I thought it would be nice to collect some recent learnings and share them on this blog so I have a motivation to write down what I consider to be interesting
In design, people sometimes tend to get stuck with their one solution. This seemed similar to confirmation bias to me. I wondered why there is this tendency. Confirmation, disconfirmation, and information in hypothesis testing (pdf) offers an interesting perspective on the phenomenon: The bias is not so much towards confirmation, but to testing what is likely to happen according to previous experiences.
In many situations it carries more risk if what is actually »bad« (let's say a design idea or a new car or a person you may trust) is mistaken to be »good« (false positive) then way around, so it makes sense to test what you assume to be a good choice. And if you search for something that occurs not very often it makes sense to use an hypothesis to identify cases in which the diagnosis is is assumed to be right. The chances to get an informative falsification are much higher in this case compared to testing cases which are supposed to not have the target property anyway (a needle-in-the-haystack) Actually, in some cases this can be the logically sound way to disconfirm a hypothesis – namely, if your hypothesis is too broad and encompasses the correct assumption.
Context in UX Studies
Doing UX Studies in the user's environment (as opposed to a testing lab) should help you to get data closer to actual behavior. I discussed the topics with my colleges. Particularly interesting was what context can mean for a study. What context is or rather: what matters about it can vary and should be considered (since providing that context may be easy or difficult). What could be context that matters for the user and use?
- Being at a familiar place
- Using known devices or device setups, a screensize one is used to etc.
- Ones own computer or smartphone
- The daily work environment, tasks, pressure, colleges…
- Familiar people
- Environmental conditions like light, wind, sound etc.
Grounded Theory (again)
I came to read the original Grounded Theory book by Glaser and Strauss again. Again, I'm impressed by their style and their suggestions and how they care about empowering researchers by changing and choosing methods.
(Like design, Grounded Theory is abductive and iterative. I like that.)
Should I worry about the applicability of my research
(Research here is not end user research, but research on design and research methods)
Long: It is likely, that there is always some factor (actually quite many) not under my control. So I may consider that user research does not even take place if a manager does not consider it as important, and I'll find that problematic, but I can't solve every problem :-)
What I learned recently by Jan Dittrich is licensed under a Creative Commons Attribution 4.0 International License.