I have to do the statistical analysis of experimental data from an experiment which was done in the same way (replicated) in two different sites, with contrasting soil properties, where soil properties were determined in 5 different epochs . The experiment have the following levels: 2 soil; 4 treatments; 5 sampling epochs.
Which is, in your opinion, the best method to statistically analyze experiment conducted in different (contrasting) soils?
This is a simple RBD factorial experiment having two different environments (soil type or locations) as variables.
Be aware of some things.
I do not know how many repetitions you have, but the simplest way is to face that you have 2 experiments, one at each site (i guess the site is the factor soil), because probably your 3 factors were not properly randomized in the field. If you do not randomized the factor ephocs in the field along with your treatments, maybe you are dealing with a split-plot design in time. So you could do a split-plot ANOVA separed for each site and then compared the sites with ANOVA Conjoint Analysis. Or you could try a Mixed Model (GLMM), and model the site as random.
Am suggestilng to use Anova for the variabipity of the soil properties and t-test for the difference of soil properties between the two soil type. Pls can you highlight your objectives so that the most appropriate statistical techniques can be suggested
You can analyze the results according to a design in complete blocks at random, in this way you can evaluate and weigh the variability of the results due to the type of soil used and the variability that is really due to the treatments applied.
- Badges
- Science topic
- Similar topics
- Agriculture
- Soil Science
- Soil