When we think about model assumptions, we tend to focus on assumptions like independence, normality, and constant variance. The other big assumption, which is harder to see or test, is that there is no specification error. The assumption of linearity is part of this, but it’s actually a bigger assumption.
What is this assumption of no specification error? (more…)
When interpreting the results of a regression model, the first step is to look at the regression coefficients. Each term in the model has one. And each one describes the average difference in the value of Y for a one-unit difference in the value of the predictor variable, X, that makes up that term. It’s the effect size statistic for that term in the model.
(more…)
If you analyze non-experimental data, is it helpful to understand experimental design principles?
Yes, absolutely! Understanding experimental design can help you recognize the questions you can and can’t answer with the data. It will also help you identify possible sources of bias that can lead to undesirable results. Finally, it will help you provide recommendations to make future studies more efficient. (more…)
There is a lot of skill needed to perform good data analyses. It is not just about statistical knowledge (though more statistical knowledge is always helpful). Organizing your data analysis, and knowing how to do that, is a key skill. 
(more…)
It’s easy to develop bad habits in data analysis. When you’re new to it, you just don’t have enough experience to realize that what feels like efficiency will actually come back to make things take longer, introduce problems, and lead to more frustration. (more…)
classes, and colleagues and journal reviews question your results because of it. But there are really only a few causes of multicollinearity. Let’s explore them.Multicollinearity is simply redundancy in the information contained in predictor variables. If the redundancy is moderate, (more…)