Stats for

Start From

Your Research Question

Find the question you're asking and pick from the methods that answer it. Then check each one's “This Is Not Your Test If…” before you commit.

Compare Groups or Conditions

Compare a Group to a Known Value

Does a group's average score, typical score or share of yes answers differ from a guideline, norm or published figure?

Narrow These Down →

Compare Two Groups

Do two independent groups differ on average, or does one tend to score higher?

Narrow These Down →

Compare Two Conditions in the Same People

Do the same people score differently in two conditions or at two times?

Narrow These Down →

Test Whether One Effect Depends on Another

Does the effect of one factor change with the level of another (an interaction)?

Narrow These Down →

Compare Proportions or Categories

Do groups or conditions differ in how often each category or outcome turns up?

Narrow These Down →

Look at a Relationship

Explain How an Effect Happens

Does one variable relate to another through a third, intervening variable?

Narrow These Down →

Test Whether a Relationship Depends on a Third Variable

Is the link between two variables stronger, weaker or reversed for people who score higher on a third (a moderator)?

Predict an Outcome

Account for Clustered Data

How do predictors at each level relate to an outcome when people come in groups, such as students in classrooms or patients in clinics?

Narrow These Down →

Predict a Yes/No or Category Outcome

Which variables predict whether something happens, or which category someone ends up in, and how much does each change the odds?

Narrow These Down →

Check a Measure's Quality

Check That a Measure Measures What It Should

Do a measure's scores reflect the construct it's meant to capture, and not other constructs?

Narrow These Down →

Test a Structure Named in Advance

Do a set of measures reflect the underlying traits a theory says they should, in the pattern it predicts?

Evaluate the Items on a Test

How well does each item on a test or scale work, and which ones need revising or dropping?

Narrow These Down →

Find the Structure behind a Set of Items

How many underlying dimensions do a set of items reflect, and which items go together?

Evaluate a Screening or Diagnostic Test

How well does a test, questionnaire or cutoff pick out the people who have a condition, judged against a reference standard?

Narrow These Down →

Describe or Explore Data

Describe Your Data

What's a typical score, how spread out are the scores, and what shape do they take?

Check a Distribution's Shape

Do the scores follow a named distribution, such as the normal curve?

Find Groups of Similar People

Do people fall into groups, or profiles, on a set of measures, and what does each group look like?

Model Change over Time

Model Change over Time

How does an outcome change across repeated measurements of the same people, and do people differ in how they change?

Narrow These Down →

Model How People Move between States

How likely are people to move from one category to another from one occasion to the next, and where do they tend to end up?

Other Ways to Start