I have data relating to the presence or absence of electrolytes in water, and the volume of water drunk. I need to do an analysis to see if electrolytes can be calculated based on the volume drunk.
So far, I've decided that binary regression is what needs to happen, as the Y response is either 0 or 1. Volume of water will be the X1 variable.
I've said that the model is ln (ODDS) = y(hat) / 1 - y(hat) = b0 + b1X1
To perform the analysis, I went to analyse -> regression -> binary logistic
The SPSS output gave p -value for the constant of 0.393 and for volume the p-value was 0.376.
My questions are: Is my model correct? Since my p-values are not statistically significant, do I end here, or is there further testing to be done?
Many thanks for any help you can offer.