Intro to Quantitative Education Research Methods

Wednesday 1:30pm-4:00pm UMC 247
Mini Workshop

Convener

Tim Shipley, Temple University
This session is for those who want to increase their ability to conduct education research and to evaluate published research. Participants will have a chance to learn about principles of statistical testing and interpretation, limitations on interpretation, current concerns about false positives in research reports, and will have time to work on application to research plans. This session is intended to complement but not require the research design workshop.

Goals

  • Understand when to use the most common statistical tests (t-test, ANOVA, Correlation)
  • Understand the difference between significance level, power, effect size
  • Understand how to minimize false positive results
  • Practice aligning statistical tests to research questions

Program

1:30 Welcome

1:40 Survey participants for areas of research interest

1:45 Statistical testing

Important concepts (variability, significance level, power, effect size)

Common statistical tests (t-test, ANOVA, Correlation)

Integrating design and statistical testing

2:30 Small group practice

2:30-3:00 Comparing tests within a group

3:00-3:30 Comparing groups

3:30-3:45 Group discussion of observations and potential pitfalls in making inferences from statistical results

3:45 Discussion of possible statistical approaches to education research of interest to the group (e.g., how to test the effectiveness of a teaching idea)

4:00 Complete workshop survey

4:15 Adjourn

Resources

Statistical tests: http://statpages.org/#Comparisons

2 sample test: http://www.evanmiller.org/ab-testing/t-test.html

Correlation link 1: http://www.fon.hum.uva.nl/Service/CGI-Inline/HTML/Statistics/Correlation_coefficient.html

Correlation link 2: http://scistatcalc.blogspot.co.uk/2013/10/pearson-correlation-calculator.html

Chi-square: http://statpages.org/ctab2x2.html

Sample size: http://www.sample-size.net/sample-size-means/

some useful files:

scatter plots of correlation from zero to one (PowerPoint 2007 (.pptx) 596kB Jul8 15)

chance of not getting .05 over multiple tries (Comma Separated Values 581bytes Jul8 15)

Excel sheet to visualize correlations (Excel 2007 (.xlsx) 36kB Jul8 15)

Some random numbers from 0-100 (Comma Separated Values 431bytes Jul11 15)


      Next Page »