Are we truly innovating in social science research, or are we just building better rope?

When I think of innovation in survey research, I think of innovation in our "sister" field of land surveying.  (A field I think of too often as I correct confused extended relatives about the type of surveys that I'm involved in.)

In 2600 B.C. Egypt, "rope stretchers" were some of the first land surveyors.  They stretched rope to measure land distances.  But rope had its flaws - it was often weak, stretched, expensive, and in large amounts could be very heavy.  Small innovations took place to bring lighter, stronger, less stretchy, and less costly rope.  Eventually rope became metal and/or chain.

But through those innovations, surveyors were still "rope stretchers".  Until the game was changed.   A disruptive innovation hit the land surveyor market - the laser.  With one piece of equipment, lasers provided the land surveyor with the ability to measure any distance between two points.

As we have moved from in-person interviews, to phone surveys, to mail surveys, to web surveys, to mobile surveys, we have made great strides and we have enhanced the quality of our science.  However, have we not simply been doing to surveys what the land surveyor was doing to rope for all those years?  We're still just asking people questions.

We can (and sometimes do) capture a massive amount of paradata during surveys - are we using it?

Mobile devices are not just small web browsers that allow us to put a web survey in front of a respondent wherever they are.  Why are we ignoring most of the advanced capabilities that mobile devices bring us?

Consider these data collections:

  1. Should we really be asking respondent how much alcohol they drank today as a way to measure their blood alcohol content (BAC)?  Or should we start measuring BAC directly via bluetooth connected breath devices or skin alcohol measuring ankle bracelets?
  2. Do we ask respondents about their heart health, or should we capture their heart rate, activity levels, blood oxygen levels, and other health data through their devices and connected wrist bands?
  3. We know that respondents' moods can impact how they interpret and respond to questions.  We also know that we can detect people's mood  by looking at key indicators in their face.  Can we capture images of respondents using on-board mobile cameras to learn about mood, and thus assist us in analyzing their responses to the questions we ask?
  4. A significant number of answers to human behavior may reside in our genetic code.  Why are we not using that data in our analyses (or even our skip logic in surveys)?

Surveys are vital, but they reflect an artificial way to measure humans.  They are filled with sources of error.  Normal human behavior stalls and normal thinking stops while one is participating in a survey.  Simply updating the mode in which surveys are delivered will not change this much.

Maybe it is time to rethink the rope and look for the laser of our field.