In an attempt to quantify evidence of student learning, our Lesson Study Project developed a pre- and post test for students who discussed the materials in a classroom setting in which the lesson was taught in the way it is normally taught and for students who learned in the revised classroom setting. If the instrument can adequately measure understanding, we should be able to see a difference in the outcomes of the pre and post-tests in the two classroom settings.
Please find the instrument below.
The results are presented in the following manner. The variables Control_pre and Control_post are the results of six questions on the topic of free trade. The “Control” indicates that the tests were performed on a class without the “LSP” treatment. This class was taught without the LSP enhancements, or the way that we usually treat trade in an introductory microeconomics course. The variables Treated_pre and Treated_post are totals from the same test given to the treated class, which was based on the LSP curriculum developed by the Microeconomics team.
Looking at basic averages of the pre and post totals as well as the difference in the pre and post values, it is clear that non-treatment class (the non-LSP curriculum) improved the most.
Control_pre Control_post Control_Difference Treated _Pre Treated _Post Treated _Difference
Mean 2.2195 5.0930 2.9767 2.3714 2.4857 0.1143
Standard Deviation 1.2147 1.2876 1.8707 1.2531 1.5203 1.7981
The following results are from a simple regression using the post-test results from both the control and the treated results aggregated as the dependent variable. Independent or predictor variables included the pre-test results and a dummy variable for the treatment group (treatment = 1). You will note that the treatment itself had a significant negative effect on the post-test result.
Regression with robust standard errors Number of obs = 112
F( 2, 109) = 48.01
Prob > F = 0.0000
R-squared = 0.4427
Root MSE = 1.444
------------------------------------------------------------------------------
| Robust
post | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
pre | .1028293 .1098087 0.94 0.351 -.114808 .3204667
treatment | -2.630579 .27074 -9.72 0.000 -3.167177 -2.093981
_cons | 4.872441 .327238 14.89 0.000 4.223866 5.521016
------------------------------------------------------------------------------
One cannot conclude from these results, however, that the non-LSP lesson is necessarily better. It is possible that this instrument is too blunt to measure what was learned with the LSP curriculum. The instrument itself was developed under the original curriculum, and may not be adequate for the LSP curriculum.
In addition to the quantitative measure, we also had observers present in three of the lesson-study-project lessons during the spring semester of 2004. We have transcribed three of the observation sessions (see attached).
Contact: Lisa Giddings
Previous Logs: 1 | 2 | 3
Comments