Market Research Biases and How to Remove Them

The Association of Qualitative Research based in the UK describes bias as a “systematic error.” The definition suggests that this systematic error is responsible for deviating results from the factual findings of the research.

 

Various biases can impact qualitative and quantitative studies and potentially occur at any research phase, from study design and sampling to data collection and analysis.

 

 

 One important note to consider when discussing market research biases is that they are not dichotomous variables. This condition means that researchers can’t interpret the tendencies simply by asking whether it’s present or not. A 2010 study on the identification and avoidance of research bias suggests that surveyors and researchers must focus on the degree to which the bias affects the research and how it was and how to avoid it.

 

Understanding bias is critical information for any researcher engaged in market research as some degree of bias affects most studies and can potentially distort their findings and conclusions.

 

While this seems like an end-all situation and may prompt one to rethink the study design, market research experts know that biases can find their way in.

 

Before we delve right in, we must get familiar with the human elements that drive said market research biases, namely the respondents and researcher. Today, we’ll explore the primary sources of these biases and what researchers can do to minimize them to prevent your study from inferring skewed results.

 

 

 

Researcher Market Research Biases

 

We’ll begin with researcher biases as they begin at the study design source – the researcher or surveyor. While many forms of researcher biases exist, we focus on those relevant to the marketing field.

 

 

Confirmation Bias

 

Peter Cathcart Wason, an English cognitive psychologist, is responsible for coining the term “confirmation bias” to describe a form of cognitive bias he observed.

 

Also referred to as confirmatory and myside bias, it describes a systematic error that stems from the researcher’s tendency to form hypotheses that agree with their preconceived notions. The researcher then uses their study’s findings to confirm that belief – hence the name.

 

In other words, the researcher sees what they want to see.

 

Confirmation bias is a fundamental and more challenging market research bias to watch for when conducting research. A researcher afflicted by confirmation bias tends to dismiss responses that sway away from their hypothesis.

 

 

From there, the bias extends to other phases of research, including analysis as they use their findings to dispel other hypotheses, severely distorting study conclusions.

 

Since this form of bias stems from a very baseline natural human tendency to filter relevant information that suits their beliefs, it can be hard to eliminate. This is why preventing its emergence is most effective.

 

Researchers need to reevaluate their intentions for the study and change how they perceive respondent answers. They must challenge and overcome any preexisting notions about the research objective and hypotheses. Research advisors often play the third-party observer’s role that can identify confirmation bias in research before it’s complete.

 

 

Culture Bias

 

Culture bias often comes after confirmation bias in social research as it is considered an extension of the latter. Instead of preconceived notions and beliefs, the researcher defaults to make assumptions about other cultures based on his own cultural values.

 

For instance, an American researcher may undermine his Indian respondent’s answers based on the stereotyped understanding of Indian culture that his American cultural beliefs influence.

 

Culture bias falls on the spectrum of ethnocentrism, which is the phenomenon of judging another culture solely based on one’s own culture. To remove culture bias from your study, shifting towards cultural relativity and reassessing your personal assumptions about other cultures.

 

Cultural relativism refers to assessing and understanding an individual’s cultural values based on said individual’s cultural context and not your own.

 

 

Question-Order Bias

 

As a researcher, when you form questions for your responders to answer, you put an idea in their head. The words you choose and how you present an idea prime the respondents’ feelings and thoughts towards the subject.

 

How you structure (or order) your questions can change how one idea changes the respondent’s attitude towards subsequent survey questions. This creates a question-order bias and unknowingly prompts the respondent to choose options or answers that are not completely accurate and answer relatively than objectively.

 

Let’s consider you are conducting brand tracking on smartphone brands. You may ask your respondents to rate two different smartphone brands on a satisfaction Likert scale. Whatever rating they give the first product will be based on what they think about it, but they are more likely to rate the second product on how it compares to the first rather than how good or bad it is.

 

This type of market research bias slips into surveys very easily, and it’s oftentimes unavoidable. One way to minimize this bias is by asking general questions before specific and positive questions before negative.

