The researcher noted that AutoTutor has been tested with more than 1000 students and delivered "learning gains of approximately one letter grade".
i know how these studies are tested, the students are cherry picked for the best possible out come, like connect 4 was back in the early 90's.
kids left to study what ever they wanted at their own pace at their own will. they barely beat out the 'cream of the road apples' students that were purposely excluded from the test group by 1/2 a letter grade.
the finding was when you put all the best students in one class with one teacher and all the other students in the other class, the teacher was better able to focus the class to the needs and requirements of the students with out having to cover such a large and wide margin of best to worst to keep them engaged, which also led to throwing out the x factor of optimal teacher to pupil ratio. but liberals and parents who found out the kids on the low end were all grouped together called this racism and it was immediately shot down and banned from discussion.
and yet china does this. so the government can better assign it's people.