Not everything is black and white
This article is part of a series by Katie Harris, principal at Zebra Connections (if you missed the other parts, start here).
In this post I respond to a (very good) question that I was asked about researcher bias and subjectivity, specifically what to do about it. Interesting topic this one!
Researcher bias and subjectivity in qualitative research
It’s true, without doubt, that personal biases and subjectivity shape the way researchers perceive and interact with research participants. The way someone looks, the way they talk, how many mini-muffins or slices of pizza they eat during a focus group session: all these factors influence the way a researcher ‘hears’ (or doesn’t hear) them. In effect, it’s probably true to say that personal biases and subjectivity shape the research output.
(I say ‘probably’ here, because, of course, it can’t actually be proven. The impact of researcher bias and subjectivity on the research output could only be assessed if a control group was used for comparison. But the control group would be subject to the very same forces. There’s no escape! Let’s just assume however, probably quite sensibly, that bias and subjectivity do exist and they do have an impact.)
They’re everywhere you look!
When we talk about researcher bias and subjectivity, it’s often within the context of the qualitative interviewing process. But it’s important to note that bias and subjectivity exist across the entire the research process.
Bias and subjectivity determine how the project is conceptualised, the shape and focus of the research, the methodology used, etc. And, even if we were to crack that nut and design the perfectly controlled research study (impossible), and then proceed to conduct it with scrupulous objectivity (impossible), our interpretation would still be subjective and biased. Completely.
The main point to note here is that qualitative or quantitative, market, social or academic, there’s no such thing as objective research. Truth be told, it’s all just one big subjective, bias-orama.
What to do about it
Bias and subjectivity in qualitative research can, and should, be acknowledged and where possible, minimised. There are many steps researchers can take to do this – I’ve listed three useful ones below:
One way to minimise any particular bias or subjectivity is to moderate the effect by throwing several more of them into the mix.
To this end, gathering input from a wide range of sources and perspectives, including non-researchers, to develop research design, hypotheses etc. is good practice. The more the merrier. Any particular bias will lose a little of its edge when tempered by the crowd.
Expect the unexpected
Interview skills 101 can teach researchers how to listen, with empathy, and thus with a little less bias and subjectivity. But it’s largely through experience alone – unearthing enough unexpected insightly gems from research participants who often don’t look as if they’d have much to contribute – that you begin to seek and expect the unexpected. And expecting the unexpected is a great way to control your biases.
Online research is about as close as you’ll get to bias free interviewing (until you trip over some spelling mistakes or a misplaced apostrophe and your literary biases flare up). Spelling mistakes aside however, there are less hooks in an online session on which to hang our biases.
To illustrate the point, let’s compare two scenarios here – offline and online focus groups.
Scenario 1: An offline or ‘live’ focus group
A research participant arrives to take part in a focus group. And here he is (see image below)!
What’s he like? Just by looking at him, in a nanosecond, my biases have informed at least 10 assumptions about him. And assumptions about what he’ll think and say about topics X, Y or Z…
Scenario 2: An online focus group
In this scenario, we don’t actually meet our research participant in real life. From a moderator’s viewpoint, this is what Harry looks like:
He looks like 24 point font, just as you see it on your screen right now. That’s all the other research participants can see too.
What’s he like? I’ve no idea. I can’t make assumptions based on how he looks, how he talks, how he smells(!). And, importantly, this minimises my assumptions about what he’ll think and say about X,Y and Z…
[For the record, I’m not a great fan of real time online groups, but I do think groups which are conducted over time – days, weeks etc – can be very useful!].
So there you have it. Bias and subjectivity exist throughout the entire research process. In cases where they go unseen and unchecked they can, and should, be of great concern. It’s definitely a problem if they aren’t acknowledged, are allowed to shape the research, and no one is any the wiser.
So make sure your research consultant acknowledges them as a first step. Make sure they work to minimise them; that they employ a multitude of perspectives and are experienced enough to expect, and seek, the unexpected. In appropriate cases, consider conducting your research online.
Check out my next post in which I look at how to get buy in/explain the value of qualitative research. What’s the ROI on a qualitative project?