Tasks
Most people arrive knowing what they need to do, not what the tool is called. These pages start from the job — what decides the choice, what tends to go wrong, and which two or three tools are worth your time.
Statistics and plotting
Is my treatment different from my control?
Compare two groupsChoosing and running a two-group comparison, and reporting it usefully.
Which of my conditions differ from each other?
Compare three or more groupsANOVA, its non-parametric equivalents, and what to do about post-hoc tests.
Does one measurement predict the other?
Test whether two variables are relatedCorrelation, linear regression, and the difference between them.
Do these categories occur at different rates?
Compare counts or proportionsChi-square, Fisher’s exact test, and when each applies.
What n do I need for this experiment?
Work out how many samples you needPower analysis, done before the experiment rather than after.
I ran thousands of tests — which results are real?
Correct for multiple testingBonferroni, Benjamini–Hochberg, and choosing between them.
How should I show this data instead of a bar chart?
Plot your data so the reader can see itShowing distributions and replicate structure rather than hiding them.
How do I plot fold change against significance?
Make a volcano plotTurning a differential expression table into a labelled figure.
How do I visualise an expression matrix?
Make a heatmapClustered heatmaps, and the scaling decision that changes what they show.
How do I make a diagram of my model?
Draw a schematic or graphical abstractDiagrams and figure assembly, as distinct from plotting data.