Analysis - how do you analyze your lovely data?


#1

I’m having a surprising amount of trouble figuring out how to analyze data from interviews and user tests.

Admittedly, I just finished my first project, but I’d prefer to feel a little less perplexed when it comes time to do things with the data I collect.

For my interviews, I ended up putting summaries on stickies which I then put on whiteboards under topic titles, and looked for patterns. For my user testing, I looked for major problems or concerns and discussed with the project owner. But neither of these seem like real analysis to me, at least not compared to psychology research.

What do you all do when you have user interview or user prototype test data?


#2

If there is too much data from the interviews and test, the sticky note summaries are the best way. Of course, this is most effective when used with qualitative data.
I follow the same method too, I write down the main ideas on stickies and randomly paste them on a wall.
Later I try to classify them using a method that seems fit : Affinity Diagramming, Concept Model, Flow Model, etc.
This is pretty much the only way I’ve come across to analyze qualitative data.
There is a great article about this Here.


#3

Make sure you join us for Joe’s session tomorrow (or today, depending on where you are) on Slack which is all about unpacking the results of user testing and translating them into functional requirements.

There are also some really good tips on what to do with data that you collect from contextual enquiries/site visits in this video by Gerry Gaffney.

Hope that helps a bit.


#4

The key to being successful here is pattern recognition. There are many different methods of doing this, not the least of which is the sticky note approach you took.

That said, pattern recognition is a skill that you can (and will) hone over the course of time. Things may seem a little overwhelming now, and that’s a good thing! It means you got enough usable data to make things useful and interesting for you, a far better result than not having enouh data. I’m sure Joe’s Slack session will provide some good help, as does the article from @enlightened_06. Take what you can, and gear up for the next round! Focus on getting more comfortable with your process, and you’ll be in better shape going forward.


#5

I would totally love to join for Joe’s session, except I have plans this afternoon! Alas. Will check out the transcript after the fact, for sure!


#6

Ah, thank you. That link should help, as should reassurance about the idea that it really is a reasonable way to do things. :slight_smile:


#7

Thank you. :slight_smile: