Inference Flowchart

Trying to decide between a 2-sample t-test and a Chi-Square test? This 90-second AP Stats review covers every inference procedure you need for the 2026 exam. The math is only half the battle—the real challenge is the logic behind choosing the right test. In this video, I break down a proven, step-by-step flowchart method to help you pick the correct procedure every single time without overthinking it.

Don't let "brain fog" ruin your score. Follow this simple path and clear the confusion before exam day!

Grab you FREE Inference Cheat Sheet below.

Transcript

The hardest part of the AP exam is not the math, it is the logic. Knowing which inference procedure to choose is critical. I am going to show you a proven method for how to pick which inference procedure.

The first thing you want to ask is whether the question is about the slope of a regression line. If it is, then it will be a linear regression t test or a linear regression t interval.

If it is not about the slope of a regression line, then the next question is whether there is a table of frequencies. If there is a table of frequencies, then you know it has to be a chi-square test.

Next, you ask how many groups or samples there are. If there is just one group or sample, then you ask how many variables are involved. If there is one variable, it is a chi-square goodness-of-fit test. If there are two variables, it is a chi-square test of independence.

If there is more than one group or sample, then it is a chi-square test for homogeneity.

If there is not a table of frequencies, then you ask whether the data involve proportions or means. If it is proportions, then you ask how many groups or samples. One group means a one-proportion z test or z interval. Two groups means a two-proportion z test or z interval.

If it is not proportions but means, then you again ask how many groups or samples. One group means a one-sample t test or t interval. Two groups means a two-sample t test or t interval.

That is it. By asking these few questions, you can always identify the correct inference procedure. Do not overthink it on the test. Just follow the path.