Analysis of Variance (ANOVA) for On-farm Trials

Interpretation of trial results is an important part of agricultural research. It is important to understand the reasons for differences in the production means (averages) between treatments in replicated trials.

Growers are well aware of the variation in production across the farm, between paddocks and even within paddocks. This variation needs to be factored into the analysis of any trial data to determine whether the differences in treatment means are really significant, as a result of the treatment/s, or just anomalies that can be expected as a result of normal variation.

This interpretation of trial results requires complex statistical software to sort out what is natural variation in the production system, and what is an actual difference between the treatments. There are many software packages capable of doing this analysis - some are expensive and complex to use. The data for simple trials can be analyzed using the statistical analysis package contained in Excel, part of the Microsoft Office Suite.

The calculator below can be used to analyse data from a Randomised Block Designed Experiment with 3 treatments and 3 replications (3x3) (note: The calculator can not be used for any other design, and it must contain a complete data set i.e. 9 entries). The trial must have 3 treatments (including the control) and 3 replications, with the treatments randomized in each replication block. If your data set does not match this requirement, or is missing a value due to some unforeseen problem, don't despair, the data can be analysed with more complicated software. You will need to contact researchers or agronomists familiar with the software.

Input Data

TreatmentBlock 1Block 2Block 3
Designed by Arris Pty Ltd