topic Re: goodness to fit in SAS Studio

goodness to fit

laurenhosking — Tue, 07 Jan 2020 19:58:24 GMT

How would I interoperate this goodness to fit

Re: goodness to fit

laurenhosking — Tue, 07 Jan 2020 20:44:11 GMT

@PaigeMiller @Rick_SAS

Re: goodness to fit

PGStats — Tue, 07 Jan 2020 21:25:55 GMT

From your output, there is no evidence that variable wt isn't normally distributed.

Re: goodness to fit

laurenhosking — Tue, 07 Jan 2020 21:29:11 GMT

thank you! how do you know this from the small table?

Re: goodness to fit

Rick_SAS — Tue, 07 Jan 2020 21:34:52 GMT

You can check a textbook for precise language. What I would say would depend on my audience. If I want to be casual, I would say "the GOF tests indicate that the data are approximately normal with the given mean and standard deviation." If I am talking to a colleague, I might say, "there is not sufficient evidence to reject the null hypothesis at the 0.05 significance level," where the "null hypothesis" is that the data are generated from a N(75.008, 13.925) distribution.

Re: goodness to fit

laurenhosking — Tue, 07 Jan 2020 21:40:51 GMT

Thank you rick. can I see that I can jeject the null hypothesis from the small table below too?

Re: goodness to fit

Rick_SAS — Tue, 07 Jan 2020 21:47:44 GMT

Yes, because the p values for the GOF tests are greater than 0.05.

Re: goodness to fit

PGStats — Tue, 07 Jan 2020 22:07:41 GMT

@Rick_SAS I guess you meant No: you can't reject the null hypothesis that the data is distributed normally.

Re: goodness to fit

Rick_SAS — Tue, 07 Jan 2020 23:44:21 GMT

Right. Thanks, PGStats. To clarify, the OP asked, "can I see that I can reject the null hypothesis from the small table below too?"

But she should have asked, "can I see from the table that the GOF tests FAIL TO reject the null hypothesis."

The answer is yes, you can see that from the table because the p values less than 0.05.