圖形設計作業創造力指標的發展與分析：行為及神經科學取向

Abstract

設計流暢性（design fluency）或非語文流暢性（non-verbal fluency）是個體流暢地產出不重複的圖形設計能力，是執行功能的一種表現，同時也是擴散性思考的指標之一。然而，擴散性思考不只有流暢性，還有變通性與獨創性。過往的圖形設計作業並沒有這兩個指標。本研究旨在發展並分析圖形設計作業的創造力指標。預備研究開發標準數位化「圖形設計作業（Figural Design Test, FDT）」。該測驗共三個情境：流暢情境、抑制情境、轉換情境。計算情境間正確設計數可得流暢（流暢情境的正確設計數）、抑制（在抑制情境下所能產出的正確設計比例，以流暢情境作為基準值）、轉換（在轉換情境下所能產出的正確設計比例，以流暢、抑制情境作為基準值）等指標。298位參與者被施測FDT以考驗信度，結果顯示具有良好的內部一致性信度。其中取樣118位參與者並完成效度工具（類別流暢性、轉換西蒙測驗），結果顯示流暢指標與語文流暢性表現有顯著正相關；抑制指標、轉換指標各別與轉換西蒙作業之抑制指標、轉換指標有顯著正相關。上述結果證明FDT能有效測量設計流暢、抑制、轉換等指標。研究一旨在發展圖形設計作業的設計變通性（設計方式的類別變化）、設計獨創性（少有的設計反應）指標。302位參與者完成了FDT。創造力指標是基於參與者在情境一的反應所發展。情境一有200張符合規則的圖形。本研究整合旋轉與鏡像等特徵的相似圖形，最終可畫出21種圖形。在設計變通性中，將21種圖形依外型分成八種類別，並檢驗設計變通性八成分模式。結果顯示，具有良好的整體適配度，但內在結構品質不佳，可能受圖形出現頻率影響。設計變通性指標的得分為所有正確設計涵蓋的類別數。分數範圍介於0～8分。在設計獨創性中，依參與者的每個正確設計在200張設計常模的出現頻率給分（小於2%得2分、2～5%得1分、大於5%得0分），再加總成設計獨創性總分，分數範圍介於0～198分。研究一發展了圖形設計作業的設計變通性與設計獨創性指標，能客觀且量化的評估設計變通性與設計獨創性。研究二旨在分析圖形設計作業創造力指標與典型創造力作業的相關性。135位參與者完成了FDT以及語文版與圖形版的擴散性思考測驗與遠距聯想測驗，共五個測驗。結果顯示設計流暢性與圖形擴散性思考的流暢性有正相關；設計變通性與語文及圖形擴散性思考的變通性有正相關；設計獨創性與圖形版擴散性思考的獨創性有正相關。此外，設計流暢性與語文版及圖形版遠距聯想測驗皆有正相關。上述結果顯示設計流暢性、設計變通性、設計獨創性之間具有區辨效度。支持設計變通性、設計獨創性指標的有效性。研究三以圖形理論分析大腦白質網路之結構，並檢驗拓樸屬性指標與FDT五個指標（設計流暢性、設計變通性、設計獨創性、抑制、轉換）的相關。72位參與者完成了FDT並完成擴散性張量影像的掃描。結果顯示設計流暢性、設計獨創性與群聚係數、小世界屬性、全局效率、局部效率呈顯著正相關，與特徵路徑長度呈顯著負相關；而轉換指標與群聚係數、小世界屬性、全局效率呈顯著負相關。在節點效率中，設計流暢性與雙側額中迴、右側額下迴的節點效率有正相關，設計變通性與右側額下迴的節點效率有正相關，設計獨創性與左側額內迴、右側顳中迴、左側顳上迴、雙側中央前迴，且設計流暢性與設計獨創性共享雙側額上迴、右側顳上迴的大腦節點。抑制指標和轉換指標與大腦節點效率之間並無顯著相關性。上述結果以生理證據支持設計流暢性、設計變通性、設計獨創性的有效性。本研究之貢獻：在研究工具上，發展電腦化施測、自動化計分的圖形設計作夜，並擴充其他創造力指標，可視為有正確解答的擴散性思考測驗；同時能評估創造力與執行功能的重要指標。在理論層面上，指出設計流暢性可能是創意思考的基礎，而設計變通性和設計獨創性則與更高層次思考歷程有相關。整體而言，本研究結果對於創造力的理論探究與評估應用皆有所貢獻。

