Homework Assignment 8

Introduction to Analysis of Covariance: Week 8 Introduction to Analysis of Covariance: Week 8 Program Transcript DR. ANNIE PEZALLA: We've done a lot thus far in this class. We've learned \ about moderation and mediation. We've covered logistic regression, facto\ rial ANOVA's, and ANOVA's with repeated measures.

You're becoming increasingly well-equipped to answer more sophisticated research questions to determine mean differences across groups. But what\ if you want to test for the effects of an independent variable while controllin\ g another variable? That's what the test this week is all about.

We'll be learning about ANCOVA's this week, which stands for analysis of covariance. ANCOVA, simply put, is a one way ANOVA that adds a continuou\ s variable as a control. Let's again return to our statistics professor an\ d his students as an example.

This professor may wish to understand if there are gender differences in asking for help. But he's not so naive to think that gender alone predicts help\ asking behavior. He's well aware that his students come to his classes with dif\ fering levels of comfort with math. So what he wants to do is essentially contr\ ol for math comfort, and focus on the role of gender above and beyond the explanator\ y power of math comfortability in explaining the variance within help seek\ ing behavior.

So why isn't a one-way ANOVA appropriate for such a research question? W\ ell, a one -way ANOVA does not control for the confounding aspect of the level of math comfortability a student has coming into the course. With ANCOVA, t\ he professor can administer a pretest at the beginning of the term, and now\ control for previously acquired statistics knowledge when comparing means.

In this week, you'll practice applying ANCOVA to given data sets, and on\ ce again analyze quantitative research articles that apply this method. But first\ , let's watch Dr. Jones demonstrate this test in SPSS.

Introduction to Analysis of Covariance: Week 8 Additional Content Attribution FOOTAGE: