Testing the validity of implicit measures of wanting and liking

Over the last decade, there has been a surge of studies examining implicit processes underlying addiction. Some implicit measures are assumed to reflect "liking" whereas other implicit measures are assumed to reflect "wanting". There is, however, little evidence to back up this claim. We examined whether implicit and explicit measures of wanting and of liking are differentially sensitive to manipulations of wanting and expected that these manipulations would affect primarily measures of wanting.

Smokers and non-smokers performed both implicit and explicit measures that are assumed to assess wanting and liking for smoking. Smokers were tested once immediately after smoking, and once after 12 h of nicotine-deprivation.

IAT results suggested that smokers showed more implicit liking for nicotine when they were deprived than when they were satiated, whereas there were no differences in wanting. Smokers also seemed to show both more implicit wanting and more implicit liking for nicotine compared to non-smokers. Explicit measures did yield the expected results in that smokers reported to want nicotine more when they were deprived, whereas there were smaller differences in liking.

We found little support for the assumption that implicit measures of wanting and liking capture different processes. Researchers should thus be cautious in drawing conclusions about wanting and liking on the basis of these measures.