Protecting the Science: Evaluative Priming for Consumer Insights
How Does Evaluative Priming Work?
A prime is any stimulus that precedes some target behavior of interest. In typical EP studies, the behavior of interest is the ability to quickly sort words or images into their appropriate category. The automatic associations that people have with the prime have an influence on the ability of participants to quickly and accurately sort the target words or images. This can be an invaluable tool for gathering consumer insights.
To illustrate, in a typical racial evaluative priming study participants are instructed that they will be sorting words into categories (e.g. positive “e” key and negative “i” key). In a baseline exercise, positive and negative words appear on the screen and participants sort them into their respective categories (note there is always a right answer). Their response times are measured as well as their accuracy. This creates a baseline for comparison to the experimental condition which follows.
Implicit Associations for Consumer Insights
In the experimental condition, white faces and black faces appear as primes very briefly (e.g. .5 seconds) before the words. The primes influence participants’ ability to make the judgments either by facilitating the judgment (if the target word is consistent with the implicit association evoked by the prime) or hindering the judgment (if the target word is inconsistent with the associations with the prime).
So for example, for a person with a positive automatic association with black faces, the categorization of a positive word following the presentation of a black face is faster, and fewer errors are made. It is easy to imagine how brands, products or packages could be used in place of faces to assess the automatic reaction to those common market research subjects.
There are a few design, analysis and implementation considerations for EP studies.
1) EP does not suffer from the same relativity issue as IAT, meaning #mrx professionals can capture a measure of implicit association with a brand that is not dependent on a comparative brand.
2) In concert with removing the relatively issue, EP studies can provide an absolute degree of positive versus negative sentiment.
3) Repeated measures are critical for stable assessment. A prime target pairing should be assessed multiple times in a study to reduce noise in the measurement.
4) The target words are representations of the categories. What is being measured is the relationship between the primes and categories. However, that measurement is only as good as how well the target words represent the category construct. This indicates that the target words should be chosen very carefully, and/or tested for representation of the categories.
5) Any construct continuum can be used including the most common (i.e. positive to negative) to the very brand specific attribute level (e.g. high quality to low quality). This opens up EP studies to much broader application for #mrx purposes.
Firms that are experts in evaluative priming should be able to provide guidance on the number of repeated measures you’ll need relative to your sample size for each implicit association you want to measure.
In addition, experienced consumer subconscious research firms can provide clear consulting on how these measures of System 1 implicit associations are best integrated with other quantitative measures of the more deliberate thinking of System 2 processing.
