How to fit a dataset into a theoretical distribution?

Question

I want to fit a dataset into a theoretical distribution to see the best-fit modes. I tried Analyse > Distribution > Continuous Fit > Fit all. I can get the AIC and BIC values, but not the Anderson–Darling values. Essentially, I want to produce a table like the one below. Can anyone help?

Distribution

Anderson–Darling

AICc

BIC

Rank

Lognormal

0.78

552.1

556.8

1

Gamma

1.04

556.9

561.5

2

Weibull

1.32

560.7

565.4

3

Normal

4.51

590.3

594.8

Poor

Mark_Bailey · Answer

What you want is not directly available. It will take some work on your part.

After selecting Fit All, press and hold the Control key on Windows or the Command key on Mac, click the red triangle next to one of the fits (e.g., Fitted Normal Distribution), and select Goodness of Fit. This command will produce the A2 statistic for each fit. Note that this procedure is compute-intensive and might take a couple of minutes to complete. Now right-click on one of the A2 tables and select Make Combined Data Table. That is the second half that you are missing. Go back to the top of the fits, right-click on the table with the AIC values and select Make into Data Table. This is the first half. Now use the Tables > Join command to merge them.