Sunday, September 2, 2012
UNIT 1: Individualized Instruction, PSI v A-T
In the ever-changing landscape of e-learning, institutions of higher education are constantly striving to make both knowledge acquisition and student experience equivocal in some way to that of what is available on campus. Both the PSI and A-T models offer some insights in how that might be accomplished. Though first conceived in the 1960's to break away from traditional "stand and deliver" lectures, parts of the models lend themselves surprisingly well to today's increasingly asynchronous worlds.
What I think works about both of these models is the ability of students to self-regulate. However, this would only work for students that are disciplined. The ability for students to move through at their own pace allowing those with prior knowledge to go more quickly or those very interested to slow down and search for supplemental material is well-suited for e-learning. The interim quizzes proctored in PSI and worked into the small assembly meeting for A-T are a wonderful way to assess learning. I do however see a big problem in holding future materials from students until a 100% is achieved in the PSI model.
The Audio Tutorial approach seems too time intensive for most content considering the quick versioning of literally everything. Even if you develop the multimedia assets for a history class where nothing changes, you still have everything around it changing. What media will the audio assets be stored in? CD, mp3, DVD, BluRay? Changes and upgrades in LMS could suddenly not support your media. Having students come to a lab isn’t feasible. Changes like Apple effectively killing Flash or W3C instituting HTML5, ever changing accessibility standards, all of this swirling around makes it almost impossible to produce an evergreen piece. This method makes itself overly reliant and therefore vulnerable to the changing tides. Also, the need to meet in small assembly groups with the professor and the weekly synchronous general assembly sessions adds to the time intensity. This would not be cost effective.
The PSI holds its own difficulties for e-learning. Chief among them is the proctor giving feedback on the exam and then releasing the next lesson. Because of the self-paced nature of PSI I would assume students could be working late. If they take a test and the proctor is unavailable to review the test and provide feedback, the student is stuck waiting. In a world where busy professionals are carving out chunks of time for school, I find this unacceptable. This seemed like a key to the model, but I couldn’t tell if the examples in the reading actually held the next unit until mastery was acquired. Also, though Koen professes an adherence to PSI it seems that students were not required to obtain a 100% to complete the lesson, which I find in direct opposition to the model.
I do believe the self-paced model will begin to take hold. The truth is education is now where newspapers were ten years ago. Everyone knows fundamental change is coming but believe they have such a quality product that they are not in danger. That is not true. I’m not sure what the future holds, but I imagine providers like Coursera, MIT’s OpenCourseWare, and Khan Academy will be leading the way as the new model. Check out Duke’s Coursera offerings: http://cit.duke.edu/blog/2012/07/duke-coursera/
Subscribe to:
Post Comments (Atom)
Good point about the "shelf life" of media developed for one of these models. Modules can quickly be outdated if your platform is no longer supported or updated to a new version. Flash and Flash Paper modules for example, no longer viewable on some browsers/platforms. So a media-heavy module like PSI/A-T, where all the content is packaged to be delivered, can be an expensive proposition and one that should take into consideration the ability to repackage materials into multiple formats as needed. Good catch bringing in open courseware and services like Khan Academy into this discussion. They do tend to follow self-pacing type models, so as to avoid the need to hire/retain instructors and keep costs low (most likely).
ReplyDeleteRachel, while I agree what you are saying (about how a student having to take a test, have it graded, and then move on) can be cumbersome, there are ways around that without a live proctor. A self-remediating test can be developed where a minimum allowable score is set and once that score (or better) is obtained, then the next module/reading, etc, would load for the student to move on to. That way if a proctor was not going to be available (for any reason), this setting could be selected and students could move on. If a student does not meet the minimum score, then they have the self-remediation from that test, they can look over it, reread as necessary, and then take the test again. Of course a random selection test bank would need to be built, but it's not difficult to do and that way they aren't just retaking the same test, and with the self-remediating tests, the students learn what the correct answer is and why. I use this every single day and it works really well for my students, but they are adult students, and that may make a difference since it is job related.
ReplyDeleteRachel,
ReplyDeleteI think you bring up some good points regarding the PSI model. As it stands, the PSI model is very difficult for students who are fully online and also work full time, have a family, etc. I see real value in the human factor (proctor) of the PSI model being replaced by self-grading assessments within a Learning Management System. For example, in Blackboard, we can set up what they call "adaptive release." In simple terms - content can be released to the student based on their score of a particular assessment completed in Blackboard. So, a student could be presented with a test in Blackboard and only if they score 100% would the next unit (of content) be released to the student. I don't see a lot of faculty using this function (adaptive release) for this purpose - maybe because we don't have many of pass/fail courses. Or, maybe because it takes so much time to build all those different assessments - as I recall, using PSI, each time a student is presented with an assessment, it is different. So, they would not be taking the same test over and over until they reached 100%. Instructors would have to have a lot of questions in their test banks!
So, after typing up this comment, I paused and took a moment to read what a fellow classmate wrote in her comment above(Melissa Hill). Surprise! It is very much in alignment with what I just wrote. I agree with Melissa - having a random selection test bank would be beneficial in implementing the PSI model in an online environment. I'm glad to see that someone else had these very same thoughts.....