A few years ago, I answered the phone.  I’ve since learned my lesson and silenced the landline.  When someone leaves a message there now, the tiny blue light flickers forlornly until I log on to the interwebs to listen and laugh at the voice mail.  For those particularly entertaining, I forward to my wife’s email for her bemusement.  But on this day, I answered the phone.  On the other end I found an individual conducting a survey on behalf of Freddie Mac and Fannie Mae.  For reasons I can neither recall nor fathom, I listened and agreed to participate.  Once told of the subject, I told the person that I had no connection or experience with these organizations.  It turned out, that did not matter.  She continued to ask me questions about the firms (whose names she read en toto for each question for the next ten minutes); probing all around my completely vacant perception of them.  I wondered aloud how useful this information was, and briefly considered making up outrages or plaudits just to make her day more interesting.

Today, there are new stories about these firms’ attempts to improve their branding and message.  I suspect my interview was part of that, and no doubt rolled up and considered insight into the public mind.  Some unnamed (and named) consultants made serious coin analyzing these results and suggesting ideas to improve the numbers.

How does my experience resemble political polls, which today make up approximately 67% of all news stories? (Statistics are fun to make up, try it yourself!)  How do people respond to questions about how they will vote in a little less than a year?  How many of them take that call as seriously as I took my Fannie Mae / Freddie Mac survey?  How is it so many people still use landline phones, apparently the only method by which these survey firms reach people?

A student of mine opined recently on the qualitative method by declaring it inferior, only useful for setting up the hypotheses for more grownup quantitative methods.  These quantitative methods feature, often, scientific polls with established margins of error.  Far better to consider the aggregate of poll results, careful diced and analyzed; over the anecdotes and full narrative of experience.  Such is the domain of the soft science.  Where “data” relies on those people who are eager to give honest answers to a stranger interrupting their day with a ten-minute questionnaire.

I don’t mean to impugn completely the survey method.  I just wonder how much of what passes for ‘data’ should be taken with a few grains of your favorite seasoning.  Layering time-honored mathematical models on top of an individual’s representation of their thoughts and intentions may not affect, it turns out, the quality of that information.

Comment