I am currently optimizing the acquisition time of an experimental Physics setup. To achieve our goal, we need to be able to properly estimate the error bars of each of the 200-time domain points of an interferometric signal obtained with a single-photon detector and then fit this signal to a model using non-linear least squares.

The problem is that, while I can now repeat the measurements hundreds or maybe even thousands (which would take us around 3 days at least) of times to study and get a proper estimate of the noise, in a 'real' experiment we would be able to only perform two quick measurements before changing the experimental conditions. What would be the proper way for me to tackle this problem?. Right now I have multiple 100-measurement samples that I can study, so I was thinking of maybe bootstrapping with N=2 to estimate the standard deviation of a two-measurements signal, but I am not sure if this will work. Any help or suggestion will be very appreciated. Apologies in advance for my rusty English and Statistics skills.



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.