Lying in design research
Funny highlight of the day: Dan Saffer's talk at Design Research 2007 entitled "How to Lie with Design Research". Have a look at the slides (.pdf, 9.5Mb) or watch the video. The talk is a funny description of tips about how to lie in design research. Saffer describes how to "Deliberately misinterpret data", " Willfully confuse correlation and cause. Pick the reason you like the most", " Toss out data you don’t like" or "don't be objective", etc.
Further out, I was very intrigued by this presentation because of his first point. He started the talk by showing picture from Japan, describing their implications and finally throwing possible design principles... and then tell the audience "All the images you just saw, I collected off Flickr in one afternoon. Voila, saved myself a trip to Osaka. I’ve never even been to Japan.". Which lead him to these tips:
"1. Don’t do any design research. Make it all up. / Don’t go into the field unless you have to. Why do research when you don't really need to? Most of the time companies are just looking to have their ideas validated. Why not give them what they want using carefully chosen photos and "stories" from the internet. TIP: Wacky cultural practices always impress. For "international" research, be sure to throw in a couple of unexpected cultural practices to make people feel that they've really taken the time to consider diverse perspectives. TIP: Don’t lie about the easily (dis)provable.
This resonates with the discussion about second-hand data, their values and their implications in social sciences (leading to terms such as "armchair anthroplogists"). It made me think of what Anne wrote last year about design ethnography: "If armchair anthropology was a product of colonialism, then design ethnography is a product of capitalism" to some extent.
Why do I blog this? Working in academia and thinking about this hints with regards to social sciences methodology makes it even more hilarious. He makes here a really good point about how user/context research can be abused and to what extent results are skewed to meet the needs of researchers, companies or other stakeholder in the process. As usual, it's not only fun but very relevant to see the mistakes, flaws and problems of a research process.