We never intended for people using Discover to transfer their data into Lighthouse Studio (previously known as “SSI Web”), but there are two ways to go about this. The first way I’ll describe below assumes you really do want to move the data into Lighthouse Studio. The second way is easier and assumes you are going to use the standalone versions of CBC/HB, CBC Latent Class, and the new standalone Choice Simulator (you are licensed for all the standalone versions of these software systems, not integrated within Lighthouse Studio; if you have a CBC license within Lighthouse Studio, you just need to download the standalone versions of the software from the Downloads area of our website).
To copy Discover CBC data into a new Lighthouse Studio (previously known as “SSI Web”) CBC project:
1. Click “Survey Data”
2. At the right-hand menu, click “Advanced” and then click “Export CBC Data”. A mystudy.csv file is downloaded. The mystudy.csv file is a comma-separated file with the following columns: RespNum, Task, Concept, Attribute1, Attribute2 (etc.), Answer. You can open this file in Excel. It contains only respondents who finished all the CBC tasks.
Now, you need to modify this file. If you included a “None” concept in your survey (most researcher’s do), then you need to delete the row representing the final concept from each choice set (it’s the concept with 0s for the attribute levels). Use the sorting capabilities of Excel to sort the file by concept # and then to delete all these rows representing the None concept. Then, use the sorting procedure again within Excel to resort the rows by RespNum then by Task then by Concept (a multi-level sort) to get the file back into the proper order. Then, delete the final Answer column. Next, you need to renumber the respondents so they go in sequence: 1, 2, 3, etc. through as many respondents as you have (note this procedure will only work if you have 999 or fewer respondents, because that’s the maximum number of design versions that Lighthouse Studio’s CBC can generate). Save the modified mystudy.csv file.
3. Click “Download Data” and a studyname_data.csv file is downloaded. This file is also a comma-separated file and has a number of columns in it. Typically, you are only interested in the columns: RespNum, mystudy_Choice1 through mystudy_Choicen (where n is the total number of tasks in your CBC study). And, you are only interested in respondents who have completed all the CBC tasks. Delete any respondents (rows) from this .CSV file for respondents who did not complete all CBC tasks. Also delete the not needed columns in this file (unless you want to keep some variables for segmentation and/or weighting purposes, then don’t delete the segmentation and weighting variable columns!) and save the file as a new name called studyname_choices.csv (you will use this in a later step below). And, you need to renumber the respondents as you did with the previous file so that the run from 1, 2, 3 etc. through as many respondents as you have. (Make sure you didn’t change the sort order of the completed respondents before renumbering so that the respondents match up between the two files you have been modifying!).
After completing these two steps, you have two .CSV formatted files: mystudy.csv which contains the information about what product concepts the respondents saw in each choice task and studyname_choices.csv which contains the answers that each respondent (who completed the CBC tasks) gave to each choice task (plus any segmentation or weighting variables if you think you’ll need them within Lighthouse Studio). Also, you’ve renumbered the respondents to go in sequence for both files from 1, 2, 3, etc. without skipping any successive integer.
We are going to “trick” Lighthouse studio into thinking that it fielded this study via “paper-and-pencil”.
(Continued in next comment, because this Forum is limited to 8000 characters per post)...