Talk:Adaptive learning: Difference between revisions
imported>Pat Palmer m (Talk:Adaptive Learning moved to Talk:Adaptive learning: more compliant with article naming policy) |
imported>Pat Palmer (a comment on "relatively new") |
||
Line 14: | Line 14: | ||
[[User:Howard C. Berkowitz|Howard C. Berkowitz]] 14:01, 25 July 2008 (CDT) | [[User:Howard C. Berkowitz|Howard C. Berkowitz]] 14:01, 25 July 2008 (CDT) | ||
==beef with "relatively new"== | |||
I'm not sure that I agree that this is "relatively new" as an area of academic enquiry (I worked as a programmer for a psych. prof. at Univ. of Tenn. in 1982 programming algorithms in Fortran for adaptive learning in 1982). I think it would be helpful to be a little more specific--is it new in the marketplace? How new? Relative to ''what''?[[User:Pat Palmer|Pat Palmer]] 20:16, 5 August 2008 (CDT) |
Revision as of 19:16, 5 August 2008
Adaptive testing as well?
This is an interesting subject. You might want to consider adaptive testing; I've worked with developers for it but unfortunately can't give you any immediate links.
This was in adult learning and industry technical certfication. The idea was to avoid unfairly penalizing a student if, for example, one question were asked on a particular topic, and the question were poorly worded, or for some reason wasn't clear to that learner. If the initial question in an area were answered correctly, the test engine would move to the next topic.
If the question were missed, and sometimes the question was deliberately advanced and possibly testing more than one concept, the test engine would ask an alternative question, still a legitimate challenge but to give the student another chance to demonstrate understanding. If the question was one that required understanding of multiple topics, the test engine would be sure to ask one that applied only to a single topic.
It wasn't extremely hard to implement: the question database had each question keyed with the topics it addressed. This also let us use the same question database for different testing requirements: if certification A required a demonstration of knowledge of subjects 1, 1.2, and 4, while certification B wanted to check knowledge is areas 1 and 5, they could both pull from the pool of category 1 questions. 1.2, in this example, would be a more specialized subset of subject 1.
By having a minimum of 3-5 questions on each topic, which the test engine might pick in random sequence, it made it impossible to cheat by memorizing an answer sheet.
Howard C. Berkowitz 14:01, 25 July 2008 (CDT)
beef with "relatively new"
I'm not sure that I agree that this is "relatively new" as an area of academic enquiry (I worked as a programmer for a psych. prof. at Univ. of Tenn. in 1982 programming algorithms in Fortran for adaptive learning in 1982). I think it would be helpful to be a little more specific--is it new in the marketplace? How new? Relative to what?Pat Palmer 20:16, 5 August 2008 (CDT)
- Article with Definition
- Developing Articles
- Nonstub Articles
- Internal Articles
- Computers Developing Articles
- Computers Nonstub Articles
- Computers Internal Articles
- Education Developing Articles
- Education Nonstub Articles
- Education Internal Articles
- Psychology Developing Articles
- Psychology Nonstub Articles
- Psychology Internal Articles
- Computers Underlinked Articles
- Underlinked Articles
- Education Underlinked Articles
- Psychology Underlinked Articles