Bias in Marketing Research: How to Avoid and Minimize

Author’s Note: I’ve always been interested in biases. A poster of The Cognitive Bias Codex hangs in one of our home offices; and, I’ve bookmarked and often look at that interactive wiki to learn more. Just like I looked to two industry experts to learn more about bias in research for this article: Jeff Henning and Ray Poynter. Many thanks to you both for serving GreenBook as expert sources. We value your contribution to this piece, just as we value our readers.

Have you ever asked a question while assuming you already knew the answer? That’s a simple example of a bias. As humans, we inadvertently experience bias in every aspect of our lives. Our personal biases influence the lens through which we look at the world. And while some biases are innocuous, some are harmful; and, bias in market research can have a negative effect on the findings.

What is Marketing Research Bias?

Research bias is defined as any type of feedback that skews the market research’s results. There are five types of market research bias to look out for:

1. Social desirability bias – This bias boils down to research participants thinking they should answer questions in a way that paints them in a positive light. For example, if surveying on gender discrimination, no one wants to admit they have misjudged someone solely based on their gender.
2. Habituation – This type of bias can be avoided if survey questions are well-formulated and differentiated. Habituation bias is when participants begin giving similar answers to similar-sounding questions.

   Related

   Food for Thought: Challenges and Ideas for Addressing Bias

3. Sponsorship bias – Most likely one of the more common types of biases, results are skewed in preference of the organization paying for the market research. Popular examples of this often appear in the food industry, such as The Coca-Cola Company funding obesity research.
4. Confirmation bias – This type of bias occurs when the researcher intentionally or unintentionally seeks to “confirm” their hypothesis by crafting leading questions or cherry-picking data.
5. Culture bias – When researchers analyze data and make assumptions through their own unique lens of their upbringing, heritage, and community, this is cultural bias.

What is the Most Prevalent Type of Marketing Research Bias?

“The most prevalent bias is the ‘say/do gap’ – people saying they will do things that it turns out they don’t do.”

Experts also have differing opinions regarding the most prevalent types of bias in research. According to Roy Poynter, Chief Research Officer at Platform One, “The most prevalent bias is the ‘say/do gap’ – people saying they will do things that it turns out they don’t do.” For Jeff Henning, Chief Research Officer at Researchscape International, the most prevalent bias boils down to a nonresponse. Other opinions state confirmation bias is the most common type.

How to Reduce and Avoid Marketing Research Bias

Why even bother reducing market research bias? First, it’s problematic because it often skews results. It’s important to critically and independently collect and report on data. Second, strong biases make it difficult to come to a true scientific conclusion that can be repeated and fact-checked by third parties. Researchers should do everything they can to proactively minimize and avoid bias.

4 Ways to Minimize Marketing Research Bias

There are specific ways to avoid each specific type of market research bias. Howe, it’s important to keep your personal perspective in check in both quantitative and qualitative research.

1. Follow Random Sampling Best Practices

Avoid bias in your sampling pool by ensuring each element has an equal chance of representation. Use randomization tools to select an assortment. For digital research, work to avoid bots with layered fraud mitigation techniques.

2. Pay Attention to How Questions Are Worded

Asking the wrong questions can lead you to the wrong answers. Since survey questions are written by humans, it’s inevitable that we subconsciously put our personal perspective into our language. Use open-ended questions to avoid steering subjects in a specific direction. Henning said, “Researchers control the questionnaire, which can often add bias. Studying questionnaire writing or using an expert minimizes instrument bias.”

3. Use a Survey Method That Makes It Easy for All Types of People to Participate

For example, if you are surveying how many commuters walk to work, and you only stand out in the street with a clipboard, you inherently miss everyone who drives or bikes to work. Poynter explained, “We need to use approaches that have been developed to be robust in the face of bias, and we need to triangulate and take different readings from different points to get a more predictive result.”

4. Use an Agile Research Method

A scalable, agile research method allows for innovation and a customer-specific focus. With continuous feedback, a minimal viable product, and sprint work, research can be human-centric and easily adaptable to changes. Researchers can get more mileage out of current efforts and focus on other initiatives simultaneously.

Bias Isn’t Completely Avoidable – But It’s Mitigable

Research will never completely avoid bias, but market research experts can get better and better at spotting bias before it influences results. An agile approach can help quickly find, solve, and document biases so they’re avoided in the future. Less bias means more accurate results, which furthers your field of research and leads to true breakthrough.