Kantar's Profiles Blog

Are your surveys healthy?

Posted by Joris Schellekens on Sep 6, 2017
Do I want to know? Yes - you need to know!

Today’s researcher is all about gathering data to create insights that provide guidance and strengthen the core business, future-proofing decision making for their company. In truth that only covers the ‘need-to-know’ part; there is a large share of ‘nice-to-know’ insights as well that are gathered for various reasons (e.g., to verify general thoughts or just to satisfy the board, directors, managers and other influential people within organizations). Need-to-know and Nice-to know insights are both based on data.

Let’s have a conversation about the term ‘data’;

Joris: “Above probably makes sense. But if you look more closely, what is data?”

You: “[insert your thoughts]”

Joris: “Great. Let’s talk about survey data only here. What factor would be key if this data would be used as input to run and optimize your business?”

You: “[insert your thoughts]”

Joris: “That’s very interesting indeed. I think I heard you mentioning something about ‘quality’ also - correct? So, what is quality data? The market research industry always tells our research participants that there are no wrong or false answers! But how wrong can we be? Yes, there are false answers.

You: “Why Joris? We are asking for their opinions and thoughts?” – Sorry, I filled the gap here for you.. ;-)

Joris: “Yes, but the wrong answers would be considered those answers that are untruthful or just-click-through answers.”

If you use untruthful data, you or your clients are ultimately being misguided in the business decisions that you are basing off of these facts and figures. So, what can we do to avoid this and drive ‘quality’ into the right direction?

What all companies with panels (should) do:

  • (1) Eliminate bots by preventing them to enter the pool of respondents that would potentially enter your survey (via PII data)
  • (2) We should verify if unique people who want to enter your potential pool of respondents are validated (IP & ID check)

Something useful to add to the bone would be to use our patented ‘lie-detector test’ for all potential research participants:

  • (3) Use our Honesty Detector to validate if the potential respondents that we can invite have the tendency to over-report (deliver untruthful data). Basically what we do is analyze panelists’ responses to a unique combination of high and low probability statements, as well as a benchmark question to identify these over-reporters. Next, we would remove the people from our panel that are not likely to provide credible responses

After we re-assured that our potential research participants have the intention to participate in a credible way, we also need to make sure that we help them by transforming this ‘good’ intention into truthful answering questions that deliver us the quality data that we are keen on for creating guidelines and input for our business;

  • (4) The buzz word here is “engagement”. Participants need to be able to participate in the easiest and most enjoyable way, without losing sight on the actual goals of the survey

We have a fantastic group of experienced professionals in our company. Also in the MR market place in general there are great professionals and experts that can help getting to a ‘quality and engaging survey’.

The MR professionals can help with (1) formulating the right questions and methodologies (foundation of getting the right data and a nice flow).

The more technical research professionals can help with (2) optimizing your survey to be compatible for all devices (less dropout, better accessible, get to the hard to reach audiences and in general the next steps into the future – mobile).

The creative research professionals can help with (3) creating an appealing look & feel and potentially integrating gaming techniques for an enjoyable and ‘challenging’ experience (less dropout and more engagement and focus on questions).

Above seems like a lot of ingredients for a great cocktail. One that guarantees optimized quality data, right? Yes, the ingredients are there, but “NO”, you probably don’t know if you are indeed engaging enough! The key here is getting data to support that you (with or without other experts) created a survey that resulted in engaged participants as mentioned in point 4. Also, wouldn’t it be terrific if you had data about perceived quality of the survey from research participants to compare your survey to? Norms and standards per country that are based on real data of all participants within a certain period of time in a certain country?!

All the valuable opinions of all the people who actually participated in all the different research studies ,yours and others, all over the globe and on all possible devices, have been captured during the past year. This is the foundation and real data to determine how “healthy” your study is in relation to the buzz word “engagement." I invite you to take a deep dive together with me into this wonderful world of need-to-know-data and see if we can find ways for monitoring and improving the health of your survey and therefor try to further improve the needed quality of data. 

Topics: Honesty Detector, Data Quality, Survey Health Score

Subscribe to Email Updates

Recent Posts

Posts by Topic

see all