Date:         Mon, 12 Nov 2007 10:22:35 -0600
Reply-To:     Mary <mlhoward@avalon.net>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Mary <mlhoward@AVALON.NET>
Subject:      Re: What test to use?
Comments: To: Sharon <bssr.wedding@GMAIL.COM>
Content-Type: text/plain; charset="iso-8859-1"

Hi, Sharon,

You could compare the response rate of the two methods, and see which one is higher.  You could also look at the number of questions that were responded to- often people will stop a phone questionaire part-way through- so it would be interesting to see if more questions were answered on one method than another.    If you really want to get sophisticated, you could use use logistic regression or discriminate analysis to try to distinguish the differences between the two methods, but you ought to have formal training on these techniques, either through your courses or through SAS Institute courses (Categorical Data Analysis, and Multivariate Analysis), to do this.

-Mary
  ----- Original Message ----- 
  From: Sharon 
  To: SAS-L@LISTSERV.UGA.EDU 
  Sent: Thursday, November 08, 2007 9:10 PM
  Subject: What test to use?


  Hi,

  I'm a student and I am conducting an experiment.

  I have one survey but administered it in two ways. One way was through
  the web and the other way was through the phone.

  I have a response rate for each type of method.

  Which test should I use to test if one method was better than the
  other?