One of the ways we do this is through experimentation. We typically have undergraduates at Oakland University participate in our analogy experiments in exchange for course credit. These undergraduates are taught how to use analogy to generate ideas. They are then provided with a target domain (solving the parking problem on campus) they are familiar with and very motivated to provide solutions for. We then manipulate various factors that affect analogical reasoning such as the source domain being used (e.g., parking at the mall) or the mapping between domains (e.g., map the similarities between domains). We then assess the solutions generated by the research participants to determine which conditions yield the most creative solutions.

One of the issues we have had to deal with in these analogy and creativity experiments is how to assess the “creativity” of the results.

As I have mentioned in another post, the definition of what should could as a creative solution to a problem is a matter of great debate. At SAC, we lean towards the viewpoint that a creative solution to a problem is one that is both novel and appropriate (i.e., it solves the problem).

Consequently, in our experiments we assess the novelty of each solution (how many people in the experiment generated that solution) and the practicality of the solution (generated by having a group of individuals rate the practicality of the solution on a scale from 1 – 10 where 10 = very practical). Our measure of the creativity of the solution is the average of that solution’s practicality and novelty scores. By this measure the most creative solutions are those that are both very novel (generated by few people) and very practical. We call this our “objective” measure of creativity.

There are some creativity researcher though that argue that a creative solution to a problem is one that is both novel and has value. Accordingly, we created a second measure of creativity. We have a group of research participants rate the creativity of each solution on a scale from 1 – 10 where 10 = very creative. This “subjective” measure of creativity is the average creativity rating for that idea.

Now one would think that our objective and subjective creativity ratings should be fairly well correlated. After all, people probably determine the value of a novel solution by assessing how actionable the idea is. However, our research suggests this is NOT the case.

There is actually a fairly weak, inverse relationship between our objective and subjective measures of creativity (r = -.174). What this means is that the objective measure of creativity rarely corresponds to the subjective measure and when it does it moves in the opposite direction.

For example, one solution to the parking problem on campus is to reduce the number of students at OU by flunking students who’s GPAs drop below a 2.0. This solution actually receives a fairly high objective creativity rating because only one person in the sample generated that solution (it is novel) and it was rated as a fairly practical solution to the problem. However, this solution received very low creativity ratings by those asked to rate the creativity of the solution.

Another example is the solution to the parking problem on campus that involves providing valet parking for students who are running late to class or who don’t want to bother with searching for a parking spot. This solution was rated high in creativity by OU students. However, its objective creativity rating was fairly low. This was because even though the rated practicality of the solution was high, the novelty of the solution was very low. Over 30% of the people generating solutions to the parking problem on campus who used “parking at the mall” as their source domain came up with valet parking as a potential solution to the parking problem on campus.

So what does this lack of correspondence between “objective” and “subjective” measures of creativity tell us?

It suggests that people’s assessment of the creativity of a solution involves more than an unbiased assessment of “practicality,” “appropriateness,” “novelty,” or “value.”

We know from talking with our research participants that many of them believe the ideas they generated will be used to inform future decisions regarding the parking problem on campus. Consequently, that belief may cause them to dislike solutions involving flunking students with 2.0 GPAs. This dislike causes them to give low creativity ratings to those solutions regardless of how practical, valuable, or novel those solutions might be.

It also provides further support for the proposition that individuals working with the same domain during analogical problem solving generate the same solutions to the problem. From the standpoint of minimizing conflict and competition during group ideation sessions, this is a selling point for the methodology. It also suggests that group consensus concerning the best solution to a problem does not translate into a perceived lack of creativity on the part of the general public concerning that solution. It doesn’t matter how many people in the ideation session generated that solution. If those assessing the creativity of the solution are unaware of the process by which the solution was generated, that process can’t affect the perceived creativity of the solution.