I have a problem with finding the right weight method in sigmoidal curve fitting. As background information: I measure different compounds to create dose response curves. Every concentration will be measured three times and the resulting activity data will be fitted by a sigmoidal curve fit. My Problem is: What is the right weighting method for fitting this data? Can I use the "no weighting" parameter? Or must I use statistical weighting or instrumental weighting. Normally I would say I should use the instrumental weighting. Are their any arguments for or against these weighting methods? I would be very happy to hear some aspects from you :)
In your case you don't need to use weight and each data point should be treated equally. You use instrumental weighting when you only have the information of the mean and uncertainty of measurements. While here you already have the original data in which case running fitting with no weighting is fine. Thank you.
Finding the right weighting method
Posted - 09/18/2018 : 05:15:41 AM
