形成假設,檢驗假設從而發現規則是人類進行分類、區辨、形成概念、建立因果關係等思考活動時所需具備的能力,也是科學研究的必備能力。本研究採用心理學中常用的規則發現作業-「246作業」,以實驗的方式,探討假設產生與檢驗的能力,並分析具有科學潛力的人才在此一認知能力上的特色。我們比較一般大學生與通過教育部甄選進入大學就讀的科學資優生在此作業的表現,如所預期,後者的正確率的確較高。分析其策略,他們在測試法上較一般生多採用「雙假設測試法」,而非過去研究所關注的反例法。此外,如所預測,在修正假設的過程中較一般生產生更多新角度的假設。這和我們過去發現一般大學生中成功者的特色相仿。我們討論採用雙假設測試法的可能優勢並考慮工作記憶資源的限制,據此我們指導另一組大學生採用一種「有限的雙假設測試策略」,結果一般大學生在假設檢驗推理作業上的成功率由26.7%大幅提升至81.8%。
Our research intended to find out what strategy makes a reasoner perform better in a rulediscovery task. In Experiment 1, two groups of 29 college students were tested with the 2 4 6 task, a popular rule-discovery task. One group was composed of undergraduates who were judged by experts and professors as top students in nature science and life science when they were high school students. These students subsequently were enrolled in relevant programs in National Taiwan University. The results showed that these students performed better on the task than did the contrast group, composed of students from the same university. Particularly, compared with the contrast group, these students generated more alternative hypotheses as well as applying more dual-hypothesis testing strategy. They, however, did not use negative-test strategy more often than the contrast group. We argue the way how a reasoner revised his/her hypothesis based on new evidence is crucial for the success on the 246 task. In addition, the potential advantages and constraints of using a dual-hypothesis testing strategy were discussed, as well as its relation with the reasoners' working memory. In Experiment 2, a group of university students were instructed on the limited dual-hypothesis testing developed by the researchers. Their performance in the rule-discovery task significantly improved when compared with the contrast group.c