By Gursagar Singh · AgriLab
In this activity, students identify the experimental unit, check block order in a spreadsheet and explain why a data-entry error changes the ANOVA without changing treatment means. The nine observations are illustrative leaf counts from three treatments in three complete field blocks.
Before running the analysis
| Treatment | Block 1 | Block 2 | Block 3 |
|---|---|---|---|
| A | 9 | 15 | 21 |
| B | 13 | 17 | 25 |
| C | 15 | 23 | 27 |
Mark the unit that received a treatment. Then calculate treatment B’s mean. Explain what must stay consistent when you copy the three columns into the calculator.
Run the two cases and compare results
- Open the prefilled calculator and select Run Analysis. Record the original-data error mean square and 5% critical difference (CD). Compare the original and square-root CDs in the comparison table.
- Reverse only the first and last readings for treatment B, making that row 25, 17, 13. Run again and record what changes. Explain why treatment B’s mean stays the same.
- Restore the correct order. Explain which means you would compare with the square-root CD and what you would record about the transformation in a thesis.
Answer key
Each treated plot is an experimental unit. Treatment B’s mean is 18.33 in both cases. Rearranging one row changes which readings the model assigns to each block and increases the residual variation.
| Run | Error mean square | CD at 5% |
|---|---|---|
| Correct order, original data | 1.3333 | 2.6177 |
| B reversed, original data | 49.3333 | 15.9226 |
| Correct order, square-root data | 0.02088 | 0.3276 |
The reversed-row result illustrates the data-entry error. Use the correctly ordered observations for analysis. Compare transformed means with the transformed CD and report the exact square-root formula used.