Dr. Womack's EDFD6993 Action Research Page

Page last updated September 26, 2005. I am planning to work with students in Dr. Ramsey's Monday night EDFD6993 Action Research class tonight to offer some technological help with statistics. As the day goes on, I will be posting spreadsheets along with Word scenarios about how and when these statistics would be used in research. Be sure to bring your flash drives tonight so you can get these files and modify them for your individual use.

We now have a way to access my Power Points and handouts from off campus in much the same way as if you had been on campus and within the Local Area Network. Either click ftp://tdata.atu.edu or paste this location into your browser. Either Netscape or Explorer will take you to that FTP area; from there, just go to swdata to find folders for the classes I teach.

What about a descriptive study with one variable and a need for descriptive statistics--with one variable? for the narrative, click here. For a spreadsheet that shows it done, click here.

Need an example of how to set up a correlation study? Gottatestum's.doc is the narrative and Testedum.xls is the Excel worksheet. The narrative asks students (I used to teach both EDFD6003 and EDFD6993 and developed this handout for them) to do some things other than just correlation, so if correlation is all that you are looking for, skip over the standard deviation and percentile stuff and go right to question 3, "Find the reliability of his test."

What about the T test for paired data, perhaps the simplest of the statistical difference tests? Paired_T is the narrative for the problem so that you can see what kind of situation in which it might be used. Paired_T.xls is the spreadsheet.

What about comparing two groups of completely different people? Even if the groups are not equal in size? Click on 6003Flea.doc for the narrative and FleabodkinT.xls to see how the independent or uncorrelated T test works. If you want to see a handout on the calculations, see HOW_TOT.doc.

When we are comparing more than two samples, we use ANalysis Of VAriance or ANOVA instead of using T tests multiple times. It was found by hard experience in industry in the 1920s and 30s that when T tests were used multiple times, the .05 or .01 levels of protection were not maintained. So ANOVA was developed to keep the .05 etc levels steady across many comparisons. See the Hornblower narrative and Hornblower spreadsheet for an example of how and when to use ANOVA to compare 3 or more groups to see if they are equal.

Chi-square or ANalysis of COVAriance situation? See Dr. Womack individually.

Click here for a Microsoft Excel spreadsheet that sets up the problem for a 20-question survey questionnaire. It shows how to get descriptive statistics for individual items and for the total survey. It calculates split-half reliabilities for the 20-item questionnaire.

Here is the link to the ATU webpage that details the review process before research is done. Visit that site for information and for the necessary forms to have proposed research reviewed by the Ethic Committee before human subjects are used.

Important dates. These dates were representative of what I (Dr. Womack) might have done two years ago in the course, but Dr. Ramsey and Dr. Stephenson may choose to do otherwise:

My Favorite Links

How to tell magazine articles from journal articles

Educational Research Basics Power Point
Populations, Samples, and Data: Theory: Library Resources
Writing hypotheses and research questions
Instrumentation issues in quantitative research: reliability and validity
Helpwith choosing between the nine types of studies
Library of Congress
Phi Delta Kappa search facility
ERIC searches on-line
Search engine for legal information
Pathways to School Improvement
Arkansas Department of Education
Arkansas Tech University library search

Student work

Sonya Hawkins--Effects of Whole Language Immersion Melissa LyByer--Action Study in School Discipline
Carla Brigance--Power Point to Improve Writing Scores Back to Dr. Womack's EDFD6993 news page


Return to Dr. Womack's home page

@atu.edu