Skip to content

Data and charts printed out and placed on a table.

Are leading words causing you biased survey results?

Many departments use surveys to collect information about peoples’ opinions and experiences; however, if a survey isn’t designed correctly and uses improper wording, it may not yield essential, accurate, or valuable data. A significant issue in survey design is using biased survey questions with language to usher respondents’ answers to a particular outcome.

One dilemma with biased surveys is using leading words, which means the survey uses language that pushes a respondent to answer in a specific way. Leading words might include a positive or negative bias and may steer a respondent consciously or unconsciously toward a certain kind of answer. For example, if you send a survey to guests after attending a work event to gauge their opinions, you ask, “How much did you enjoy this event?” You have established a positive bias that implies that the respondent enjoyed the event, leading the respondent to a more favorable answer and skew the data you are collecting.

If positive or negative biases and leading questions are present, the purpose of disturbing the survey becomes diluted because the data will be biased based on the leading question. It is important to remember that having biased questions in your survey will result in biased results. You want your survey data to be as objective as it can be to help demonstrate the views and opinions of your respondents within your data collection.

So how do you remove leading words that may cause biased responses? One suggestion would be to structure questions as objectively as you can by using rating scales. If you look back at the previous example, “How much did you enjoy this event?” you could restructure the question by asking, “On a scale of 1-5, with 1 being the worst and 5 being the best, please rate today’s event?” Another option is providing balanced multiple-choice answers. For example, you could ask survey respondents, “How would you rate today’s event?” and offer the multiple choices options:

  • Excellent
  • Good
  • Average
  • Poor
  • Terrible

The choices above include a balanced range of survey choices and improve your chances of collecting objective and accurate results.

It is essential to confirm that the multiple-choice answers you provide are not leading respondents to more positive or negative options. For example, if you used the same question as above, “How would you rate today’s event?” but used the unbalanced multiple choices options:

  • Excellent
  • Very Good
  • Good
  • Average
  • Poor

A survey respondent may sway to a specific answer more positively because there are more positive choices within the options list, which causes a biased result. The unbalanced multiple-choice options could be problematic for a respondent who thought the event needed improvement but didn’t think it was a bad event. This respondent would probably not choose “poor” because it’s the lowest option, and they believe the event wasn’t at the “poor” level. The more positively ranged choices have caused the respondent to choose either higher or lower answers than they originally wanted to pick. Having unbalanced options may confuse the respondent to select a higher scaling choice than they believe the event represented.

Generally, surveys help gain and collect knowledge, and using an unbiased format when creating questions allows participants to answer truthfully without any suggestion from the survey’s wording. To honestly assess survey respondents’ thoughts and feelings, the person or department designing the survey must avoid using leading words that could sway the respondents’ responses positively or negatively.