Design fluency or non-verbal fluency refers to an individual’s ability to generate a fluent and non-repetitive series of figural designs. It is considered an executive function and a measure of divergent thinking. However, previous tests did not include these two indicators (flexibility and originality). The purpose of this study was to develop and analyze the effectiveness of creative indicators. A preliminary study was conducted to develop a standardized digital test called the Figural Design Test (FDT). The FDT included three conditions: fluency, inhibition, and shifting. Indicators measured were design fluency (performance in fluency condition), inhibition (proportion of correct designs relative to fluency condition), and shifting (proportion of correct designs relative to fluency and inhibition conditions). The FDT’s reliability was tested with 298 participants, showing good internal consistency reliability. Validity was examined in 118 participants through category fluency and Shifting Simon Task, revealing significant positive correlations between design fluency and verbal fluency, and between FDT inhibition/shifting and Simon Task indicators. These results support the reliability and validity of the FDT in measuring design fluency, inhibition, and shifting.Study 1 aimed to develop indicators for design flexibility (category variation in design) and design originality (rare design responses) in the FDT. 302 participants completed the FDT. The indicators were developed based on participants’ responses in Condition 1. Condition 1 consisted of 200 rule-conforming figures. This study integrated figures with similar features such as rotations and reflections, resulting in a total of 21 different designs. For design flexibility, the 21 designs were categorized and an eight-factor model of design flexibility was examined. The results showed a good overall fit, but the internal structure was poor, possibly influenced by the frequency. The design flexibility indicator score was calculated as the number of categories covered by correctly generated designs, ranging from 0 to 8. For design originality, participants' scores were determined by assigning points to each correct design based on its frequency of occurrence in a normative sample of 200 designs (2 points for less than 2%, 1 point for 2-5%, and 0 points for above 5%). The total score for design originality was the sum of these individual design scores, ranging from 0 to 198. Study 1 developed indicators of design flexibility and design originality for the FDT, enabling an objective and quantifiable assessment of these aspects. Study 2 aimed to analyze the correlation between creativity indicators in the FDT and typical creativity tasks. 135 participants completed the FDT, as well as the verbal and figural versions of the divergent thinking (DT) test and remote associates test (RAT), totaling five tests. The results showed a positive correlation between design fluency and fluency in figural DT. Design flexibility correlated positively with flexibility in both verbal and figural DT. Design originality showed a positive correlation with originality in the figural DT. Additionally, design fluency correlated positively with both the verbal and figural versions of the RAT. The above results indicate the discriminant validity between design fluency, design flexibility, and design originality. This supports the effectiveness of the indicators for design flexibility and design originality.Study 3 aimed to analyze the structural characteristics of brain white matter networks using graph theory and examine the correlations between topological properties indicators and five indicators of FDT (design fluency, design flexibility, design originality, inhibition, and shifting). A total of 72 participants completed the FDT and underwent diffusion tensor imaging scans. The results showed significant positive correlations between design fluency, design originality, and clustering coefficient, small-world attributes, global efficiency, and local efficiency, and a significant negative correlation with characteristic path length. The shifting indicator showed a significant negative correlation with clustering coefficient, small-world attributes, and global efficiency. In node efficiency, design fluency showed a positive correlation with node efficiency in the bilateral middle frontal gyrus and right inferior frontal gyrus. Design flexibility showed a positive correlation with node efficiency in the right inferior frontal gyrus. Design originality showed positive correlations with node efficiency in the left medial frontal gyrus, right middle temporal gyrus, left superior temporal gyrus, and bilateral precentral gyrus. There were no significant correlations between the inhibition and shifting indicators and node efficiency in the brain. The above results provide physiological evidence supporting the effectiveness of creative indicators.Contribution of the study: In the research tools, the development of computerized administration and automated scoring for the figural design test, along with the expansion of other creativity indicators, can be seen as a divergent thinking test with correct answers. It also allows for the assessment of important indicators of creativity and executive functioning. On a theoretical level, the study suggests that design fluency may serve as the foundation for creative thinking, while design flexibility and design originality are associated with higher-level cognitive processes. Overall, the findings of this study make contributions to both the theoretical exploration and practical application of creativity.