I have just now completed my study in which there were two groups . In control group i taught with didactic lecture and other group taught with modified problem based learning . in both group pre test and post test calculated . in intervention group relative learning gain more and in control group absolute learning gain was more .what it means ?
When doing a study like yours, the problem is that the two groups never start at exactly the same point on the pretest. Just doing "gain scores" or asolute learning does not control for the starting points of the two groups. This is a fatal flaw if you compare groups simply on gain. The more appropriate analysis is the compare the post test scores relative to the pretest scores. An Analysis of Covariance will take care of this and provide a valid comparison.
This sounds like the normalized gain used in physics. I am not sure why physics has taken a lead like this in education.
Years ago in physics Hake aggregated data across the US for the pre and post test scores of the Force-Concept Inventory or FCI, and compared what he called the normalized gain according to teaching methods. He compared the more passive teaching (didactic lecture) to what he called "Active Engagment teaching"
The normalized gain is the gain in score divided by the possible gain in score. So if a student got 18 of 32 the first time and got 24 of 32 the second, they gained 6 points, and could have gained 14 points. This means the normalized gain was about 0.43. He found gains of 0.25 typical of better lecture styles, and gains around 0.4 and above for active styles.
The normalized gain sounds like a relative gain, and the absolute gain sounds like the increase in points or percent.
Erik Mazur at Harvard has gotten some interesting results in improving FCI scores using clickers.
Richard Hake published his results in the American Journal of Physics. AIP Conf. Proc. 399, 595 (1997); http://dx.doi.org/10.1063/1.53091
(The FCI is found in The Physics Teacher, Vol. 30, March 1992, 141-158)
Sorry for the delay in replying . Thanks for clearing the concept Dr.Singh ,
Dr. Lederman and Nielson . What i understood was we should go for relative gain when doing such studies .
Relative gain is more meaningful than absolute scores or gains, but it still suffers from ceiling effects. There are also some interesting and unexpected predictors of success on the FCI.
- Badges
- Science topic
- Similar topics
- Social Psychology
- Attitude