by Jeff Meyer, MPA, MBA

We’ve looked at the interaction effect between two categorical variables. Now let’s make things a little more interesting, shall we?

What if our predictors of interest, say, are a categorical and a continuous variable? How do we interpret the interaction between the two? [click to continue…]


May 2018 Member Webinar – Adjustments for Multiple Testing: When and How to Handle Multiplicity

A research study rarely involves just one single statistical test. And multiple testing can result in statistically significant findings just by chance.
After all, with the typical Type I error rate of 5% used in most tests, we are allowing ourselves to “get lucky” 1 in 20 times for each test.  When you figure out the probability of Type I error across all the tests, that probability skyrockets.
There are a number of ways to control for this chance significance. And as with most things statistical, determining a viable adjustment to control for the chance significance depends on what you are doing. Some approaches are good. Some are not so good. And, sometimes an adjustment isn’t even necessary.
In this webinar, Elaine Eisenbeisz will provide an overview of multiple comparisons and why they can be a problem.

Read the full article →

When to Use Logistic Regression for Percentages and Counts

One important yet difficult skill in statistics is choosing a type model for different data situations. One key consideration is the dependent variable.

For linear models, the dependent variable doesn’t have to be normally distributed, but it does have to be continuous, unbounded, and measured on an interval or ratio scale.

Percentages don’t fit these criteria. Yes, they’re continuous and ratio scale. The issue is the boundaries at 0 and 100.

Likewise, counts have a boundary at 0 and are discrete, not continuous. The general advice is to analyze these with some variety of a Poisson model.

Yet there is a very specific type of variable that can be considered either a count or a percentage, but has its own specific distribution.

Read the full article →

Why ANOVA is Really a Linear Regression, Despite the Difference in Notation

When I was in graduate school, stat professors would say “ANOVA is just a special case of linear regression.”  But they never explained why. And I couldn’t figure it out. The model notation is different. The output looks different. The vocabulary is different. The focus of what we’re testing is completely different. How can they […]

Read the full article →

What Is Specification Error in Statistical Models?

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? The basic idea is that when you choose a final model, you want to choose one that accurately represents the real relationships among variables. There are a few common ways of specifying a linear model inaccurately.

Read the full article →

What Is Regression to the Mean?

Have you ever heard that “2 tall parents will have shorter children”?

This phenomenon, known as regression to the mean, has been used to explain everything from patterns in hereditary stature (as Galton first did in 1886) to why movie sequels or sophomore albums so often flop.

So just what is regression to the mean (RTM)?

Read the full article →

April 2018 Member Webinar: Equivalence Tests and Non-Inferiority

Statistics is, to a large extent, a science of comparison. You are trying to test whether one group is bigger, faster, or smarter than another.
You do this by setting up a null hypothesis that your two groups have equal means or proportions and an alternative hypothesis that one group is “better” than the other. The test has interesting results only when the data you collect ends up rejecting the null hypothesis.
But there are times when the interesting research question you’re asking is not about whether one group is better than the other, but whether the two groups are equivalent.

Read the full article →

What Is an Exact Test?

Most of the p-values we calculate are based on an assumption that our test statistic meets some distribution. These distributions are generally a good way to calculate p-values as long as assumptions are met. But it’s not the only way to calculate a p-value. Rather than come up with a theoretical probability based on a distribution, exact tests calculate a p-value empirically. The simplest (and most common) exact test is a Fisher’s exact for a 2×2 table. Remember calculating empirical probabilities from your intro stats course? All those red and white balls in urns? The calculation of empirical probability starts with the number of all the possible outcomes.

Read the full article →

Poisson or Negative Binomial? Using Count Model Diagnostics to Select a Model

How do you choose between Poisson and negative binomial models for discrete count outcomes? One key criterion is the relative value of the variance to the mean after accounting for the effect of the predictors. A previous article discussed the concept of a variance that is larger than the model assumes: overdispersion. (Underdispersion is also […]

Read the full article →

Link Functions and Errors in Logistic Regression

I recently held a free webinar in our The Craft of Statistical Analysis program about Binary, Ordinal, and Nominal Logistic Regression. It was a record crowd and we didn’t get through everyone’s questions, so I’m answering some here on the site. They’re grouped by topic, and you will probably get more out of it if […]

Read the full article →