Measuring Flexibility: A Text-Mining Approach

• In Event: Innovations in Methodological Approaches in Creativity Assessments for Gifted Student Identification

Abstract

For the last century, psychologists studying creativity have perennially highlighted the important link between individuals’ creative outcomes and the flexibility of their thinking (De Dreu et al., 2011; Georgsdottir & Getz, 2004; Zabelina & Robinson, 2010). Individuals who are capable of thinking in a flexible way are seen as more likely to produce highly creative ideas and products (Acar et al., 2019; Hass, 2017; Runco & Chand, 1995) as a result of their ability to avoid cognitive fixation and consider various directions in thinking. Despite the clear importance of flexibility in the creativity research literature, empirical observation and quantification of this construct has lagged behind other creativity-relevant attributes such as fluency and originality.
Here, we build from extant research on the measurement of originality using text-mining models to theoretically posit and then empirically validate a text-mining based method for measuring flexibility in verbal divergent thinking (DT) responses. The empirical validation of this method is accomplished in two studies. In the first study, we used data from 224 participants who completed the verbal form of the Torrance Test of Creative Thinking (TTCT). Results demonstrate that our novel flexibility scoring method strongly and positively correlated with traditionally used TTCT flexibility scores supporting convergent validity of the method.
In the second study, we examined data from 296 participants using the Alternate Uses Task. By confirming the theoretically proposed latent structure, using confirmatory factor analysis, we established internal construct validity of our scoring method. Our text-mining based flexibility scoring also achieved a high level of reliability using both Cronbach’s alpha and coefficient H. By further expanding on the theoretical examination, text-mining based flexibility scores were correlated with multiple relevant external variables. Convergent validity was examined by correlating text-mining based flexibility scores with text-mining derived fluency, originality and elaboration scores. Criterion validity was evaluated by examining artistic expertise and our flexibility scoring in relation to Big 5 personality characteristics. Findings confirmed theoretical relation between concepts and our text-mining based flexibility scoring further suggesting relevance of this method. To assist researchers in applying this flexibility scoring method, we also recommend a particular freeware text-mining program and provide specific details about how to undertake an analysis of flexibility using text-mining models.

Authors

• Katalin Grajzel, University of Denver

• Denis Dumas, University of Georgia

• Kelly Berthiaume, University of North Texas

• Selcuk Acar, University of North Texas

• Peter Organisciak, University of Denver