Tags: Creativity Research · Psychological Assessment · Divergent Thinking
The Torrance Tests of Creative Thinking (TTCT) has been, for decades, the most widely used creativity assessment instrument in the world. Its conceptual architecture rests on a well-established premise in the literature: creativity can be operationalised through four fundamental cognitive dimensions — fluency, flexibility, originality, and elaboration — first formulated by Guilford (1967) and adopted by Torrance as the theoretical and evaluative foundation of his battery.
Yet a 2008 article by Leandro Almeida, Lola Prieto, Mercedes Ferrando, Emma Oliveira, and Carmen Ferrándiz raises an uncomfortable question: does this structure hold empirically when the TTCT is applied in cultural contexts beyond the North American one? The data collected in Spain and Portugal suggest that, in large measure, it does not.
The Test’s Logic and Its Structural Assumption
The TTCT exists in two parallel versions: verbal and figural. Both are designed to assess the same four cognitive dimensions across different formats and task types. The construct validity hypothesis underlying the test is relatively clear: if fluency, flexibility, originality, and elaboration are stable cognitive functions that define a subject’s creativity, then those dimensions should emerge as consistent, replicable factors across the instrument’s various subtasks — independently of the specific format of each task.
Put differently: if the test measures what it claims to measure, the factors obtained through factor analysis should cluster around cognitive dimensions, not around tasks. A creative subject should demonstrate originality when completing a drawing just as when generating unusual uses for an object, because originality would be a property of their thinking — not of the task they happen to be performing.
What the Data Revealed
The authors present results from three independent empirical studies, conducted with samples of children and pre-adolescents in Spain and Portugal (N = 649, N = 595, and N = 310, respectively). In all three cases, versions of the verbal and figural TTCT forms were administered and principal components factor analyses with varimax rotation were conducted.
The results are mutually consistent — and markedly at odds with theoretical expectations. In none of the three studies did the identified factors correspond to the cognitive dimensions postulated by Guilford and Torrance. The factors that emerge from the analysis do not group fluency scores across subtasks, nor flexibility scores, nor originality scores. Instead, factors organise around the specific subtasks of the test: Factor I clusters all scores from Subtask 3, Factor II from Subtask 2, and so on.
The only dimension that shows any cross-task consistency is elaboration — precisely the dimension considered in the specialist literature to be the least central to the definition of creativity. The supposedly nuclear dimensions — fluency, flexibility, and originality — fail to articulate stable factors that transcend the specificity of each task.
What This Finding Implies
The authors’ interpretation is direct: what determines subjects’ performance on the TTCT is less their creative cognitive capacities than the specific demands, format, and content of each task. In other words, the instrument largely measures the ability to solve that particular type of task — not a general creative capacity that would express itself consistently across different assessment contexts.
This does not necessarily imply that the TTCT lacks value as an assessment tool, but it does call into question the assumption that grants it theoretical coherence: that the four cognitive dimensions constitute a stable, generalisable structure of creativity. If that structure does not hold in the data, the construct validity of the instrument is compromised — at least when applied to child and pre-adolescent populations outside the North American context.
The authors identify two complementary hypotheses that warrant further investigation. The first concerns the age factor: it is possible that the consistency of cognitive dimensions is greater in adult subjects, for whom the test may be better calibrated. The second concerns task design: if the specific demands of each subtask distort the expression of cognitive dimensions, it may be necessary to develop more neutral tasks that allow a cleaner expression of those functions.
A Broader Question About Creativity Assessment
Almeida and colleagues’ article is part of a broader debate about the measurability of creativity as a psychological construct. Measuring creativity requires assuming that there exists something stable, generalisable, and quantifiable that we call “creative thinking” — something that expresses itself recognisably across different contexts. The data presented here complicate that assumption: if performance on creativity tasks is so strongly determined by the characteristics of the task itself, then what tests measure may be far more situated and context-dependent than the psychometric tradition has been willing to acknowledge.
For those working in the field of creative development — whether in education, coaching, or research — this finding carries practical implications. Scores on creativity tests should be interpreted with caution, particularly when used as indicators of a general, transferable capacity. Creativity, as Vygotski suggested from a different theoretical vantage point, may systematically resist the tools designed to capture it under standardised, decontextualised conditions.
Original source: Almeida, L. S., Prieto, L. P., Ferrando, M., Oliveira, E., & Ferrándiz, C. (2008). Torrance Test of Creative Thinking: The question of its construct validity. Thinking Skills and Creativity, 3(1), 53–58. https://doi.org/10.1016/j.tsc.2008.03.003