4 Basic Steps to Stats
In much of my research I use mixed methods designs, meaning I combine qualitative and quantitative information. I’ve also been teaching a psych research methods course for the past couple of years in which students design and conduct their own final research projects and questions invariably arise about what stats to run. I encourage my students to follow a similar approach to the one I use when considering analyses for a particular project or to answer a particular question.
Here are my 4 basic steps to stats:
Step 1: Selecting an Analysis
The first, and probably most challenging step involves deciding what analyses to run.
One useful resources for selecting the appropriate analyses is a page created by Anne Marenco, currently at College of the Canyons and formerly of California State Northridge University.
Scroll down the page and Marenco lays out some excellent tables that can help a researcher decide what statistical analyses to use when. At the bottom of the page Marenco writes out a few Q’s and A’s :
“When trying to decide what test to use, ask yourself the following…
Am I interested in…?:
description (association) – correlations, factor analysis, path analysis
explanation (prediction) – regression, logistic regression, discriminant analysis
intervention (group differences) – t-test, ANOVA, MANOVA, Chi square”
Another similar and useful page is UCLA’s What’s Statistical Analysis Should I Use.
I’ve also found reading through comment pages and blogs is a great way to learn from people who are wrestling with or have wrestled with similar questions.
The Analysis Factor is one blog I stumbled on the other day. Judging from a quick read it looked quite active – and the bloggers seemed really responsive to questions posted in the comments section.
Talk Stats is another active blog where a researcher can post and answer stats questions.
Hopefully these resources will be helpful for thinking through which statistical analysis to use. I also suggest talking to people about your thought process. If you have any friends who are familiar with statistics – share your ideas with them and see what they think – or email your professor and see what they think about your general direction and proposed analyses.
Once you’ve decided it’s time RUN the analyses.
Posts on Steps 2, 3 and 4 to come!