­‐05-­‐06 Uses of Learning Analy2cs • Guide and inform course and program design • Improve quality and accuracy of assessments • Improve quality of communica2on between learners, teachers, and advisors. • Provide personalized support and early alerts • Improve administra2ve data for strategic enrolment management How people are using LA and data • Profiling students – Individual risk factors (aXri2on, academic failure, subject-­‐specific deficiencies) – Individual capabili2es (targeted recruitment for graduate work, poten2al leaders, scholars needing enrichment) – Resource recommenda2ons (based on interests or needs) – Group characteris2cs (class demographics, program demographics, university demographics) 2 13-­‐05-­‐06 How people are using LA and data • Tracking student development, achievement and behaviours – Personalized adap2ve learning support – Adap2ve assessment – Adap2ve learning resource recommenda2ons – Personalized messaging and targeXed communica2on How people are using LA and data • Adjus2ng instruc2onal strategies and course designs • Evalua2on of programs based on measured learning outcomes • Longitudinal studies of reten2on and effects of administra2ve ac2ons 3 13-­‐05-­‐06 Some UofS projects • Entry census – Gathering richer data about incoming students (richer demographics such as first-­‐in-­‐family, funding sources, living arrangements, feelings of connectedness, approaches to learning, career goals and plans, etc.) – Merge with ins2tu2onal (admissions) data – Richer profiling • Builds on Large Classes Survey run 2009-­‐12 Some UofS projects • Resource recommender system – Learning and help resources in University Learning Centre • Web resources • Instruc2onal videos • Workshops • Personal consulta2ons – “Amazon-­‐style”, “Nealix-­‐style” recommenda2ons • Based on ra2ngs by “like students” 4 13-­‐05-­‐06 Some UofS projects • Adap2ve individualized learning systems – McGraw Hill, Pearson, and other publishers are rolling out personalized study aids, adap2ve tutors, adap2ve quizzes – being used in selec2ve courses now. – Online assessment and self-­‐assessment tools for diagnosis and placement in courses – Intelligent tutoring systems are finding their way out of research labs and into classrooms Planned projects at UofS • Early alert system pilot – Combining risk profiles with ac2vity tracking in courses • Use of learning management system • Clicker data • Assignment grades – Provide a dashboard display to student and prof – Encourage earlier interven2on – Poten2ally involve academic advisors 5 13-­‐05-­‐06 Planned projects at UofS • Concept inventory score tracking – Longitudinal assessment of learning gains in discipline-­‐specific knowledge – Modelled on “Force concept inventory” in Physics – Can be used to iden2fy problem points in courses in a program Privacy and ethical use of data • Profiling can be discriminatory and prejudicial • Knowledge of student risk factors can result in instructor bias (even if uninten2onal) • Students have a right to keep personal informa2on private • Importance of having privacy officer and student advocate involved • Different access for counsellors, advisors, teachers, administrators 6