```Date: Wed, 1 Jul 2009 11:42:53 -0500 Reply-To: Joe Matise Sender: "SAS(r) Discussion" From: Joe Matise Subject: Re: Survey Question Categorization Comments: To: "Mathur, Rajat" In-Reply-To: Content-Type: text/plain; charset=ISO-8859-1 On the first part, I would just use the statistical technique of 'comparison'. IE, if X in (6:7) then X_SC = 1; else if X in (1:5) then X_SC = 0; Then take the mean of that variable to get a frequency. I couldn't say what to do with the three-part variable, as it depends on what you're doing later with that. If you want to get a score of '60%' where 60% = 50% highly likely, 20% somewhat likely, 30% not likely, 50%+ (1/2)*(20%), then you could use 1,0.5,0. If you are wanting to know the proportions of each of those, either use any recode (say, 2,1,0) and do FREQ, or spread it to two or three variables (X_VL, X_SL, and if needed X_NL). -Joe On Wed, Jul 1, 2009 at 11:26 AM, Mathur, Rajat < Rajat.Mathur@diamondconsultants.com> wrote: > Hi Group, > > I am working on survey data and all the questions are on likert scale 1 to > 7 [1 as least important and 7 as most important]. I now need to decide a > cutoff on this scale so that I can create a binary variable, out of such > questions > > For example: > Question: How often to you visit XYZ type of website. > Possible Responses: 1-Least often, 2,3,4,5,6, 7-Very often > > Now can I use any statistical technique to figure out - if response is > greater than 6 (or any other number cutoff) then responders are frequent > visitors and below the cutoff then responders are not frequent visitors. > > Also, if instead of two, I had to break it into three bands (very frequent, > moderate frequent, least frequent) then what technique can I use? > > Any help would be deeply apprecited... > > Regards, > Rajat > > ________________________________ > This transmission may (i) be subject to contractual, statutory or other > obligations of confidentiality; (ii) contain "protected health information " > as defined in HIPAA or "non-public personal information " as defined under > data privacy laws; or (iii) be otherwise strictly confidential. If you are > not the intended recipient of this message, you may not disclose, print, > copy or disseminate this information. If you have received this in error, > please reply to the sender only and delete the message. Unauthorized > interception of this e-mail is a violation of federal law. Diamond's > operating subsidiary in North America is Diamond Management & Technology > Consultants NA, Inc.; in the U.K. and Europe, Diamond Management & > Technology Consultants Limited; and in Asia, Diamond Management & Technology > Consultants Pvt. Ltd. > ```

Back to: Top of message | Previous page | Main SAS-L page