連韻文2006-07-252018-06-282006-07-252018-06-282001http://ntur.lib.ntu.edu.tw//handle/246246/9678近年來我們發現影響假設檢驗推理作業成敗的關鍵在於是否能夠產生新角度的假設。在本研究中，我們進一步澄清與過去文獻與我們的發現不一致之處，藉以探討上述發現的普遍性。兩個實驗分別探討規則難易度與不同形態的作業的影響。We found that, rather than searching for disconfirmatory information, the ability to view the data from a different viewpoint was more critical for better performance in a hypothesis-testing task (246 task). The current research discuss whether the above finding would be hold for problems with various levels of difficulty and those rule-discovery tasks with different format, in which too many alternative hypotheses waited to be ruled out (e.g., concept learning tasks composed of visual features.)application/pdf202763 bytesapplication/pdfzh-TW國立臺灣大學心理學系暨研究所reporthttp://ntur.lib.ntu.edu.tw/bitstream/246246/9678/1/892413H002040.pdf