 

 

Irrational Escalation Bias

 

While observed less frequently than other forms of biases, irrational escalation can be quite damaging for marketers who have made a financial investment in an idea.

 

This bias is similar to confirmation bias in that the researcher has already developed a hypothesis they believe in and dismiss findings and results that override or undermine them. How it differs is that financial investment becomes sunken debt.

 

Instead of changing methodologies based on research findings, they escalate the importance of their own beliefs and convince themselves that the research is wrong.

 

For example, let’s consider that a concept testing study forecasts a new product’s performance in the market. The brand or company might have already rolled out the products and made them available to customers.

 

Now let’s consider that research findings project product failure. Instead of accepting their mistake and cutting their losses by ceasing production, they reject research findings and continue pushing the products, convinced that it might work.

 

This bias is quite strong and often hard to eliminate. The solution for eliminating this bias is to keep an open mind and relying on research before making such critical decisions.

 

 

 

Respondent Market Research Biases

 

Biases also incur on the respondent’s side while they fill out survey forms and answer researcher questions.

 

 

Social Desirability Bias

 

Social desirability comes naturally to most human beings. It has roots in our inherent desire to have people like us. This need creeps into respondents’ subconscious while they answer questions, distorting survey findings.

 

Respondents end up answering questions in a way they think will lead to them being liked. Regardless of your market research format, questions about their personality or sensitive subjects can cause social desirability bias to occur.

 

This bias is often apparent in “would you rather” questions.

 

To prevent social desirability biases, phrase the question with positive affirmations so that the respondents know that it’s okay to answer honestly. Or you can ask it indirectly.

 

For example, instead of asking what they think about a certain situation, ask what a third-party would think. This way, the respondents project their own beliefs without feeling the need to come off as socially desirable.

 

 

Acquiescence Bias

 

Also known as the friendliness bias, this occurs when respondents want tend to say yes or agree with whatever the researcher asks or presents. They believe that whatever idea they are presented with is good.

 

There are several reasons for this bias to occur:

• One of the more obvious ones comes from social desirability. Respondents think that agreeing with or responding positively will get the researcher to like them.
• The second reason could be that the respondent has an acquiescent personality, and they tend to agree with others in most cases. Third, the respondent might agree because they think the researcher or surveyor is the expert and has the right answers.
• And lastly, acquiescence comes with laziness. Whether it’s a long interview/ questionnaire or lack of interest, respondents may choose to acquiesce to get over it.

 

Eliminating this bias requires the researcher to ask questions that ask respondents’ for their opinion instead of presenting them with dichotomous options.

 

 

Sponsor Bias

 

This form of bias can occur easily in market research as it is often sponsored or supported by a notable brand or organization. If respondents are aware of this sponsor, their feelings or beliefs about the sponsor may influence their answers.

 

If keeping that information from respondents for objectivity is not possible, researchers should focus on maintaining a neutral stance or reiterating their independent status.

 

 

Habituation

 

Habituation bias occurs when the respondents’ brain goes into an autopilot mode, and they begin answering similarly-worded or intended questions the same way. This may be due to the sheer length of the survey or how the researcher phrases his or her questions.

 

Instead of paying attention to each question, the respondents glance to find commonalities between statements and give the same answer out of habit.

 

Rephrasing the questions to make them more engaging and creating more concise surveys can reduce habituation.

 

 

Halo Effect

 

We’ve mentioned this particular type of market research bias separately because it can occur on either side. The halo effect can influence both researchers and respondents. It’s easy for either to form an opinion about a product or brand based on one positive attribute.

 

Moderators and researchers may form positive opinions about certain respondents based on one positive answer. Similarly, respondents may respond to a question about a brand or product positively based on one positive quality, disregarding objectivity and non-consideration of other attributes.

 

For researchers, preventing the halo effect on their side requires maintaining neutral opinions about the respondents. To prevent it from occurring among respondents, researchers should ask questions concerning one brand before the second.