Choosing a Statistical Test

Start with the data

The mind map “What analysis test should be used?” starts by distinguishing qualitative, categorical data from quantitative, measurement data. It then considers whether there is one categorical variable or more than one. Use it to follow a path from the kind of data you have toward a possible test.

For another view, the “Data Analysis Flowchart” branches by the type of dependent variable and the independent variables. Comparing the paths can help you practice identifying the details a test choice depends on.

Work through the choices

Next, try the mind map “Choose stats test”. It asks whether data are categorical, whether you are looking for relationships or differences, and whether tests are parametric or nonparametric. Follow the questions in order and note where your choice changes.

The flowchart “Choosing a Statistical Test” also separates tests of difference from correlation. Its branches include independent groups, repeated measures, interval or ordinal data, and nominal data. Use it alongside the mind map to review the same choices in a different format.

Practice with cards and a quiz

The flashcards on when to use statistical tests name tests including chi-square, Mann-Whitney U, Spearman's rank and the independent t-test. Work through them after the maps to practice matching test names with the situations on the cards.

Then try the quiz “Statistics”. Its questions include choosing tests for between-groups and within-groups designs using scale or ordinal data. Return to the maps as you study the choices.

Place the topic in the unit

Choosing a Statistical Test is a section of “Common Statistical Tests (t-tests, Chi-Square, Non-parametric)” in Statistics and probability. The resources here give you ways to practice selecting among the tests named in the unit. Choose a resource and work through a path that is new to you.