Example: Random distribution for concept tests (2 of 3 concepts)

Use nested groups to show some, but not all versions of a concept of a product design.

Tip: Random distribution is one of several methods for achieving even distribution in surveys. For more information, see Quick comparison: Random distribution, Least Filled concept testing, even distribution scripts.
Example: Concept test

You want to show participants two of three new concepts of a print ad for an upcoming promotion. After the survey has collected data, you will want to see which concept participants saw first and second.

To do this, you will need to create two hidden single choice questions that will capture the order the concepts appeared to participants. Then you will need to create three paths in your survey. Each path will contain one of the three concepts; however, participants will only see two concepts that are displayed in random order.

Record the concepts the participant will see

Start the survey with two hidden Single Choice questions to capture which concepts the participant sees and the order they appeared in the survey. In each question, the answer options are a list of the concepts.

  1. Add a Single Choice question to the survey Table of Contents, and name it First Seen.
  2. Click Add Answers in Batch, and type the following answer values:
    • Concept A
    • Concept B
    • Concept C
  3. Under Additional Options, select Hide question.
  4. Duplicate the question named "First Seen" and rename it Second Seen.

Create a path for each concept

Now you are going to use a group to show each concept to participants at random. Within this group you will create three nested groups that contain a highlighter question that shows an image of the concept.

After you distribute your survey, the application will select a path at random and show the participant the concept contained in the group. After the participant has answered the question, the application will select another path at random and show the concept contained in the group. When the participant has seen two concepts, the application skips the remaining concept and moves to the next question in the survey.

  1. Add a Group to the survey Table of Contents and name it Random.
  2. Add three nested groups and name them:
    • Path 1
    • Path 2
    • Path 3
  3. Open the group named "Random":
    • From the Display objects in list, select Random order.
    • From the Number of objects to display list, select 2.
      Tip:
      • The Number of objects to display is disabled until you make a selection from the Display objects in list.
      • You are selecting 2 for the purposes of this example. However, depending on how many concepts you want to show, any setting can work as long as it's not All.

Set the value in the hidden questions

Next you will use a Set Value action to set the value for the hidden Single Choice questions named "First Seen" and "Second Seen".

  1. Add two Set Value actions to the group "Path 1" and name them:
    • First Seen: Path 1
    • Second Seen: Path 1
  2. Open the Set Value action named "First Seen: Path 1", and use the Has no value operator to set the value for the question named "First Seen" to Concept A.

    After you distribute your survey, the application will set the value to the question named "First Seen" to the first concept the participant sees.

    Example set the value in the hidden questions.

  3. Open the Set Value action named "Second Seen: Path 1", and use the Has a value operator to set the value for the question named "Second Seen" to Concept A.

  4. Add a Highlighter question to "Path 1" and name it Concept A.
  5. Repeat steps 1-3 for "Path 2" and "Path 3".
    • Use "Path 2" to set the values in "First Seen" and "Second Seen" to Concept B.
    • Use "Path 3" to set the values in "First Seen" and "Second Seen" to Concept C.
    Tip: To save time, duplicate the first path, and edit the group names and Set Value actions accordingly.
  6. Check your overall survey structure one more time.

    The survey structure should look similar to the images above.

    • There should be one top-level group that contains all the subgroups.
      • In this top-level group, Random order is set. (If you expect a sample size of less than 50 responses, choose Rotated order.)
      • The Number of objects to display is set to something other than All.
    • There should be one subgroup for each concept. Each subgroup contains a concept on its own page.

View the reports

Test your survey and click Report > Standard Report to view which concepts were seen by participants and the order they saw them in. For further analysis, you can compare a member's profile data with the order they saw the concepts in the survey. For more information, see Subgroups.

Sample reporting view:

Sample reporting view with subgroups: