Repeated-measures ANOVA: a worked example

When the same subjects are measured repeatedly — six plants, each measured on days 0, 7 and 14 — the measurements are not independent, and ordinary one-way ANOVA is the wrong tool: it treats 18 numbers as 18 unrelated observations. Repeated-measures ANOVA instead removes the differences between subjects first, then asks whether values change across time — which makes it considerably more powerful, because each plant serves as its own control.

The data

Leaf area (cm²) of six plants tracked over two weeks:

The repeated-measures layout: one row per subject, one column per timepoint — exactly how MinutePlot expects it.

The result

Time: F(2, 10) = 189.5, p < 0.001. The post-hoc letters separate all three timepoints — Day 0 (5.05, c), Day 7 (6.15, b), Day 14 (7.27, a): leaf area grew significantly in each week.

Means across timepoints with letters — all three differ.

Styling and exporting this figure: see Figure styling and Exporting & downloading.

Report it like this

“Leaf area increased over time (repeated-measures ANOVA, F(2, 10) = 189.5, p < 0.001); all timepoints differed from one another (p < 0.05, letters in Fig. 1).”

Common pitfalls

Entering repeated data as if the groups were independent is the classic error — it discards the design’s power and can bias p-values in either direction. In MinutePlot, choose the repeated-measures layout so each row is a subject; the engine then handles the within-subject structure for you.