Simulation Study of test assembly proportion in multiple problem-solving strategies

• Main Authors: Yu-Wei Chen, 陳又維
• Other Authors: Bor-Chen Kuo
• Format: Others
• Language: zh-TW
• Published: 2015
• Online Access: http://ndltd.ncl.edu.tw/handle/xqu4aa

碩士 === 國立臺中教育大學 === 教育資訊與測驗統計研究所 === 103 === This study aims to discuss problem-solving strategies in multiple cognitive diagnosis model, in the different proportion of multiple problem-solving strategies item for the estimated impact of cognitive diagnosis model, which also discuss design the multiple problem-solving strategies Q matrices.Through research and the results of the simulation data for comparison and analysis, with the proportion of test assembly to achieve the best estimate results. According to the research results of the analysis, results are as follows: A、Multiple problem-solving strategies item in different proportions, MS-DINA model correct classification rate are better than DINA model. B、On multiple problem solving strategies Q matrix, "repeat the cognitive attribute ' is designed for multiple problem-solving strategies item have a good estimate of the effect. C、When the entire test is the multiple problem-solving strategies, there is the best estimate of the effect for all item, and also found in multiple problem-solving strategies in the item of 60% to 80% of the total number of questions, there are good estimates of the effect. D、No matter what item parameters are, Multiple-problem solving strategies item can be added to inprove the diagnostic effect, but poor item parameter will decreased the recognition rate