Is there any sort of analysis that can be done, or methods such as bootstrapping that would work with a low degrees of freedom?
Running into a problem with a clinical study where the actual data gives a degrees of freedom that is too low (below the threshold) while inflating the number of cases would give a higher degrees of freedom and more satisfactory results.
Is there any way to work with limited data beyond simply increasing the sample size?
This is also a doctor-patient design (so not fully crossed).
Is there any sort of analysis that can be done, or methods such as bootstrapping that would work with a low degrees of freedom?
Running into a problem with a clinical study where the actual data gives a degrees of freedom that is too low (below the threshold) while inflating the number of cases would give a higher degrees of freedom and more satisfactory results.
Is there any way to work with limited data beyond simply increasing the sample size?
This is also a doctor-patient design (so not fully crossed).