User Testing As Justification For Opinions

August 23rd, 2006

In this article User Research Smoke & Mirrors, Part 3: Research as a Political Tool, a behaviour of using usability to win an design argument is described.

I wonder for how many other people this rings bells? I can remember many times when I was asked to run off and find the research that backs up “Not using popup windows in the checkout process”… as if there was someone stupid enough to spend time doing that research… of course there isn’t …

Now in cog-sci-hci circles there probably would be some papers on interrupting flow, surprises and cognitive dissonance but finding the ones that, with groaning stretch of imagination could be applied to an online shopping cart, something that didn’t exist when the original research was done was both frustrating and farsical.

But I have to say… at times I did it… I found research that argued that gree was nicer/better/faster/more user-firendly/profitable than red and plopped a load of papers on the desk that noone in their right mind would want to read. Although I’d had to…

And the whole time I was collecting and collating this research… I felt sick to my stomach… because I knew deep down that all that mattered at that point was my opinion… and that, at the end of the the day, my opinion, although it was based on years of reading HCi research and although it was often in consensus with everyone else round the table… counted for nothing.

“Show me the research… then I’ll believe your opinions”… really was always just about power… I wanted their money and had to dance through some ludicrous hoops for their entertainment first… and to my shame… I did.

(And yes, yes, yes… we all know the dangers of thinking we know it all, doing no user testing and producing totally irrellevant work… but a more insidious danger is to waste time on research that tells you nothing you didn’t know already)

Share me:
  • Digg
  • Sphinn
  • del.icio.us
  • Facebook
  • StumbleUpon
  • LinkedIn
  • SphereIt
  • Technorati
Tagged with:

Leave a Response