Simon Roberts, the highly engaging, smart and easily approachable chair of the EPIC conference last week, was so absorbed with all the logistics that he didn’t find the concentration to speak his mind during the conference. Now that the conference is over (and well organised it was!), it took him less than a week to type out a long blog post to position his thoughts on Big Data. It’s a long read but very much worth it, and it starts off exactly with the right criticism:
“The discussion at EPIC 2013 disappointed me a little. It was either constrained by simplistic oppositions (big data good / nothing to fear vs. big data bad / end of our profession as we know it), impoverished by a general lack of ethnographic specificity and illustration, or absented to discuss the power relations that big data entails.
Most worrying for me of all of these was the lack of specificity in the discussion and the absence of discussion about power. “
Exactly my thinking as well. There is an asymmetry in power relations that requires serious reflection and analysis, and it was dearly missing, sometimes even actively sidelined – as if irrelevant for ethnographers. There is an ethical and even political side to Big Data, that we have to very aware of, as user researchers and as designers (i.e. the professionals that mediate the relations between corporations and people).
Very helpful are Simon’s four dimensions of Big Data which articulate this power imbalance in more detail:
- Quantified self vs. Monitored Self — the difference between me assenting to monitor myself vs. being monitored
- Asymmetries of exchange — the uneven nature of the exchange between provider and analysyer/reseller of data
- Asymmetries of feedback — the importance of balanced feedback systems
- Asymmetries of judgment — the difference between the big data creating ‘fact’ and being used to create value judgements
He uses the example of the driving style tracking device that an insurance company installed in his car to raise some very good questions.
His three challenges (on incentive structures, interaction design and business risks) are spot-on. Read, read, read!