Creativity is one of the ideas we claim to value in our students, rather than just rote memorization of trivial facts, but how do we identify creativity in a student’s work? Can an automated scoring system, an artificially intelligent one, help us with this identification or evaluation?

The question has eluded researchers and educators alike since we became aware that students’ creativity was important in their growth and development. When I worked in the Fellows Program at the John D and Catherine T MacArthur Foundation in Chicago, which is the group that awards a couple dozen “genius grants” every year, we used to profess that creativity was hard to measure, mainly because it came in many forms. Artists had a few forms, scientists a few other forms, businesspeople use a few others, and there’s just a tangled web of overlap.
“Creativity is like comedy,” we used to say in the program. “It loses meaning in the definition, but some things make you laugh while others don’t.”
Now a new study out of the University of Kent in the UK, using computer programs to analyze the language people use when they talk about the creative process or creativity itself, has identified 14 general themes that may be at the root of creativity:
- Active involvement & persistence
- Dealing with uncertainty
- Domain competence
- General intellect
- Generating results
- Independence & freedom
- Intention & emotional involvement
- Originality
- Progression & development
- Social interaction & communication
- Spontaneity & subconscious processing
- Thinking & evaluation
- Value
- Variety, divergence, & experimentation
Dr Anna Jordanous, lecturer in the School of Computing at the University of Kent, and language expert Dr Bill Keller of the University of Sussex published their findings, which made use of language-analysis software to analyze and group words associated with creativity into the 14 clusters identified above, in the journal PLOS ONE.
They both said they hope the work might one day form a basis for the “automated evaluation of creative systems,” such as systems that explore the creativity many students show in the work they do all the time.
Or, we could just use teachers to evaluate this work, but as we all know, the opinions of teachers regarding the work their own students produce, can be subject to biases of various sorts.