Ki-Squared Goodness of Fit for Asymmetric Standard Deviations

[kderakhshani]

Hi

To test the fitness of a certain model to some observational data I use ki_squared goodness of fit test. As you might know, this test is appropriate for normal distributions. In fact, one uses the standard formula:

X^2 = Sum [(Oi - Ei)^2/SDi^2]

Oi average of the ith bin
Ei expected (model) value for the ith bin
SDi standard deviation for the ith bin

But all of my data in each bin are somewhat skewed to the right (asymmetric SD).

Is there any approach to evaluate the fitness for this skewed-normal distribution?

Thanks a lot in advance.