Challenging the common belief that using unrelated ideas to build on the past and create something new, a new study makes the case for working from within the nucleus of a problem in order to best solve it.
"For people needing fresh inspiration for a problem, these findings imply that you shouldn't just go off and talk to random people or read things totally unrelated to your problem," says lead author Joel Chan, a postdoctoral fellow in Carnegie Mellon's Human-Computer Interaction Institute. "These might yield novel ideas, but not necessarily...useful and novel ideas."
In the study, Dr. Chan and his mentor, Christian Schunn, of Pitt's Learning Research and Development Center crowd-sourced their data using OpenIDEO, a web-based innovation platform for problem solving in domains ranging from environmental conservation to human rights violations and job growth for youth.
Here, individuals are able to post their ideas for solving problems. They start out open-minded in a freewheeling "inspiration" phase, and fine-tune their proposals thereafter.
At the end of the 10-week process, executives from OpenIDEO select the most pertinent proposals for a shortlist and at this point in the cycle that the researchers observed, they had data from 350 individuals and thousands of ideas they had proposed.
"Creativity studies typically have many participants solve 'toy' problems or observe few participants solving real problems," says Dr. Chan. "In our study we had both, lending greater strength to our conclusions."
They plugged the data into an algorithm that replaced what is normally a subjective human process of determining whether the idea was within-the-lines or out-of-the-box.
According to the researchers, the algorithm had been vetted against human judgments and proved to be accurate.
The researchers concluded that the vast majority of ideas selected by the OpenIDEO experts for the shortlist were not the fruits of out-of-the-box thinking but reflected a more centered approach.
The study was published in the journal Design Studies.
ALSO ON HUFFPOST: