If you haven’t seen the new Leonardo DiCarpio thiller “Inception”, then you need to.  And when you watch it, watch it through the eyes of a marketer or researcher.  The film centers on Dom Cobb, a thief who enters the dreams of others to obtain information that is otherwise inaccessible.  His abilities have cost him coveted things, but a second chance allows the situation to be turned around. However, when an unforeseen enemy predicts Cobb’s every move, Cobb and his team of specialists are forced to plant an idea in a target’s subconscious.  The film’s title refers to the task of planting an idea rather than stealing one, a concept that Cobb is less acquainted with.  Here’s a link to the trailer.

Cobb must not be a marketer or researcher, for as an industry we tend to do a brilliant job planting ideas in the markets’ mind; it’s called priming.  No communications can ever get avoid priming.  It’s part of basic persuasion.  But sometimes it can be used unethically via Neuromarketing.  For more on Neuromarketing Ethics check out another recent blog on the topic.

And when it comes to research, priming also produces flawed results.

Let me highlight an example from the movie and translate it to the accuracy issues in research.  Without giving away much about the plot, one of Cobb’s colleagues Arthur, or The Point Man, is challenged by another character Saito, that you can’t plant an idea in someone’s brain.  Arthur says to Saito, “Elephants” and then asks him what he imagines.  Saito says “Elephants”.  I don’t want to reveal more, but I think you get the point.  Arthur was demonstrating to Saito the technique of priming.

So let’s translate this to research.  Here’s a question I saw on a recent survey.  “Is it important for your online school to have its administrative offices in the same state you live in?”  I’ll guarantee you YES will come up more often then the true belief.

Don’t believe me?  Here’s a test you can perform.  Ask half of the respondents that question, and ask the other half, “Is it not important for your online school to have its administrative offices in the state you live in?”  I’ll bet you there is a noticeable difference between the percentage who answered YES in the first question and NO in the second question, even though those choices are technically the same.

What’s going on here?  It’s priming at work.  And in research it’s just plan stupid and sloppy.  The first question as written subconsciously adds a bit of implied importance to answering YES to the first question.  When you ask, “Is it important…” you actually get a bit of biased toward YES.  It’s priming at work.

Mostly it’s due to poor question design and not purposeful manipulation.  The good news is, poor design is fixed easily.  Here’s how.  Simply reword the question to the inverse of the question and ask yourself, do you think there is the possibility that the results from the two sets of question could possibly be different?  If yes, well you have a priming problem on your hand and you need to reword your questionnaire.  You can also test the questionnaire doing an A/B split with the question.  If you find the questions produce substantially different results, then kill the question as priming is an affect.

You might be able to plant an elephant in someone’s mind, even a pink one.  But when you do, the answers they give you will likely be flawed.  So be careful with question design.

So what color is the elephant?