Help please

[deleted_user]

Hello SAS Gurus,

I'd really appreciate your help. I am doing a stat analysis of an injury data which comes from a probability sample of hospitals. I was told that in order to calculate national estimates i have to use PROC SURVEYMEANS, PROD SURVEYFREQ AND PROCLOGISTIC procedures. I managed doing SURVEYMEANS but having hard time figuring out SURVEYFREQ. So my questions:
1 - how do i calculate odds ratio usinf SURVEYFREQ;
2 - how to interpret SURVEYFREQ output. It gives me a bunch od chi-sguare tests that i've never heard of (e.g., Rao-Scott).

I'd relly appreciate you help. Thank you in advance.

[sbb]

Recommend you visit the SAS support website http://support.sas.com/ and do a search on the phrase "how calculate odds ratio using surveyfreq" and I expect you will find technical papers (SUGI/SGF) on this topic, as I found. Or a more general Google advanced search, using the search argument below will also yield reference documents (remove the site: parameter if you want to include other websites):

how calculate odds ratio using surveyfreq site:sas.com

Scott Barry
SBBWorks, Inc.

[deleted_user]

Well people... call me slow but i cannot find how to calculate odds ration in SURVEYFREQ using SAS 9.1 version. Only version 9.2 comes up when i try to search it. Could anybody help me please???? I'd greatly apprecaite your help.

[sbb]

Sorry, not my expertise but some crafty searches (including SAS OnlineDoc):

Topic: odds ratios or odds ratio estimates

SAS OnlineDoc - What's new in SAS/STAT 9.0 and 9.1 mentions:

SURVEYLOGISTIC Procedure


The SURVEYLOGISTIC procedure performs logistic regression on data that arise from a survey sampling scheme. PROC SURVEYLOGISTIC incorporates complex survey sample designs, including designs with stratification, clustering, and unequal weighting, in its estimation process. Variances of the regression parameters and odds ratios are computed using a Taylor expansion approximation. The SURVEYLOGISTIC procedure is similar in syntax to the LOGISTIC procedure, and it can fit link functions such as the logit, cumulative logit, generalized logit, probit, and complementary log-log functions. Maximum likelihood estimation of the regression coefficients is carried out with either the Fisher-scoring algorithm or the Newton-Raphson algorithm.


Google advanced search against SAS.COM - use this argument to find related matches, some of what are not 9.2 related, including the SUGI/SGF paper link:

+"odds ratio" +9.1.3 site:sas.com

U.S. Health and Nutrition: SAS® Survey Procedures and NHANES
http://www2.sas.com/proceedings/sugi31/140-31.pdf


Usage Note 30333: FASTats: Frequently Asked-For Statistics
http://support.sas.com/kb/30/333.html#oddsratios

[deleted_user]

Thanks, i will check it out!