詞彙聯想策略測驗的發展

Abstract

摘 要 自Guilford於1950年呼籲研究者重視創造力以來，此後五十年創造力研究即開始蓬勃發展。然由於許多研究者皆從不同角度來研究創造力，因此至今對於創造力仍未有一共識之定義。 至今最常為人所使用之創造力測量工具為擴散性思考測驗。然而有許多研究者對擴散性思考作業提出質疑，尤其是在測量方式以及獨創性計分方式上。再加上國內之擴散性思考作業常模樣本多已年代久遠，且多為「概念流暢力」的測量，而尚無「聯想流暢力」的擴散性思考工具，因此於本研究即試圖結合遠距聯想概念與擴散性思考模式，發展「詞彙聯想策略測驗」，建立流暢性、變通性與獨創性分數。在測驗中，主要有包含兩個作業，分別為「海洋」與「玫瑰」之聯想作業，要求參與者根據「海洋」或「玫瑰」進行直接聯想，並寫出所聯想的詞彙與「海洋」或「玫瑰」之間的關係為何。透過參與者所填寫的關連說明，可以排除掉不適切的反應，以有效考量想法的適切性。因此流暢性分數即以「有效聯想反應的個數」來代表；另外，研究者透過參與者所回答之關連說明，建立了十一種詞彙聯想策略，作為變通性分數的類別。而在獨創性計分上，研究者進一步去比較傳統常模分數、加權平均數與加權中位數的三種獨創性計分方式，以找出最能代表獨創性指標之計分。 在信度分析上，一共進行了內部作業信度、極端組比較與評分者間信度。研究結果顯示，流暢性與變通性與傳統常模分數、加權平均數與加權中位數皆具有良好之信度。然而，傳統常模分數與流暢性分數相關高 (r = 0.542)，顯示容易受到流暢性分數之影響；而加權平均數與加權中位數則不會受到流暢性分數之影響，較能有效測量獨創性之指標。 而效度資料上，分別進行了四種建構效度 (同時性效度、同儕評定效度、前半與後半之比較、聯想策略) 與兩種區辨效度 (與頓悟性問題及新編中文遠距聯想量表之關係)。結果顯示，流暢性與變通性分數在所有效度資料上，皆具有良好之結果，雖然兩者之間的相關很高 (r = .621)，顯示變通性容易受到流暢性分數之影響。然而其與同時性效度的新編創造思考測驗仍具有顯著之相關 (r= .314)，顯示詞彙聯想策略測驗的變通性分數仍能有效測得「產生不同聯想反應」的能力。 進一步比較三種獨創性不同計分方式，結果顯示，傳統常模分數與新編創造思考測驗的同時性效度不高 (r = .254)；而加權平均數，除了與新編創造思考測驗之同時性效度結果不佳 (r = .153) 外，聯想策略團體差異的建構效度也不好；然而加權中位數在各個效度分析上，皆具有良好之結果，且不會受到流暢性分數的影響，能有效代表獨創性之指標，因此建議應以加權中位數來取代傳統常模分數，作為獨創性分數的指標。 整體而言，詞彙聯想策略測驗具有良好之信度與效度，能有效測量擴散性思考以及遠距聯想能力，可作為創造潛能之有效測量工具。Abstract The researches on creativity have been prospering for around 50 years since Guilford mentioned its importance in 1950. However, due to different research paradigms, there is no one definition of creativity that is accepted by all researchers. Since the divergent thinking test, the most commonly used measuring tool nowadays, has been criticized for its measuring and scoring procedures, and since the domestic divergent thinking task focuses largely on measuring fluency with its norm being out of date, this research aimed to develop a Word Association Strategy Test by combining the concepts of remote association and divergent thinking. The test is composed of two association subtests, “the ocean” and “the rose”. Participants were asked to associate “the ocean” or “the rose” with something in thinking and describe their relations, through examining which the researcher set up three parts of scoring. The fluency score is represented by the number of valid associations excluding the inappropriate responses. The flexibility score is represented by the number of strategies used in producing word associations. And the originality score is set up by comparing three scoring procedures, including the traditional norm score, mean and median of the weights. In reliability analysis, the researcher conducted intertask reliability, extreme groups comparing, and inter-raters reliability. It is demonstrated that the fluency and flexibility and three kinds of originality scores are fairly reliable. And mean and median of the weights are more effective measuring index for originality than the traditional norm score which is easily influenced by the fluency score. In validity analysis, the researcher conducted four kinds of construct validity, and two kinds of discriminative validity. It is demonstrated that the fluency and flexibility scores are fairly valid. Though the high correlation between the two scores shows that the flexibility score is easily influenced by the originality score, the high correlation between the flexibility score and the revised creative thinking test also shows that the flexibility score is still valid on measuring different association responses. By comparing the three scoring procedures for setting up the originality score, the researcher found that with fair reliability and validity, the median of weights is the most representative score for originality. Totally speaking, the Word Association Strategy Test can effectively measure individuals’ divergent thinking and remote association ability with fair reliability and validity.