The Valentine’s Day Brain Challenge
Valentine’s Day has been over us, and the ever recurring question for every man, boyfriend and lover is: what is the best gift for my most precious one? This question haunts very man (only 2% of men in Denmark like this particular day…), and a recent Valentine’s Day Experiment demonstrated some neurophysiological benefits of Valentine’s Day, as true VD behaviour increased oxytocin release (or, in men, vasopressin release).
Our sympathy goes to her husband…
In a TV stunt soon to be broadcasted at the Danish broadcast company, Neurons Inc took on the challenge of predicting consumer choice. More specifically, we were asked: can we predict a woman’s Valentine’s Day gift preference better than her long-term boyfriend/husband?
To add to the challenge, we had to make the analysis within only a fraction of the time that we usually do, leaving us to our preferred metric – the well-known frontal asymmetry motivation score. Instead, we relied on arousal, both from brain based measures and pupil dilation.
During the test, the participant (the actual program presenter) went through a store and found 5 different items. One item she loved, one item she hated, and three she was more or less neutral towards. The task for us, and for her husband, was to guess which item she desired, and to avoid choosing the hate-gift.
Unfortunately, her husband fared the worst: he selected the item she absolutely despised. Our sympathy goes to her husband…
Did our neurometric test fare any better?
By looking solely at our pupil dilation measure, we found that we were actually spot on with the prediction! The highest arousal was for item nr. 2, which is the one she chose.
Looking more into the dynamics of our physiological and neurological measures, we have subsequently run an analysis of the best possible model using our scores. We ended up with a model that combined the following features:
- * total viewing time (from eye-tracking)
- * arousal (from pupil dilation)
- * arousal (from EEG neurometrics)
- * interaction effects between the above parameters)
(technically speaking, we ran an exploratory model using stepwise multiple regression with backwards elimination)
By using this model, we end up with a model that predicts 81.7% of the variation in choice. That is, the model is 81.6% likely to predict the actual choice made.
Should men use neuromarketing to predict what they should give their loved ones? Following our finding, one should say a resounding “yes”. Perhaps neuromarketing can be used to more than merely predicting consumer choice. Following the Valentine’s Day challenge, neuromarketing can be an actual marriage saver 😀