To have a valid experiment that will lead to the proper conclusions, the researcher must have a strong experimental design, and Lessons 20 and 21 dealt with this particular issue. Far and away, the remark that I remember most from these two classes is when Dr. Norman told us to stay away from preexperimental designs because they are non-interpretable. From listening to him, I understood the importance of randomization in testing, and I also realized that randomization is a big part of one’s experimental design. The researcher must be very careful when gathering his or her sample population to make sure that the results are not forced, and randomization is probably the best way to guard against that type of fallacy.

Of the types of true experimental design discussed in the lecture, my favorite was the Solomon four-group because of its completeness aspect. This design was preferable to me because it looked at every possibility, a Control Group that received the treatment and one that did not, and an Experimental Group that received the treatment and one that did not. From a mathematical point of view, every test situation is examined, so the proper conclusions should be more easily achieved.

I found the discussion about the causal-comparative research and the problem of spurious results very interesting, especially the class example of the number of drunks in a town and the number of churches. This example explained to me how easily a researcher can draw the wrong conclusions if he or she does not properly evaluate the evidence.

Back To TLEE's Home Page

Index to Selected Essays And Book Reviews

Lesson 24. Descriptive and Survey Research

Back to Lesson 21

Send email to: tlee6040@aol.com