While working on a grant from the Russian Science Foundation, researchers from South Ural State University have patented two databases for assessing creativity levels in children and adults. A neural network trained on the unique datasets delivers almost instant results for the Torrance creativity test, significantly reducing the time required for expert analysis.
The scientists collected and analysed data obtained from a sample of 3007 Russians who had previously completed the Torrance Test of Creative Thinking. For the first time in Russia, the two databases include diagnostic results on the structure of creative thinking not only in children, but in adults as well. These databases offer researchers unique opportunities: first, to develop creativity norms for adults, as there is a strong practical demand but very little data on adult creativity; and second, to train artificial intelligence systems to evaluate test responses.
“Traditionally, Torrance testing was conducted only for children and adolescents under the age of seventeen,” explained Iuliia Dmitrieva, Associate Professor at the SUSU Department of Psychology of Management and Professional Activities. “However, the world is changing rapidly, and new technologies are developing quickly, which naturally increases employers’ interest in highly creative potential employees. Without exaggeration, creativity is becoming a key competitive advantage across many professional fields, regardless of age or occupation.”
As part of the grant project, the Chelyabinsk researchers, jointly with their colleagues from the Institute of Psychology of the Russian Academy of Sciences, developed an artificial intelligence model based on the patented databases. The model processes creativity assessment results from several projective tests, simplifying and accelerating psychologists’ work. Already today, anyone can take an online creativity test using either the Urban method or the Torrance method on a dedicated Internet portal and, thanks to the trained neural network, receive results within minutes without visiting a psychologist.
In addition, within the frameworks of the grant, the researchers are developing an original methodology for comprehensive mass creativity assessment. Unlike the existing approaches, the new test takes into account the specifics of the Russian population sample and expands beyond traditional creativity indicators by including metacognitive indicators, personality traits, and personal narrative characteristics.
The test makes it possible to assess creativity levels across different age and professional groups.
“Our test consists of only three parts: the first two are drawing-based tasks suitable for children from the age of four or five, while the third part requires participants to invent a story about real or fictional characters, including characters from books or films,” said Vera Griazeva-Dobshinskaia, Professor at the SUSU Department of Psychology of Management and Professional Activities.
The traditional Torrance Test of Creative Thinking evaluates six key characteristics: originality, fluency, flexibility, elaboration, abstractness of titles, and resistance to premature closure.
The new test being developed by the Chelyabinsk researchers assesses not only these six universal creativity factors, but also additional metacognitive indicators (self-transcendence as supra-situational activity, cultural code indicators, and dynamics indicators in test series), personality traits (creative self-efficacy and personal identity), and personal narrative indicators (ego positions, drivers, and types of personal choice).
The original test and the existing neural network can be used for large-scale creativity assessment, for example when selecting students for educational institutions at all levels or employees for organizations in various sectors.