What will a ANOVA test do and do I need one?

Question

I have four subject groups. Three test groups and a control, plus I am grouping the test group into a aggregate for comparison to the control. The groups are being measured for seven varibles with seven different scores on standardized test both pre-and post-treatment. What will a ANOVA do? Will it measure the variance between the test? Which data sets should I include, Pre and Post or post only? Do you know of a good on-line calculator for them?

Answer 1

The ANOVA looks to see if the means of a population are the same (your null hypothesis) or if they differ between populations (your research hypothesis) by looking at the variances.

I guess an ANOVA could work if you just use the post only results, but I dont really know for sure as as I said, cant quite get my head around what you are doing. Could you go into further detail?

I dont know of an online calculator, but if yo have Microsoft Office, you can install an Excel thing that allows you to do data analysis (the program is called that from memory) and it will allow you to perform all sorts of statistical tests.

Answer 2

What Is An Anova Test

Answer 3

An ANOVA will tell you if there is a statistically significant difference between group means (averages) based on group variances and sample sizes.

An ANOVA can do this to multiple groups simultaneously, so there is no need to group the 3 test groups, that can actually hurt you. For instance, if there was a mean shift in test group 1 and not in the others, the ANOVA may not detect the difference, and if it does, you wouldn't know what test group was different.

ANOVA requires statistically equal variances among your groups, so that will need to be verified before you can count on your results. ANOVA does not test for differences in variance (standard deviation).

The test groups that you should perform ANOVA on depends on what question you are trying to answer. If you are asking "Is there an average difference between pre and post tests?" Then I would recommend you do multiple 2-Sample T-tests: group 1 pre and post, group 2 pre and post, etc. If you are asking "Are there average differences among pre or post groups?", then I would use ANOVA with the pre scores only for all groups, then a second ANOVA with the post scores for all groups.

I use Minitab for such calculations and am not aware of an online ANOVA calculator. If you Google 'ANOVA calculator', there are just a few hits. I broused around a bit and found most of them to be decent. Try this one for instance: http://statpages.org/ then http://www.physics.csbsju.edu/stats/anova.html This one even has a graphical output in PDF!

Good luck!

Answer 4

This Site Might Help You.

RE:
What will a ANOVA test do and do I need one?
I have four subject groups. Three test groups and a control, plus I am grouping the test group into a aggregate for comparison to the control. The groups are being measured for seven varibles with seven different scores on standardized test both pre-and post-treatment. What will a ANOVA do? Will it...