This is my continued report on my attendance at the 2015 AECT Convention in Indianapolis, Indiana.
I attended a session on learning analytics. There were several presenters, and the first presentation by Dongho Kim and associates from the University of Georgia entitled What Makes a Learning Analytics Dashboard Successful? Suggestions for designing a learning analytics dashboard include the following:
- use an intuitive and simple interface
- use concise and clear charts and graphs
- provide a description of how to interpret visualized information
- use specific examples to explain information
- discover and take into account what the learners seem to find useful
- remind learners of the learning process
- provide a tutorial to help learners understand how the dashboard functions
In the next presentation, entitled Course-tailored Prediction Model: Enhancing Learner Performance Prediction by Adaptive Use of Proxy Variables by Meehyun Yoon and associates, presenters noted that when working to predict student success in a course that is discussion-heavy, the following categories and variables were used. Note that they used what they called "proxy variables" which represented the component they were trying to measure.
- Active Participation
- Total time spent on LMS
- LMS visit frequency
- Discussion board frequency
- Number of postings
- Engagement with discussion topic
- Length of posting
- Discussion time per visit
- Consistent Effort and Awareness
- LMS visit regularity
- Discussion visit interval regularity
- In-degree centrality
- Out-degree centrality
Using these proxy variables, the researchers were able to predict student success in the course with nearly 70% accuracy in the first week and up to 90% accuracy by the middle of the course. Again, this is a discussion-heavy course, but the results are quite interesting.
I also attended a session on doctoral mentoring. This session had two excellent presenters giving separate presentations- Swapna Kumar and Kathleen Andrews. Dr. Kumar discussed an experiment in which doctoral students were mentored as groups of 4-5 students who moved through the program together. They shared their resources and gave one another support throughout the learning process. When dissertation mentoring took place, it was done with the whole group intact so others could see how their peers were rated or judged and adapt and learn. It seemed like a good model, though they did acknowledge that it may not have been more time-efficient for the faculty member.
Dr. Andrews shared a set of competencies for mentoring doctoral students (Schichtel, 2010). They included
- Developmental Learning
The presenter mentioned that she thought there was a need for training doctoral mentors more fully. A participant noted that many of the for-profit groups were much more effective at providing training and support than more traditional universities - an interesting finding.