What is response bias, and how can you reduce it? 

Response bias is a form of survey bias that occurs when respondents answer questions in ways that don't accurately reflect their experiences, opinions, or behaviors. It can happen consciously or unconsciously, and it can skew results even when a survey is well designed. 

Who participates, how options are phrased and ordered, and even how respondents want to describe themselves can all influence what people say. That matters for consulting and investment teams, where survey findings may inform recommendations, test investment hypotheses, support due diligence, and evaluate value-creation decisions. Understanding the different ways response bias can creep into a survey makes it easier to identify potential problems and reduce the effect they may have on the results. 

What are common types of response bias? 

Social desirability bias 

This type of bias occurs when respondents give answers they believe will be viewed more favorably by others or that align with perceived social expectations. For example, an employee may be reluctant to give negative feedback about their workplace if they believe the responses could be traced back to them. Respondents may also unintentionally give answers that reflect how they see themselves, such as overestimating how much they exercise. 

Acquiescence and dissent bias 

Acquiescence bias occurs when respondents tend to agree with statements or select positive responses regardless of what they actually believe. Dissent bias is the opposite: when respondents tend to disagree or select negative responses. 

Survey fatigue can make both biases more likely, as respondents may stop paying attention to questions and rush to the end of the survey. 

Central tendency and extreme response bias 

Central tendency bias, also called neutral responding, occurs when respondents repeatedly choose the midpoint of a scale rather than expressing a preference. They may be reluctant to take a side or simply looking for the quickest way through the survey.  

Extreme response bias is when respondents consistently choose the highest or lowest options without considering the choices in between. 

Here’s what that looks like in practice: On a five-point Likert scale, central tendency bias may show up as a respondent repeatedly selecting 3, while extreme response bias would show up as repeatedly selecting 1 or 5. 

Recency bias 

Recency bias occurs when respondents give more weight to recent experiences than to the bigger picture. For example, a customer might give a vendor a low NPS rating because of an order mix-up last week, even after years of otherwise excellent service. The most recent experience can loom larger than everything that came before it. 

Position bias  

Position bias occurs when the order or placement of response options influences which answers respondents select. For example, respondents may be more likely to choose options that appear first or last in a list, giving those choices an advantage that has nothing to do with what respondents actually think. 

Favoring options at the top of a list is known as the primacy effect. Favoring those at the bottom is the recency effect. 

(Note: Don’t confuse the recency effect with recency bias, described above.) 

Non-response bias 

Non-response bias is not technically a type of response bias, but it's closely related and can have a similar effect on survey results. It occurs when people who should be represented in the research don’t provide a response because they never participated, were incorrectly removed by the screener section, or left the survey before completing it. If those respondents differ in meaningful ways from those who complete the survey, the results may not represent the target population.  

Frequently asked questions about response bias 

Why does response bias matter? 

Response bias can skew survey results, misrepresent trends, and lead to poorly informed decisions. Even when the data looks plausible, it may not accurately represent what people think, want, or do. 

What is the difference between survey bias and response bias? 

Survey bias is a broad term for factors that can skew survey results. Response bias is one type of survey bias that specifically affects how respondents answer questions.  

Can response bias be eliminated completely? 

No. Response bias can’t be eliminated entirely, but thoughtful survey design can reduce the opportunities for it to show up.  

How do you reduce response bias? 

• Protect respondents’ privacy: When possible, keep responses anonymous or confidential and explain those protections clearly. Respondents may be more willing to answer candidly when they know their individual responses won’t be shared. 

• Write neutral questions: Avoid leading, loaded, or emotionally charged language that could steer respondents toward a particular answer. 

• Keep the survey concise: Ask only the questions needed to meet the research objectives. Long surveys can lead to fatigue, rushing, and less thoughtful responses. 

• Randomize answer choices: Randomizing the order of response options can help reduce position bias. Keep options in a logical order when sequence matters, such as with rating scales. 

• Avoid double-barreled questions: Ask about one thing at a time. Combining two topics in a single question can leave respondents unsure which one they’re answering. 

• Use balanced rating scales: Include an equal number of positive and negative response options and, when appropriate, a neutral midpoint. Make sure each point on the scale is clearly defined. 

• Pilot the survey: Test the survey with a small group before fielding it more broadly. This can uncover confusing wording, problematic response options, or other issues while there’s still time to fix them. 

• Screen the data: Good survey design can reduce response bias, but questionable responses can still slip through. Look for straight-lining, unusual response patterns, and unusually fast completion times to identify responses that may require closer examination. 

Related resource: IncQuery’s guide to writing effective survey questions. 

Reducing response bias in consulting and investment research 

A completed survey isn’t necessarily a successful survey. Response bias can produce perfectly plausible results that don’t accurately reflect what respondents think, want, or do. Taking steps to address potential bias while the survey is still being designed helps ensure the findings provide a solid foundation for whatever comes next. 

‍