PHARM/PCS V5 · Procedure 35 ANOVA and Comparisons
Analysis of Variance III: Two-Way, with Replication
Test two factors and their interaction, with several observations in every combination, to see whether the factors act independently.
What it computes
Two-factor ANOVA with r replicates per cell. Tests for treatment effects, block effects, and interaction.
Inputs and outputs
You provide
- Replicate values for each combination of treatment and block, entered in a grid organized by block and treatment
PHARM/PCS reports
- F ratios for treatments, blocks, and their interaction, each marked significant or not
- An interaction plot, where non-parallel lines signal an interaction
When to use it
When two factors act together and you have multiple observations per combination, so you can ask not only whether each factor matters but whether they interact, the question Procedure 34 can't answer.
Reading the results
Read the interaction term first: if it's significant, the factors' effects aren't independent, and interpreting either main effect alone can mislead, since the story lives in the combination. Only when interaction is absent do the main effects stand cleanly on their own.
The design at a glance
The entry grid is organized by block and treatment, so the two-factor structure is visible as you type, and the interaction plot shows at once whether the lines run parallel. Results match the 1986 program.
This page describes the procedure. Calculations run inside the licensed PHARM/PCS V5 application. PHARM/PCS is a trademark of MCS Scientific LLC.