Posts tagged “subjectivity”

ChittahChattah Quickies

  • [from julienorvaisas] Is America Mature Enough for This Line Graph of Gay Marriage Poll Results? [VF Daily | Vanity Fair] – [Here's a novel interpretation of a seemingly—ahem—straight-forward infographic; as well as an acknowledgment that your interpretation says as much about yourself as it does about the data.] A new survey of same-sex marriage poll results posted to suggests support for gay nuptials is—ahem—on the upswing. As one commenter points out, the site’s statistics overlord, Nate Silver, may also have giggled upon compiling the data: Silver describes a regression line as “fairly sensitive on the endpoints,” which, contextually, is rather raunchy. Do you see a bloated asparagus stalk turned sideways? A close-up of the mouth of a Pac-Man? The tail of a Na’vi? The answer may explain to you your own feelings about same-sex marriage. It’s a Rorschach test of inscrutable accuracy, probably as prescient as the polls themselves.

Experienced pollsters know: people “lie”

As I’ve said before, garbage in, garbage out. From Rob Walker’s Consumed

Recently, Stardoll did a study of its own, polling United States users about their brand preferences. Apparently they saw real-world brands on the same plane as the half-dozen or so invented brands that exist only within the site. (Some respondents even made the – clearly impossible – claim that they wear the strictly digital Goth-style brand Fallen Angel to school.)

These sorts of stories always crop up in market research and business case studies. And they are wonderful because they illustrate the depth of meaning the products, services, brands, and stories we create can be to the people that consume them. So meaningful that they will conflate pretend brands online and tangible experiences offline. Wow, we marvel, that tells you how great our stuff is; they will lie about it.

But the flip side to that is that if you are going to ask people what they think and do and want, you better have a way of triangulating their responses against other data. If you don’t know more about the person than their response, how can you contextualize it? If you don’t know what they are really saying when they answer the question – if they understood the question or are answering it in the way you intended – then you must be very careful in what you conclude and how you act on those answers.


About Steve