If you have conducted an opinion poll, for example, and you realized that some of your questions may have prompted a respondent to answer a certain way, acknowledge that in this section. “The survey questions included a statement that may have indicated to the participant that our research was being sponsored by the school. This statement was listed toward the end, and most likely only influenced the two remaining questions. ”
If you are interviewing a coworker or a colleague, avoid asking directly if they are unhappy in their current job. Reframe the question so that it is not as direct. “What do most of your coworkers think about the management in your office?”[4] X Research source This indirect question about a third party can foster a honest response from the participant.
Don’t ask a participant a close-ended question that they can answer simply. Instead of asking someone who they voted for in the last election, ask them to describe how they felt about each candidate. If you are trying to gauge if a new workflow process in your office is helpful, ask employees how the process has influenced their work. “How has this process helped or hindered your workflow?” This question will reveal more than simply asking whether or not they like the new process.
Try to remove any trace of the sponsor from the interview or observation, and do not express your own personal feelings or opinions. [5] X Research source Do not add your company’s logo or your school’s seal on materials provided to the participants. If a person is being polled on how they feel about the efficiency of a particular school, for example, they may provide a biased answer if they suspect or know if that institution is performing the research. If you are collecting opinions regarding the admissions process from current students, do not let the participants know if you work in the admissions office or sit on the admissions committee.
Instead of asking a respondent to agree or disagree in a customer satisfaction survey, ask item-specific questions. [7] X Research source Form a more direct question instead of asking a participant to respond to the statement, “My experience in the store was satisfactory. Agree or disagree. ” Ask a participant, “How was your overall shopping experience at this store? Excellent, good, fair, or bad. ” Additionally, it’s a good idea to allow the respondent to review their answers before submitting their response, as this allows them to confirm their answers accurately reflect their views.
During an election, supporters of a particular candidate may only seek out news sources that showcase their chosen candidate in a positive light. This is confirmation bias. This can affect how you perceive a candidate and can influence your decisions. [9] X Research source
Sort data into categories that make sense for your project. List it by observation type, by date, by location, or by participant background information. [11] X Trustworthy Source Kansas University Center for Community Health and Development Community-based research center focused on supporting public health development and education Go to source As you sort or code your data, ask someone to help you or to review your work. It’s likely that you will need to interpret ambiguous answers, which allows room for bias. Having multiple researchers interpreting the data will limit the risk of bias influencing your results.
Before collecting data, ask a colleague to review your methods section to look for questions or approaches that may lead to biased data. When you have written your final report, ask another mentor or researcher to review the results and the conclusions to look for signs of bias.
