I set up a research plot that had three subplots of control, three subplots of cut and poison, and three subplots of cut, posion, and planting of spicebush. Within each subplot, I sampled along a transect using a quad. Within each quad, I recorded the percentage cover of honeysuckle, spicebush, wood, grass, leaf, and shade coverage. I would like to look into the differences between the control and two treatments. I would also like to look at the effect of shade coverage on the ground cover. Here is what my data looks like. There were a lot of zeros so I transformed it using log(x+1). I attached a screenshot of what my data looks like.
You can use permanova, but there is no replication so your results cannot be generalised and applied to other systems
In which statistical program you would analyze your Data. I have some experience with R.
Therefore, please let me know if R is the program that you are using to provide you wth some codes to start with different models to estimate which model would show better normality of the residuals of your data.
I also would like to know if you checked the normality of your data after log transformation to avoid trying different models if the analysis of the log-transformation already solves the problem of non-normality in your data (if the residuals are already not-normally distributed).
try firsly to see how you data are distributed (normality test) it will be oriented you in parmetric or no-parametric test to do.
secondly, seeing that you have 0 in your data set, you can try to see these effect without log transformation.
I agree with Chimi Djomo Cédric 's answer! You must first perform a normality test. If the population follows a normal distribution law, you can perform an analysis of variance (ANOVA), and compare the means in case of significance using the Duncan test. If not, you must perform a non-parametric test (chi-square test). Given the "0" in some columns, it is likely that you will perform non-parametric tests for some parameters.
- Badges
- Science topic
- Similar topics
- Correction