Survey Bias: What Is Response Bias, Its Types, & Preventive Techniques

By ,
Survey Biases

Photo Credit:

In our previous (which was also the first) article of the series, we talked about survey bias and one of its types, Sampling Bias.

For those of you who are joining us for the first time and want to learn from the beginning, here’s a quick refresher about survey bias:

Survey bias refers to any act, intentional or otherwise, that influences the results of the survey unfairly towards or against a certain objective, thing, or person. It’s basically any deviation from the truth. Any act or decision that takes the survey results away from what’s really happening is called survey bias.

Important Disclaimer:

It’s almost inconceivable to create a survey questionnaire that’s free from all kinds of biases. Too many external factors play a role in survey administration, and many of those factors can be out of your control. The goal, therefore, is not to eliminate bias but to minimize it as much as possible.

By understanding what causes bias and ways to minimize it, we can create surveys that deliver the most accurate results and keep the insights as close to reality as possible.

Different Types of Survey Biases:

Survey bias manifests itself in three unique ways:

  1. Sampling bias: When you choose a sample that fails to accurately represent all strata of the target population.
  2. Response bias: When respondents answer questions inaccurately or unfaithfully or under different types of influences or assumptions.
  3. Interviewer bias: When survey creators or interviews consciously or unconsciously try to influence the survey process, resulting in skewed findings.

Different Types of Survey Biases

In today’s article, we’ll focus on Response Bias, and learn how to prevent it and minimize its impact on our surveys and research.

Response Bias

Response bias is any inaccuracy, distortion, or error in the data that is caused due to the way respondents have answered questions. Various factors can lead to respondents answering questions inaccurately. This typically happens for two major reasons:

  • – They don’t want to answer a question correctly.
  • – They do not have sufficient knowledge to answer the question correctly.

Response bias is quite dangerous as it has a huge potential for rendering false results. Relying on these inaccurate findings can cause you to commit resources and efforts to misguided goals and hence create huge losses.

Response bias has several types, based on different factors that cause it.

• Demand characteristic bias

When respondents might try to align their responses to what they think are the purposes of the study or the expectations of the researcher. In other words, they’ll modify their behavior or answers to suit what they believe the researcher wants to hear.

Example: Employees start working harder when they know a productivity survey is just coming around the corner. Their hard work is not due to any real desire to be more productive but just because of the effect of being observed.

• Social desirability bias

Respondents might provide answers that they believe are socially acceptable or desirable, rather than their true opinions or behaviors. This can lead to over-reporting of positive behaviors and under-reporting of negative behaviors.

Example: In a mental health survey, some respondents downplay their anxiety symptoms to present themselves in a more socially acceptable light.

• Acquiescence bias

It is also called yay-saying or nay-saying. Some respondents may have a tendency to agree or disagree with statements regardless of their actual opinions. This bias can occur when respondents are not fully engaged in the survey or don’t carefully consider their responses.

Example: When you conduct an employee engagement survey and most employees seem to be agreeing or disagreeing with almost everything you say. Even to a point where their statements might contradict one another.

• Non-response bias

This bias occurs when certain groups of people are less likely to respond to the survey, due to a sampling error or just the way you have designed or administered the survey. Non-response bias leads to an unrepresentative or underrepresentation data sample.

Example: If a survey is conducted primarily on mobile, it might exclude individuals who are not comfortable with technology, potentially biasing the results.

• Extreme response bias

When respondents provide extreme or exaggerated answers instead of expressing a moderate opinion.

Example: In a survey with rating scales, respondents consistently choose the most extreme answers such as, Strongly Agree or Strongly Disagree, with very few answers with other options.

• Neutral response bias

It refers to a tendency of respondents to consistently choose responses that are in the middle of a response scale. It usually indicates a lack of interest on the respondent’s part.

Example: In a Likert scale survey where respondents are asked to rate their agreement or disagreement on a scale of Strongly Agree to Strongly Disagree, respondents frequently choose the middle option (Neither Agree nor Disagree).

