Attend Spring Meetings on Development topics from Apr 17-21. Comment and engage with experts. Calendar of Events

Syndicate content

Taking What We Don't Know Seriously

Antonio Lambino's picture

There are three criteria for testing hypotheses on why people think and/or act the way they do.  First, there must be covariation.  That is, one phenomenon, such as watching a shampoo ad on TV, must be associated with another phenomenon, such as liking or buying the shampoo being advertised.  Second, there is the criterion called temporal order.  For instance, watching the ad should come before buying the shampoo.  Third, we must make a case against other likely explanations.  Following the shampoo example above, it's likely that a person who is predisposed to buying a product is also likely to pay attention to an ad about it.  Perhaps there's something about the person, such as having particularly thick hair or split ends, that explains both behaviors.  So how do we know that watching the ad causes the purchasing behavior?  We usually don’t know for sure.  But that doesn’t seem to stop intelligent and highly educated marketing and advertising executives from making decisions that lead to a whole lot of advertising. 

Social scientists, including communication scholars, have studied the possible determinants of a host of behaviors -- not just buying products, but also volunteering for community organizations, participating in political activities (e.g., voting), binge drinking, and using condoms.  And empirical research designs range from experiments, with carefully manipulated conditions (attempting to meet the three criteria above) to cross-sectional studies that do not bring us quite as close to being able to say whether a behavioral hypothesis is supported.  There are, of course, pros and cons for various designs and methods, which are standard fare in most introductory methods courses.

But not all questions related to influencing human behavior lend themselves to this type of testing.  For example, it’s well and good to try and find out whether violence in the media leads to aggressive behavior.  But should findings suggest that it doesn’t affect, say, the general population all that much or even not at all (not true, according to some of the best research in this area), does that mean policy makers should sanction pushing the limits of our fellow human beings by bombarding them with all sorts of blood and gore?  For those with mental and developmental issues, are we willing to say that because media violence doesn’t affect the general population (again, not what the evidence suggests), it won’t affect them either?  Furthermore, what kinds of studies might allow us to collect evidence, either way?  Will they pass ethical muster?  There are obvious issues with gathering a group of people with special needs and exposing some of them to violent messages to test whether they might act differently from those who aren’t exposed -- while controlling for a host of potential confounding variables.

So we can split hairs and run all the studies we want, but the fact remains that human beings are complex and so are the societies in which they live.  The contributions of sound social science research are certainly valuable, but limited and incremental.  We will never find all the certainties we seek.  That said, I believe that most of us are not willing to allow children access to pornography.  Most of us wouldn’t support providing public access to “how to” guides on producing dirty bombs.  And perhaps relevant to recent events in some countries, do we think that prominent individuals who enjoy the adulation of large groups should be able to use incendiary language against others in the public arena? 

I think most reasonable people would agree that words, sounds, and images can influence behavior to varying degrees.  It is probably also the case that our opinions on these matters depend on subject, degree, emphasis, context, and other considerations.  While I believe that prominent among them should most certainly be the existing evidence base, research findings must be used as an input to illuminate not only what we know, but also don’t know and might never know.  And when it comes to the terrible things human beings can do to themselves and each other, societies might then judiciously use public policy to hedge against the uncertainties research renders a little more evident.

Photo credit: Flickr user Marco Belluci

Add new comment