I’m at Cornell for the Common Solutions Group Meeting.
First part of workshop will deal with e-learning, afternoon with collaboration tools.
Anne Moore from Va Tech is talking about categories for thinking about evaluation of success for learning technologies. She starts by talking about the 1999 National Academies report on Being Fluent in Information Technology. One point they made is when you look at critical thinking and sustained reasoning, you need to look at those skills in an environment that is technology assisted, but in a domain. The report hasn’t been largely read or applied.
Being able to assess higher level of skills is becoming more important due to the emphasis on accountability. We’ve been more focused on inputs, rather than outcomes – academic institutions are largely still not focused on being able to demonstrate learning outcomes.
Joel Smith from Carnegie Mellon talks about the Open Learning Initiative at CMU. He call this Scientifically Informed Digital Learning Interventions.
The challenge is to design and build fully web-based courses which by rigorous assessments are proven to be as good or better than traditional teaching methods. There are multiple ways of building those courses.
Why? Increased access, improved effectiveness, providing flexibility, contain costs.
The current structure of higher ed presents substantial roadblocks to the application of proven results and methodologies from the learning sciences. We depend on individual faculty to develop courses – before we had lots of info from cognitive and learning sciences teaching may have been more of an art than a science. But it’s not fair to saddle each faculty with having to know all that cognitives science. There’s an opportunity in e-learning interventions, developed by collaborations of experts, to embed the knowledge of learning sciences to make the practice more effective.
OLI Guiding Assumptions:
– Digitial learning interventions can make a significant different in learning outcomes.
– Designs grounded in learning theory and evaluation have the best chance of achieving the goal.
– A possibe, acceptabe outcome is failure or mixed failures and successes – not promoting technology for its own sake.
– Formative assessment iwll be a major feature (and cost component, like 40% of budget) of designs and improvement of courses.
– IT staff working with faculty is too limited a partnership – learning scientists, HCI experts, and assessment experts must be part of design, development, production, and iterative improvement.
OLI courses are available in http://www.cmu.edu/oli . Don’t expect an “OCW experience” this project has a different set of goals than OCW. These are full courses, designed for real learners. “Clicking around” will be unsatisfying: these interventions are designed to support a novice learner in acquiring knowledge workin on their own.
Key elements in OLI courses:
– Theory based –
Builds on prior informal knowledge. We know that building on informal knowledge helps people learn faster. Example is an economics course that has exercises that builds on student knowledge of markets based on eBay. Includes cognitive tutors that have just a few node trees on giving feedback of correct or incorrect decisions.
Provides immediate feedback in the problem solving context – midterm and final is hardly immediate or rich.
Promote autheticity, flexibility, and applicability. Real world problems, which are messy and not clear-cut, is much more effective in promoting better learning outcomes.
– Feedback loops (The killer app) – courses record student activity for robust feedback mechanisms. Can feed info back into database or to faculty – this can change the nature of education. Can also give feedback to course designers and learning scientists.
There are papers and evaluations of outcomes on the OLI web site. One example is in statistics – in the first iteration the online students did as well as the students in the traditional course which itself had been worked on for ten years with cognitive scientists. Then they taught the course in a blended mode (meeting with faculty once a week, using OLI as the textbook) in half the time. Students (randomly selected) showed significantly greater gains than the traditional course. Now considering teaching all of the sections that way.
Courses are instrumented to provide instructors with lots of feedback. Faculty can be far more effective when they know what concepts the students are getting and where they’re having problems. The vision is to have a digital dashboard for faculty and students.
“Improvement in post-secondary education will require converting teaching from a ‘solo sport’ to a community-based research activity” – Herbert simon
Deborah Heyek-Franssen, from Colorado is talking about Carts & Horses in the Collaborative, Social Space. Technology is the cart, pedagogy and content should be the horses.
The basics – understand elements of learning, articulate content goals, find pedagogical method, and choose appropriate tool.
Some elements of learning –
Working memory – limited, seven “chunks” at a time. What does this mean for pedagogy? Chunking activity and keeping working memory available for learning. Can technology help? It can, but it can also harm it – e.g. slides with gratuitous images and animations. Cognitive load of looking at images or simulations is lower than reading about it.
Engagement – students get engaged in challenging, complex, multidisciplinary tasks involving sustained amounts of time. What does it mean for pedagogy? designing in and out of class activities that engage, including lecture and readings. Collaborative and social tools can help engage students.
Motivation – what motivates students? building motivation into course – rewards for desired activities.
Reflection – explaining and then critically evaluating own and others explanations. Wikis and blogs can help reflection.
Building on past knowledge – students now have opportunity to build global knowledge – e.g. wikipedia.
Deb notes that simulations can be addictive, and Greg comments that addiction doesn’t equate with learning. While that’s right, it seems to me that learning is at least more likely to occur if you’re highly engaged.
Shel notes that in research universities, even when faculty really want to teach, they’re mostly consumed with their research and even when we have tools and staff resources to help them, they’re not particularly interested in spending the time to work on really improving course methodologies. Joel notes that it works much better to engage with whole departments at curricular levels rather than individual faculty.
Cliff notes that collecting lots of real-time data on student activities has a creepy element about it and wonders about what the policy and privacy issues are. Joel says that there’s at CMU there’s an opt-in informed consent form they can assent to. And feedback is not granular at the level of individual students.
Greg says that the problem with assessment is whether or not people will make any changes based on the assessment, and if we don’t have institutions that make changes based on data then it may not be worth spending money on assessment.
Shel says they did a survey of large courses and 80% of students were using Facebook, but only 20% were using Sakai. So when they put the courseware into Facebook, the students didn’t use it there either.