https://communities.sas.com/t5/Statistical-Procedures/When-to-use-Proc-reg-and-when-to-use-proc-glm/m-p/301440#M16034 <P>If you're new to SAS and statistics, the first SAS Statistics course, which covers regression and analysis is free here, bottom left corner.&nbsp;</P> <P><A href="http://support.sas.com/training/tutorial/" target="_blank">http://support.sas.com/training/tutorial/</A></P> Thu, 29 Sep 2016 05:51:43 GMT Reeza 2016-09-29T05:51:43Z https://communities.sas.com/t5/Statistical-Procedures/When-to-use-Proc-reg-and-when-to-use-proc-glm/m-p/301438#M16032 <P>Hello,</P><P>&nbsp;</P><P>I am a bit confused about when to use <STRONG>proc reg</STRONG> and when to use <STRONG>proc glm</STRONG>. I want to find out the factors effecting Exists.</P><P>&nbsp;</P><P>Below is my data table</P><P>&nbsp;</P><TABLE><TBODY><TR><TD>Bounces</TD><TD>Exits</TD><TD>Continent</TD><TD>Sourcegroup</TD><TD>Timeinpage</TD><TD>Uniquepageviews</TD><TD>Visits</TD><TD>BouncesNew</TD></TR><TR><TD>0</TD><TD>0</TD><TD>OC</TD><TD>(direct)</TD><TD>18</TD><TD>1</TD><TD>0</TD><TD>0</TD></TR><TR><TD>0</TD><TD>0</TD><TD>N.America</TD><TD>(direct)</TD><TD>4</TD><TD>1</TD><TD>0</TD><TD>0</TD></TR><TR><TD>0</TD><TD>0</TD><TD>N.America</TD><TD>Others</TD><TD>35</TD><TD>1</TD><TD>0</TD><TD>0</TD></TR><TR><TD>0</TD><TD>0</TD><TD>N.America</TD><TD>public.tableausoftware.com</TD><TD>70</TD><TD>1</TD><TD>0</TD><TD>0</TD></TR><TR><TD>0</TD><TD>0</TD><TD>N.America</TD><TD>public.tableausoftware.com</TD><TD>81</TD><TD>1</TD><TD>0</TD><TD>0</TD></TR><TR><TD>0</TD><TD>0</TD><TD>N.America</TD><TD>public.tableausoftware.com</TD><TD>75</TD><TD>1</TD><TD>0</TD><TD>0</TD></TR><TR><TD>0</TD><TD>0</TD><TD>N.America</TD><TD>public.tableausoftware.com</TD><TD>186</TD><TD>1</TD><TD>0</TD><TD>0</TD></TR><TR><TD>0</TD><TD>0</TD><TD>N.America</TD><TD>(direct)</TD><TD>710</TD><TD>1</TD><TD>0</TD><TD>0</TD></TR><TR><TD>0</TD><TD>0</TD><TD>OC</TD><TD>(direct)</TD><TD>712</TD><TD>1</TD><TD>1</TD><TD>0</TD></TR><TR><TD>0</TD><TD>0</TD><TD>AS</TD><TD>Others</TD><TD>344</TD><TD>1</TD><TD>1</TD><TD>0</TD></TR><TR><TD>0</TD><TD>0</TD><TD>EU</TD><TD>Others</TD><TD>27</TD><TD>1</TD><TD>1</TD><TD>0</TD></TR><TR><TD>0</TD><TD>0</TD><TD>EU</TD><TD>visualisingdata.com</TD><TD>0</TD><TD>1</TD><TD>1</TD><TD>0</TD></TR><TR><TD>0</TD><TD>0</TD><TD>N.America</TD><TD>Others</TD><TD>294</TD><TD>1</TD><TD>1</TD><TD>0</TD></TR><TR><TD>0</TD><TD>0</TD><TD>N.America</TD><TD>public.tableausoftware.com</TD><TD>111</TD><TD>1</TD><TD>1</TD><TD>0</TD></TR><TR><TD>0</TD><TD>0</TD><TD>SA</TD><TD>(direct)</TD><TD>1430</TD><TD>1</TD><TD>1</TD><TD>0</TD></TR><TR><TD>0</TD><TD>0</TD><TD>N.America</TD><TD>(direct)</TD><TD>29</TD><TD>1</TD><TD>1</TD><TD>0</TD></TR><TR><TD>0</TD><TD>0</TD><TD>N.America</TD><TD>Others</TD><TD>637</TD><TD>1</TD><TD>1</TD><TD>0</TD></TR><TR><TD>0</TD><TD>0</TD><TD>N.America</TD><TD>google</TD><TD>10</TD><TD>1</TD><TD>1</TD><TD>0</TD></TR><TR><TD>0</TD><TD>0</TD><TD>N.America</TD><TD>google</TD><TD>42</TD><TD>2</TD><TD>1</TD><TD>0</TD></TR></TBODY></TABLE><P>&nbsp;</P><P>&nbsp;</P><P>&nbsp;</P><P>Thanks