Questionnaire Design Best Practices Guide

vSurveys+™: Best Practices Design Guide

This article will provide best practices for designing surveys that interact efficiently with behavior data enabled through the Studio Sandbox.

Prior to MFour Studio, attitudinal data collection methods and behavior data sets were rarely integrated. Survey researchers dealt with attitudinal data while data scientists handled behavior data analysis separately. However, with MFour Studio, the integration of these data sets is now possible. Survey researchers must now learn to factor behavioral data analysis needs when designing surveys. Creating a questionnaire for analyzing behavioral data does not require a drastic change in approach. To facilitate the analysis of data through Studio Sandbox, consider these helpful tips for designing an effective questionnaire.  

Tip #1: Identify your questions for behavioral data segmentation

Selecting appropriate questions is critical to creating meaningful segments or audiences for analysis. These questions may pertain to demographics, opinions, or stated behaviors. By proactively determining and choosing the right questions, researchers can achieve clear distinctions between respondents and gain valuable behavioral insights. Adequate planning can also ensure that the questionnaire includes the necessary questions to define the desired segments.  

Tip #2: Keep response options simple and differentiated

It is recommended to keep response options for key segmentation questions simple and clearly differentiated. Simple response groups allow for easy differentiation between attitudinal segments, making it easier to view behavior differences. To efficiently analyze cohorts, MFour Studio experts suggest limiting multiple-choice or segmentation groups to five or fewer options.  

Don’t do this:

What is your favorite color:

1. Fire engine red
2. Dragon’s breath red
3. Rose red
4. Light green
5. Grass green
6. Mean green
7. Teal green
8. More blue than green
9. Light blue
10. Baby blue
11. Blue blue
12. ….Are you getting the point?

This is better:

Which of these is your favorite color?

1. Red
2. Green
3. Blue 
4. Black

Tip #3: Create Collapsed Variables (CVs) for Segments from Multiple Questions

While MFour Studio survey design and analysis Sandbox is a highly effective tool, it does have certain limitations. For instance, it does not allow the addition of customized segments built from single or multi-question response stubs once the survey has begun fielding.  So “Collapsed Variables” (or, “CVs” as we call them) must be created during survey programming to make these segments available in the data set.

Collapsed Variables are used to define categories of survey data in the background. A collapsed variable can be used to collect data into groups or buckets.

Example: I want to see a segment for behavior data analysis built from the “Strong” or “Lukewarm” ACME fans who said they were “very interested” in the product.

Q1: Are you a fan of this ACME Brand?

1. Strong
2. Lukewarm
3. Not at all
4. Never heard of

Q2: How interested or uninterested in the ACME ROCKET BACKPACK+ROLLER SKATE ROADRUNNER CAPTURE KIT are you?

1. Very interested
2. Somewhat
3. Not really
4. Not at all

“Target Market” CV = Q1.1 or Q1.2 AND Q2.1

Note: Some attitudinal questions types cannot be analyzed in the Sandbox at this time.  These include open ends, heat maps and other question types that do not require a single selection by a respondent. Additionally, questions collected through  “matrix” style format can be viewed in the Sandbox, but presentation is challenging.  Consult your support team during survey design about your needs.

Be sure to coordinate with your Research Consultant or Project Manager on your banner plan and custom variable needs during the design and programming phase. 

Tip #4: Keep it short and to the point

This is more of a best practice for survey design overall - rather than one specifically for vSurveys+™: Keep it simple!

Effective surveys should ideally take no more than 10 minutes to complete. Surveys that exceed 30 minutes can cause respondent fatigue, resulting in less useful feedback after the 10-minute mark.