Become a better data scientist

Sign up to our newsletter

No thanks

Statistical tests – interactive tool

By Data Tricks

Which statistical test?

Choosing which statistical test to use can be confusing, with a seemingly endless list of options depending on whether you have interval or nominal data, paired, unpaired or parametric. The interactive tool on this page has been developed to help you choose an appropriate statistical test.

How to use this tool

The tool asks a set of questions about your objective and your data. Select the answer that best matches your problem and the tool will recommend an appropriate statistical test. Click the link at the end for instructions on how to perform the test or calculation in R.
Instructions are being updated regularly so please bear with us if the instructions for a particular test are not available.

What are you assessing?

Differences between samples

I have split my data into groups and want to compare those groups

Relationships between variables

I am looking for links between variables

How many samples do you have?

One sample

I have a single sample and I want to see if the average is different from a fixed value

Two samples

I have two samples and want to see the differences between them

Multiple samples

I have more than 2 samples and want to see the differences between them (note if your independent variable has more than 2 levels then you effectively have >2 samples)

What's the measurement scale of the variable you want to test?

Interval

Your scale has an order and known differences between the values, eg. cm to measure distance, or height

Ordinal

Your scale has an order but different distances between values, eg. a rank order

Nominal

Your measure has no quantitative value, eg. gender

What is the distribution of your data?

Parametric

Non-parametric

Are your samples paired?

Yes

My samples are paired

No

My samples are not paired

What's the measurement scale of the variable you want to test?

Interval

Your scale has an order and known differences between the values, eg. cm to measure distance, or height

Ordinal

Your scale has an order but different distances between values, eg. a rank order

Nominal

Your measure has no quantitative value, eg. gender

What is the distribution of your data?

Parametric

Non-parametric

What's the measurement scale of the variable you want to test?

Interval

Your scale has an order and known differences between the values, eg. cm to measure distance, or height

Ordinal

Your scale has an order but different distances between values, eg. a rank order

Nominal

Your measure has no quantitative value, eg. gender

What is the distribution of your data?

Parametric

Non-parametric

Are your samples matched?

Yes

My samples are matched

No

My samples are not matched

What's the measurement scale of the variable you want to test?

Interval

Your scale has an order and known differences between the values, eg. cm to measure distance, or height

Ordinal

Your scale has an order but different distances between values, eg. a rank order

Nominal

Your measure has no quantitative value, eg. gender

What is the distribution of your data?

Parametric

Non-parametric

What's the measurement scale of the variable you want to test?

Interval

Your scale has an order and known differences between the values, eg. cm to measure distance, or height

Ordinal

Your scale has an order but different distances between values, eg. a rank order

Nominal

Your measure has no quantitative value, eg. gender

What is the distribution of your data?

Parametric

Non-parametric

Do you have dependent and independent variables?

Yes

There are clearly defined dependent and independent variables

No

The variables are interdependent

What's the measurement scale of your dependent variable?

Interval

Your scale has an order and known differences between the values, eg. cm to measure distance, or height

Ordinal

Your scale has an order but different distances between values, eg. a rank order

Nominal

Your measure has no quantitative value, eg. gender

How many categories does your nominal dependent variable have?

Two

More than two

What's the measurement scale of your variables?

All are interval

All variables are on an interval scale

Interval and/or ordinal

I have interval and/or ordinal variables but no nominal variables

All are nominal

All my variables are nominal

What is the distribution of your data?

Parametric

Non-parametric

How many categories does your nominal variable have?

Two

More than two

Result

Binomial logistic regression

Result

Multinomial logistic regression

Result

Ordinal linear regression

Result

Repeated measure ANOVA

Result

One-way ANOVA

Result

Friedman test

Result

Cochrane Q

Result

Kruskal-Wallis test

Leave a Reply

Your email address will not be published. Required fields are marked *

Please note that your first comment on this site will be moderated, after which you will be able to comment freely.

You might also like