Department of Psychology, Nanjing University, Nanjing, Jiangsu, China.
Department of Psychology, Zhejiang Normal University, Jinhua, Zhejiang, China.
Behav Res Methods. 2024 Dec;56(8):8640-8657. doi: 10.3758/s13428-024-02495-0. Epub 2024 Sep 10.
In educational and psychological assessments, benefiting from back random response (BRR) is a major type of rapid guessing in misfitting item score patterns. Person-fit statistics (PFS) based on cumulative sum (CUSUM) and change point analysis (CPA) from statistical process control (SPC) are better than other PFS for detecting aberrant response. In this study, we developed new person-fit statistics based on three algorithms from CPA procedure and CUSUM method for detection of person misfit with dichotomous or polytomous items. By means of simulated data, the effectiveness of the new statistics to detect test-takers with BRR was investigated.
在教育和心理评估中，受益于后随机反应（BRR）是不拟合项目得分模式中快速猜测的主要类型。基于统计过程控制（SPC）中的累积和（CUSUM）和断点分析（CPA）的个体适合度统计（PFS）比其他 PFS 更适合检测异常反应。在这项研究中，我们开发了基于 CPA 过程和 CUSUM 方法的三种算法的新个体适合度统计，用于检测二项式或多项式项目的个体不匹配。通过模拟数据，研究了新统计数据检测具有 BRR 的测试者的有效性。
