- What's the difference between and ?
- How can you have negative covariance? Is this how one cancels systematic uncertainties?
- Is there possible asymmetry in the error matrix, ?
- Is it true that given a set of measurements, we can always find the transformation that removes correlations from the transformed error matrix? Is there any physical/experimental significance to doing this?
- In the small asymmetry case, we get if , . Does this result make sense? It seems that we cannot draw any conclusion from this.
- Why do , obey a Poisson when not fixed, while they obey binomial when is fixed?
- In the binomial treatment, what if ? It seems they should not have the same variance.
- When comparing the binomial fixed asymmetric setup to the Poisson, did we go from Poisson to binomial or binomial to Poisson and show they were equal?
- How is the fixed situation different from the not fixed case?
- Say means uncorrelated. If just by chance you had it come out this way, then you say that are uncorrelated when there might be a physical reason not to believe that. How does this make sense physically?
- In an experiment, how would you determine if two variables are correlated?

Kate Scholberg 2014-11-10