What best predicts useless research

Running a UX industry survey through a predictive model

What makes great, usable UX research? While working on a new predictive feature for a survey company, akin.nz, we decided to give it a test and upload an old survey we had lying about. For the feature, first choose a question and answer to learn about then see a list of answers and their predictive power.

The survey in question was of around 150 UX researchers, marketers, designers, writers and beyond with a mix of demographic-characteristic, behavioural, motivational and open questions.

Now that we have a survey of UX researchers and some predictive software, let’s get exploring. Since there were a slew of motivational and behavioural questions it made it super simple to have a question and see what other questions best predicted that question.

What akin.nz looks like.

First I was curious about what best predicted people answering ‘over 9 years’ to ‘How long have you been a researcher for?’ Which led me to:

The biggest predictor of being over 9 years in research is that no one uses your research.