Understanding Statistics Plots

Statistics becomes much easier to interpret if you choose the metric and plot type intentionally.

The same data can look very different depending on whether you view counts, proportions, or density.

Choose the metric first

Start by deciding what quantity best matches the biological question.

Common options include:

• Total Cells for overall counts
• ON Cells for positively classified cells
• % ON for proportion of positive cells
• Cells/mm2 for overall density
• ON Cells/mm2 for positive-cell density

In practice:

• use counts when absolute abundance matters
• use % ON when relative composition matters
• use density when region area or coverage should be taken into account

Heatmaps

Heatmaps are useful when you want a broad overview across many animals and regions at once.

They are especially good for spotting patterns such as:

• one region standing out across the cohort
• one animal looking different from the rest
• region clusters that behave similarly

Heatmaps are best for pattern recognition, not for precise reading of small numeric differences.

Mirrored plots

Mirrored plots are useful when you want direct comparison between grouped summaries.

They work well when the main question is comparative, for example:

• control vs treatment
• one condition vs another
• male vs female

This plot type is often the clearest choice when you want to compare the same regions across two sides of a contrast.

Anatomical representations

Anatomical representations are useful when spatial interpretation matters most.

They show where a selected metric is concentrated in anatomical space, which helps connect summary values back to brain structure.

This is often the most intuitive plot when the key question is:

Where in the brain is the effect strongest?

Grouping changes the level of interpretation

Statistics can group values by metadata such as:

• group
• condition
• sex
• combinations of these

This changes the level at which you are comparing the data.

For example, grouping by sex can reveal differences that are hidden when all animals are pooled together.

In contrast, grouping too finely can make the display sparse and harder to interpret.

In practice, start simple and only add more grouping when it helps answer a specific question.

Read differences carefully

A visual difference in a plot is not always a biologically meaningful difference.

Before interpreting a strong pattern, ask:

• Is the metric the right one for this question?
• Is the comparison based on the subset I actually intended to analyze?
• Does the pattern still make sense when I check the underlying Results?

A practical approach

A reliable workflow is:

1. pick one biological question
2. choose the matching metric
3. choose the plot that best supports that question
4. cross-check surprising findings against the underlying Results

That usually leads to more defensible interpretation than trying to read every possible pattern from one plot.