Before you start your study, here’s a quick summary of the main types of bias that might creep into your work, and how to minimise them.

Researcher bias / Subjective bias

Your background, experiences, demographics etc will affect what you study and how.

Reduce the bias with reflexive questioning: Reflect on how your background and your biases may affect your study.

Subject bias / Withholding or distortion

People might not tell you the truth or might distort the truth in some way. For example, people in a sport team may not give genuine feedback about the quality of the coaching, for fear of retaliation.

Reduce the bias by building trust: Be clear about what you are studying and why. Engage with subjects over time so they get to know and trust you. This will encourage them to open up. Be clear about how you’ll protect privacy and confidentiality.

Contextual bias / Environmental influences on behaviour

People act differently, and share different information, in different contexts.

Reduce the bias by choosing a relevant context: Try to study people in the most relevant context, based on your research question. If you’re studying team dynamics, talk to players at practice. Spend time in that environment, so people get used to you being there.

Make it easy for people to share: Be considerate. Don’t ask personal questions in a crowded environment. Try to adjust the context or environment so that it’s easy for people to share.

Research method bias / Biased recruitment

No matter how you recruit people, there will be some inevitable bias in terms of who you reach and who decides to participate. For instance, recruiting via social media will skew towards a more tech-savvy audience.

Reduce the bias by aiming for balance: If you are aware that your recruitment process will favour one type of person (e.g., tech-savvy people), you can try to balance this with other forms of recruitment (e.g., in-person sign ups).

Be clear about the limitations: It’s difficult to eliminate all bias, so make sure to acknowledge the bias that exists.

Biased analysis 

Bias can accidentally creep in at the data analysis stage, as you decide what to pay attention to, and what it all means.

Reduce the bias by getting a second opinion: If you can, get a second person to run over your data set. What do they notice?

Triangulate your data: Compare your findings with other findings. You could triangulate within your study, to see if what you heard in interviews aligns with what you heard in focus groups. Or you could triangulate with sources outside your study and see if your findings are generally consistent with what other researchers have found.

Try to prove yourself wrong: Humans are wired to seek out information that aligns with what we already think: confirmation bias. So it’s useful to actively try to find information that goes against what you believe. If you find data like this, don’t try to bury it. Instead, study it more deeply and see what you can learn.

“A proposition deserves some degree of trust only when it has survived serious attempts to falsify it.” (Chronbach cited in Brink, 1993)

Lack of transparency

Lack of transparency is an issue if you want your study to be ‘confirmable,’ or repeatable.

Reduce the bias by being crystal clear about your processes: Include very clear notes about how you’ve conducted your study, and the decisions you made at each stage. PLUS TABLE Source of bias etc