I calculated LOQ and LOD for ELISA but it seems that the LOQ smaller than LOD. According to EPA (1996) and Armbruster & Pry (2008), LOQ cannot be lower than LOD. Where did I could possibly go wrong with this?
Craig Silver 80% of my ELISA data is non-detectable. I've tried establishing the LOD and LOQ by Excel Data Analysis using Regression. I can't pinpoint at which step did I mess up.
I use the same metrics for qPCR, LOB, LOD, and LOQ. On some situations you can have an LOD that is equivalent to an LOQ, but typically LOQ is higher because you have more faith in the data and validation data.
Craig Silver Thank you for the link. Btw, are you familiar with maximum likelihood estimation? I've tried using normal distribution but apparently it didn't fit my data. So I tried Weibull's distrib. However, my average value is negative so that too didn't work out. Is there any alternative distribution that I can use?