抗作假人格迫选测验中瑟斯顿IRT模型的影响因素

发布时间：2018-12-26 14:06

【摘要】：使用蒙特卡洛(Monte Carlo)模拟研究的方法,探讨应用瑟斯顿IRT模型对抗作假迫选测验计分时需满足的编制条件,考察了测验所测特质个数(2或5个)、每维度包含陈述数量(10或20个)、单维配对题目比例(0或20%)和正负向陈述间配对题目比例(0、20%或40%)对模型估计的影响.结果如下:1)特质个数对模型估计有显著影响,测验所测特质个数越多,模型估计越准确;2)陈述数量影响模型估计,测验包含陈述数量越多,模型估计越精确,且当特质个数较少时,陈述数量的影响更大;3)测验中单维配对题目的比例基本不影响瑟斯顿IRT模型的估计精度;4)测验中加入一定比例(约20%)的正负向陈述间配对题目可提高模型的估计精度,且特质个数较少时,该因素的影响更大.最后,在研究结果的基础上给出了开发抗作假人格迫选测验的建议.
[Abstract]:By using Monte Carlo (Monte Carlo) simulation method, this paper discusses the compiling conditions to be satisfied when using Thurston IRT model to counter the false selection test, and investigates the number of traits (2 or 5) measured in the test. Each dimension includes the effects of the number of statements (10 or 20), the proportion of one-dimensional pairs (0 or 20%) and the proportion of positive and negative statements (0% or 40%) on the model estimation. The results are as follows: 1) the number of traits has a significant impact on the model estimation, the more the number of traits measured, the more accurate the model estimation; 2) the number of statements affects the estimation of the model. The more statements the test contains, the more accurate the model estimation is, and the more the number of statements is when the number of traits is small; 3) the proportion of one-dimensional paired questions in the test has little effect on the estimation accuracy of Thurston IRT model. 4) adding a certain proportion (about 20%) of positive and negative statements to the test can improve the estimation accuracy of the model, and the influence of this factor is greater when the number of traits is small. Finally, based on the results of the study, some suggestions on the development of anti-false personality forced selection test are given.
【作者单位】：北京师范大学心理学部;未来教育高精尖创新中心;
【基金】：中央高校基本科研业务费专项资金资助
【分类号】：B841.7

【相似文献】