and regards,</P><P>&nbsp;</P><P>Aditya</P> Thu, 29 Sep 2016 05:41:40 GMT https://communities.sas.com/t5/Statistical-Procedures/When-to-use-Proc-reg-and-when-to-use-proc-glm/m-p/301438#M16032 AdityaKir 2016-09-29T05:41:40Z https://communities.sas.com/t5/Statistical-Procedures/When-to-use-Proc-reg-and-when-to-use-proc-glm/m-p/301439#M16033 <P>What values does exits take on? 0/1?</P> <P>&nbsp;</P> <P>PROC GLM does most of what PROC REG does. There was a comparison a short while ago here:</P> <P><A href="https://communities.sas.com/t5/SAS-Statistical-Procedures/HAS-PROC-REG-been-out-of-date/m-p/285418#M15082" target="_blank">https://communities.sas.com/t5/SAS-Statistical-Procedures/HAS-PROC-REG-been-out-of-date/m-p/285418#M15082</A></P> <P>&nbsp;</P> <P>If your variable takes on the values 0/1 you shouldn't be using PROC REG - use PROC LOGISTIC or GLM. Since you have mostly categorical data, you can also do as already suggested by&nbsp;<a href="https://communities.sas.com/t5/user/viewprofilepage/user-id/18408">@Ksharp</a>&nbsp;and use CATMOD.</P> Thu, 29 Sep 2016 05:49:05 GMT https://communities.sas.com/t5/Statistical-Procedures/When-to-use-Proc-reg-and-when-to-use-proc-glm/m-p/301439#M16033 Reeza 2016-09-29T05:49:05Z https://communities.sas.com/t5/Statistical-Procedures/When-to-use-Proc-reg-and-when-to-use-proc-glm/m-p/301440#M16034 <P>If you're new to SAS and statistics, the first SAS Statistics course, which covers regression and analysis is free here, bottom left corner.&nbsp;</P> <P><A href="http://support.sas.com/training/tutorial/" target="_blank">http://support.sas.com/training/tutorial/</A></P> Thu, 29 Sep 2016 05:51:43 GMT https://communities.sas.com/t5/Statistical-Procedures/When-to-use-Proc-reg-and-when-to-use-proc-glm/m-p/301440#M16034 Reeza 2016-09-29T05:51:43Z https://communities.sas.com/t5/Statistical-Procedures/When-to-use-Proc-reg-and-when-to-use-proc-glm/m-p/301460#M16036 <PRE> proc reg only is suited for continuous variable. proc glm can handle both continuous and category variables. If EXISTS has only two value 0 and 1, then Logistic Regression (proc logistic , proc catmod .... ) or decision tree (proc hpsplit) is a good choice . </PRE> Thu, 29 Sep 2016 08:14:16 GMT https://communities.sas.com/t5/Statistical-Procedures/When-to-use-Proc-reg-and-when-to-use-proc-glm/m-p/301460#M16036 Ksharp 2016-09-29T08:14:16Z