Home > Statistics Every Writer Should Know > The Stats Board > Discusssion
Experimental Design
Message posted by Edwin on December 1, 1999 at 12:00 AM (ET)
Hi,
I really need help. I am doing research on mathematics
education. I intend to compare the performance of 12 groups of students.
There is one control group and 11 treatment groups.
All the students will be tested and taught individually.
Should I just lump them all into one experiment or
do I break them up into several experiments? What are the disadvantages
of lumping them all into one experiment
with regard to the statistical analysis
done. I hope to run a one-way ANOVA. The groups are as follows:
Conventional Teaching (Control)
Conventional Teaching+Use of calculator
Conventional teaching+Use of Heuristics
Conventional Teaching+Use of calculators+Use of heuristics
Worked Example
Worked Example+Use of Calculators
Worked example+Use of heuristics
Worked Example+Use of calculator+use of heuristics
Pure Worked Example
Pure Worked Example+Use of Calculators
Pure Worked example+Use of heuristics
Pure Worked Example+Use of calculator+use of heuristics
Would really appreciate any help I can get.
Please help?
Ed.
READERS RESPOND:
(In chronological order. Most recent at the bottom.)
Re: Experimental Design
Message posted by Bill on December 2, 1999 at 12:00 AM (ET)
Remember the KISS (kept it simple stupid) approach. From my experience, many studies collapse under their own weight. Unless you have a tremendous amount of resources, this design looks as if it may be a candidate for such a collapse. You may find it helpful to focus your study, i.e., conduct several smaller experiments.
Your $5 contribution helps cover part the $500 annual cost of keeping this site online.