The variables I have- vegetation index and plant disease severity scores, were not normal. So, I did log10(y+2) transformation of vegetation index and sqrt(log10(y+2)) transformation of plant disease severity score. Plant disease severity is on the scale of 0, 10, 20, 30,..., 100 and were scored based on visual observations. Even after combined transformation, disease severity scoring data is non-normal but it improves the CV in simple linear regression.
Can I proceed with the parametric test, a simple linear regression between the log transformed vegetation index (normally distributed) and combined transformed (non-normal) disease severity data?
You can just use non-parametric regression in which the predictor does not take a predetermined form, but is constructed according to information derived from the data. It's better since it uses non-transformed data and is therefore more transparent :)
if the vegetation index is NDVI or other, then additional data detailing can be carried out based on 3D visualization, such as the CPMI index: Fig/4
Conference Paper "Computer vision and hyperspectral imagery in orchards and i...
While it is recommended to have normally distributed variables in a linear regression, transformations can be applied to improve normality. However, in cases where the transformed data still does not meet the assumptions of a linear regression, non-parametric regression methods or generalized linear models with appropriate distributions may be better suited. It is important to consider other factors such as linearity, homoscedasticity, and independence of residuals when evaluating the appropriateness of the resulting model.
Why would these variables have to be normal? As far as I understand our problem, a logistic model might do well. You can try it with my software "FittingKVdm", but if you can send me some dat, I can try it for you.
- Badges
- Science topic
- Similar topics
- Filing