At the risk of sounding like a broken record in the UX space, ‘you ≠ user’. But as often as it is that this term is thrown around, its habitually easier to say than to put into practice. So why is it sometimes easier to say stuff than do stuff?
Well, Daniel Kahneman (author of ‘Thinking fast and slow’ and Nobel Prize winner in economic sciences) and his late friend and fellow psychologist Amos Tversky provided an explanation. They call these mental shortcuts: “cognitive biases & heuristics.”
In this post I’ll aim to outline an example of of one of these heuristics and (thanks to Daniel Kahneman) how knowledge of them can help us ask our users better questions, collect better data and make better decisions so that ultimately we can provide a better user experience.
Commonly when we ask a user to carry out a key journey on a live website, we identify multiple routes to a successful outcome. A real interaction with a user interface is frequently different to the designer’s ‘happy’ path. So, when we ask a participant to complete a task, the route in which they take to a successful or unsuccessful outcome can be dependent on the context or frame in which the task is set.
What I’m getting at here is that our study task instructions are especially vulnerable to biases.
So what is ‘Framing’?
Definition: “A frame is the context used to describe an idea, question, or decision. Frames heavily influence our interpretations and conclusions by emphasizing (or ignoring) certain aspects of a situation.”
Kahneman’s investigations into framing found that the exact same information can lead to opposite conclusions depending on the frame or context used to present the decision.
A remedy for this bias may seem flippantly obvious (don’t ask leading questions and don’t specify leading task instructions). For example, below are a couple of questions. One is a leading question, the other is slightly more neutral:
Is the updated design (Y) better than the old design (X)? Indicate your preference below:
Please indicate your preference for either design X or design Y below:
This (as mentioned before) may seem obvious, however, the framing effect can sometimes present itself in more subtle ways.
Imagine you are testing your site’s Information Architecture (IA) using a remote unmoderated tree test method and want to remove some items that appear more than once in your navigation tree structure.
Using a question like…
Where would you look for more information about X in the tree?
may steer people towards the ‘INFO’ level in your tree, more so than:
Where in the tree would you find X?
The first task description here is framed in the context of looking for more information about a specific item in the IA. This may influence participants to look in the level of ‘INFO’ rather than where they would normally go.
Hopefully you can see the effect that framing can have on the behaviour of your participants and could lead to incorrect conclusions being drawn from research.
Framing is, however, sometimes needed to elicit a user’s true behaviour. Let’s take an adaption of Krug’s IKEA website example to illustrate this. Imagine we want to investigate the user journey to the bookcases section. We could phrase a question like:
Imagine you want to find and purchase a bookcase using this website. Please find a suitable product, add it to your cart and complete the checkout process.
Often an instruction like this one would suffice, but imagine you observe an uplift in participants using the search bar by simply using the search term ‘bookcase’ compared to your previous research that says otherwise.
Framing the task by giving the participant a specific product to look for may be skewing the results in favour of the search bar. Here, by providing a framing scenario we are able to give the participant some context and help elicit their true behaviour:
Imagine you have several boxes of books in your kitchen and want to find a way organise them. Please find a suitable product, add it to your cart and complete the checkout process.
We may find that this real world task, may leave the interpretation more open, might grant you much richer insights into how people use or would use the site.
If you take one thing away from reading this post please let it be that framing can affect your UX research results in both good and bad ways. And next time you’re writing a task description or designing a question, imagine they’re works of art and you want to place them in the best frame possible (maybe even from IKEA!?)