• Cultural bias

When there are questions or other elements in the survey that are not relevant to certain groups of people. It happens when researchers systematically favor certain people or unfairly assume behavior patterns for a diverse set of samples.

Example: A marketing survey that wants to know what holidays people celebrate, includes this in their survey: Do you celebrate Christmas? This excludes all cultural groups from the survey who don’t celebrate Christmas but do celebrate other holidays such as Holi, Eid, or Hanukkah.

• Order bias

Order bias occurs when the placement or sequence of questions in a survey influences how respondents answer those questions. Participants may give more weight to earlier questions or recall the latest items more clearly. The order in which questions are presented can impact participants’ interpretations, and thus their responses.

Example: Asking about benefits before drawbacks might lead to more positive responses.

• Recall bias

When the mobile surveys intends to ask people questions from their past that they cannot recall. Or may recall incorrectly. It often happens when you are asking about things that happened more than a year ago from the time period of the study.

Example: A study on dietary habits where participants must recall their food consumption habits over the past year. Participants might struggle to accurately remember every meal consumed, portion sizes, and specific foods eaten, leading to inaccuracies in the reported data.

How to avoid response bias?

  1. Use audience segmentation: Divide your survey audience into segments based on their understanding and familiarity with your product or service. More knowledgeable respondents can jump ahead to later parts of the survey while those that are unfamiliar with your product may spend a few minutes answering questions that increase their understanding of what you do. It personalizes the survey and hence delivers more accurate responses.
  2. Add diverse question types: To avoid extreme response or acquiescence bias, avoid using one or two types of questions throughout the survey. Use different types of questions in your survey. A survey with Likert scales, dichotomous questions with Yes or No responses, MCQs, and essay-type questions encourages the participants to think clearly and answer with more care.
  3. Careful question phrasing: Phrase your questions more carefully. Instead of asking ‘Are you fluent in English’, consider asking, ‘What languages are you comfortable communicating in?’
  4. Short and clear questions: Keep your questions short and simple-worded. Questions that are too long, too many, or too difficult are some of the leading causes of survey fatigue.
  5. No leading questions: Leading questions subtly or directly guide respondents towards or against a particular answer. ‘Wouldn’t you agree that smoking is bad?’ is a leading question, and must be avoided. Instead, simply ask: Do you believe smoking has any effect on health?
  6. No double-barrelled questions: This type of question asks two questions in one. For example: Do you find this product useful and affordable? The participants may think the product is useful but not affordable. So break the question into two separate ones where one asks about usefulness and the next one asks about affordability.
  7. No absolute questions: Do not add questions where you are forcing participants to provide a definite answer without giving them room to be nuanced. For example, if you want to know people’s education level, don’t ask ‘Do you have a college degree’. Instead, ‘What level of education have you completed?’ is more nuanced and will give more accurate results.
  8. Avoid loaded questions: Loaded questions are similar to leading questions in their goal to influence survey responses. Loaded questions use emotive language in their phrasing and provide extremely biased answers. So avoid adding emotionally-charged terms in your language such as terrorist, junkies, concerned parents, patriots, etc., and use simple and neutral language in your questions.
  9. Simple and concise answer choices: Give respondents the room to say No or I don’t know. Survey questions that don’t include these choices force participants to choose incorrect answers.
  10. Optimized survey structure: Structure your survey in a way that considers how your audience will respond to different sections of the survey. Add personal questions at the end, randomize the order of answers to curtail acquiescence bias, and branch relevant questions together to increase respondents’ understanding of the concept.

How to avoid response bias?


As we wrap up, it is crucial to acknowledge the significance of identifying and tackling various response bias types to achieve precise research outcomes.

By being aware of acquiescence, social desirability, cultural influences, and other biases, researchers can apply proactive measures like randomized order, impartial phrasing, and inclusive participant selection. These strategies elevate the quality of data and guarantee the dependability and authenticity of research discoveries, making for more comprehensive perspectives that genuinely reflect participants’ views.

About The Author

Kelvin Stiles is a tech enthusiast and works as a marketing consultant at SurveyCrest – FREE online survey software and publishing tools for academic and business use. He is also an avid blogger and a comic book fanatic.