Split plot RCBD with repeated measure time and only one level of whole plot

[Adesuwa-34]

Hi,

I am carrying out a research from an already established field experiment (Split plot design RCBD), Grass cultivars = whole plot, 3 levels, and legume cultivars are subplots with 5 levels.

However, I only selected one grass cultivar and two legume cultivars for my current experiment resulting in

Grass cultivar =whole plot, 1 level.

Legume cultivars = subplot, 2 levels.

Sample collection = 2 time points.

Since I am using only one level of grass cultivar, is it okay to proceed with the data analysis using the split plot ANOVA model?  If not, what is the best approach for my statistical analysis?

Thanks.

[SteveDenham]

Take whatever analysis code you were going to use, and remove any reference to the grass cultivar, as it is no longer a factor in the design.

You have two levels for the main effect, and two time points that should be treated as repeated measures on the legumes.

What I worry about is that it looks like you only have 4 measurements (2 legumes, 2 times), so any analysis will be severely underpowered.

SteveDenham

[Adesuwa-34]

Thanks Steve,

It looks like I may have omitted some essential information about my experiment. Here is the total treatment combinations from the experiment.

Grass monoculture

Legume A monoculture

Legume B monoculture

Grass and Legume A mixture

Grass and Legume B mixture.

Samples from each treatment were taken at two time points.

If I omit grass as a factor, will I be able to account for the grass and legume mixture in my experiment? 

[SteveDenham]

I am going to assume you mean you would have the following 4 groups (each measured two times)

Legume A monoculture

Legume B monoculture

Grass and Legume A mixture

Grass and Legume B mixture

So that means you would have 8 measurements in total.  There is no power to detect anything except two or more fold differences.  Are there plots within each of these that were measured?  If so, you may have something that will be aided by an inferential statistical analysis.  But if it is a single measurement, you could present the entirety of your information content in a 4x2 table of the observations.

SteveDenham

[Adesuwa-34]

Thanks Steve, 

Since I used just one grass cultivar. I used a randomized complete block design with repeated measure for my analysis. Thanks