Abstract
In a follow-up to a previously published study (Epstein, Schmidt, & Warfel, Citation2008), an ethnically-diverse sample of 13,578 people in 47 countries (mainly the United States and Canada) took an online test that measures 4 trainable competencies that have been shown to enhance creative expression in individuals. The new study confirmed that the test has high internal-consistency reliability and reasonably strong predictive validity. Competency scores were good predictors of self-reported frequency of creative expression and moderately good predictors of life satisfaction and professional success. Scores were also substantially higher for people who had had creativity training, and more training was associated with higher scores. Small but statistically significant differences in scores were found for gender (women outscoring men) and race/ethnicity. Of the 4 competencies examined, capturing (preserves new ideas as they occur) proved to be the best predictor of self-reported creative expression, followed by challenging (takes on difficult tasks), broadening (expands one's skills and knowledge), and surrounding (arranges stimulating physical and social environments).
Acknowledgments
This article is based on a paper presented at the 118th annual meeting of the American Psychological Association, San Diego, California, August 2010.
Notes
1Competency refers to a functionally-related set of skills that may or may not have been previously expressed. In that sense, the competency domain is far larger than the performance domain. Although the ideal way to measure competence is to sample behavior (McClelland, Citation1973), test questions that pinpoint behavior can be excellent predictors of such behavior (Boyatzis, Citation1982; Epstein et al., Citation2008; Smith & Smith, Citation2005; Spencer & Spencer, Citation1993; Wood & Payne, Citation1998). A test item such as “I'm good at recording my new ideas” would be deficient in this regard, whereas an item such as “I always keep a recording device by my bed at night” (an item from the ECCI-i) does a far better job of pinpointing behavior.
2Nonparametric statistical tests such as Spearman's rho, the Mann-Whitney U, and the Kruskal-Wallis H are used throughout this study because scores on the ECCI-i lie on an ordinal scale. The triple asterisk is used to signify a significance level (p) of less than 0.001. The double asterisk is used to signify a significance level (p) of less than 0.01. A single asterisk is used to signify a significance level (p) of less than 0.05. Unless otherwise indicated, all test scores are reported as a percentage of total correct rather than as raw scores.
Notes. Extraction Method was Principal Component Analysis. Rotation Method was Varimax with Kaiser Normalization. Factor loadings under 0.30 are not shown.