Sawtooth Software: The Survey Software of Choice

Is there a workaround in SSI Web so we can ask Dual-Response None questions on a 5-point scale instead of the 2-point Yes/No buy scale.

The use of the 5-point Dual-Response scale has been discussed at previous Sawtooth Software events by both Jon Pinnell and Kevin Karty. The benefit of using a 5-point scale, these researchers have described, is to be able to leverage the common 5-point purchase intent scale (1=Definitely Would Not Buy, 2=Probably Would Not Buy, 3=Might or Might Not Buy, 4=Probably Would Buy, 5=Definitely Would Buy) and to be able to adjust CBC’s None parameter to reflect a “Buy” intention based on top box only (5 on the scale) or top two box (4 or 5 on the scale). For more info, see Kevin Karty’s (with Bin Yu) 2012 Sawtooth Software Paper entitled, “Much Ado About Nothing” (2012 Conference Proceedings.)

Yes, it’s possible to do this in CBC software, and here's the workaround:

  1. Create a standard CBC study using SSI Web (without a None). Add a Select-Type question on the same page beneath each CBC question. That Select-Type question is a 1-5 likelihood of purchase question.
  2. Collect and download the data.
  3. Create a copy of your SSI Web study using File + Save Study As. Let's call that new study Study2. Edit Study2 to be a CBC Dual-Response None format.
  4. Still working with Study2, "trick" the software into thinking you are using paper-and-pencil to collect the data. To do this, click Field + Create Paper & Pencil Interviews. As long as you don’t change anything (other than requesting Dual Response None) about your CBC design, it will use the same experimental design to create the paper-and-pencil files as it had used to collect the data.
  5. Using File + Data Management…, export your Respondent#, CBCVersion#, CBC responses, and answers to the 5-point scale questions to a .csv file. You'll be feeding these back yet into Study2, but telling the software that these were actually collected using paper-and-pencil.
  6. Put your .csv file you exported in step 5 in the right format. Using Study2, click Field + Create Accumulated Data Template File... This creates a .csv file with the correct labels in the first row so that SSI Web understands the paper-and-pencil data you'll be giving it.
  7. Copy your respondent data into the paper-and-pencil .csv template file, rearranging the data in the proper columns according to the data template file you created in step 6. You'll be supplying Respondent#, version#, answers to the CBC questions, and also the Yes/No buy responses (but you’ll need to recode the 5-point scale to a 2-point buy/no buy scale). Here’s when you decide what the cutoff is on the 5-point scale for buy vs. no buy. For example, if you think that top box only (5) is a "buy", then you will recode your 5-point scale so that any ratings of 5 become a "1" (buy) and any ratings 1-4 become a "2" (no buy).
  8. Import the paper-and-pencil .csv file data you created in step 7 into Study2, by clicking Field + Accumulate Paper & Pencil Data... When you click the Get Data button, the software will immediately flag that have duplicate records, since Study2 is a copy of Study1 and has all the data still. It will ask you what you want to do about the duplicate information. Change the option to “Combine existing respondent data with new respondent data” to add the new Dual-Response None responses to the existing data and click Continue.
  9. Then, proceed with analysis with Study2 as usual, as if you had originally asked a Dual-Response None.

Perhaps we'll add this as an option in the next version of CBC, so you don't have to go through the steps of this workaround.