topic Re: NPAR1WAY, SURVEYSELECT and Kolmogorov-Smirnov two-sample tests on weighted data in Statistical Procedures https://communities.sas.com/t5/Statistical-Procedures/NPAR1WAY-SURVEYSELECT-and-Kolmogorov-Smirnov-two-sample-tests-on/m-p/173036#M8976 <HTML><HEAD></HEAD><BODY><P>Not sure a weighted K-S test exists. Is there a reference describing such a test? - PG</P></BODY></HTML> Sun, 24 Aug 2014 18:00:52 GMT PGStats 2014-08-24T18:00:52Z NPAR1WAY, SURVEYSELECT and Kolmogorov-Smirnov two-sample tests on weighted data https://communities.sas.com/t5/Statistical-Procedures/NPAR1WAY-SURVEYSELECT-and-Kolmogorov-Smirnov-two-sample-tests-on/m-p/173035#M8975 <HTML><HEAD></HEAD><BODY><P>In need to compare two distributions using NPAR1WAY and two-sample K-S tests,but one of them is weighted.&nbsp; If I set FREQ to the weight I get the correct cumulative distribution for the weighted data, but NPAR1WAY calculates the p-value incorrectly.&nbsp; It thinks the number of entries in the cumulative distribution is the sum of the weights, whereas it is much lower (thus the p-values are too low).&nbsp; Given the D-statistic, which I think SAS calculates correctly from the two cumulative distributions, I believe I can recalculate the p-value from the correct numbers of entries in the two distributions.</P><P></P><P>Is there a way to get NPAR1WAY to correctly calculate the p-value?&nbsp; Problem is I have to do this for 400 different pairs of distributions!</P><P></P><P>Can I somehow use SURVEYSELECT to resample the weighted distribution to get an unweighted distribution having the original number of observations?&nbsp; E.g. if the unweighted data set has 10K entries, and the sum of the weights is 200M, can SURVEYSELECT produce a data set with 10K entries that reproduces the weighted sample cumulative distribution?</P></BODY></HTML> Sat, 23 Aug 2014 20:06:21 GMT https://communities.sas.com/t5/Statistical-Procedures/NPAR1WAY-SURVEYSELECT-and-Kolmogorov-Smirnov-two-sample-tests-on/m-p/173035#M8975 ewolin 2014-08-23T20:06:21Z Re: NPAR1WAY, SURVEYSELECT and Kolmogorov-Smirnov two-sample tests on weighted data https://communities.sas.com/t5/Statistical-Procedures/NPAR1WAY-SURVEYSELECT-and-Kolmogorov-Smirnov-two-sample-tests-on/m-p/173036#M8976 <HTML><HEAD></HEAD><BODY><P>Not sure a weighted K-S test exists. Is there a reference describing such a test? - PG</P></BODY></HTML> Sun, 24 Aug 2014 18:00:52 GMT https://communities.sas.com/t5/Statistical-Procedures/NPAR1WAY-SURVEYSELECT-and-Kolmogorov-Smirnov-two-sample-tests-on/m-p/173036#M8976 PGStats 2014-08-24T18:00:52Z Re: NPAR1WAY, SURVEYSELECT and Kolmogorov-Smirnov two-sample tests on weighted data https://communities.sas.com/t5/Statistical-Procedures/NPAR1WAY-SURVEYSELECT-and-Kolmogorov-Smirnov-two-sample-tests-on/m-p/173037#M8977 <HTML><HEAD></HEAD><BODY><P>The weights are essentially predicted frequencies, and I use them as such.&nbsp; They are based on a full survey design and using the weights/frequencies when plotting a variable should give a distribution that is close to what one would get if one sampled the entire US civilian population where every sample had weight equal to one.</P><P></P><P>The K-S test should work fine, I just want to find a way to get SAS to calculate the p-value correctly.&nbsp; It gets the d-statistic correct, I believe.</P></BODY></HTML> Sun, 24 Aug 2014 18:07:05 GMT https://communities.sas.com/t5/Statistical-Procedures/NPAR1WAY-SURVEYSELECT-and-Kolmogorov-Smirnov-two-sample-tests-on/m-p/173037#M8977 ewolin 2014-08-24T18:07:05Z