问卷调查中被试不认真作答的控制与识别  

Preventing and detecting insufficient effort survey responding

作　　者：钟晓钰 李铭尧 李凌艳 ZHONG Xiaoyu;LI Mingyao;LI Lingyan(Collaboration Innovation Center of Assessment toward Basic Education Quality,Beijing Normal University,Beijing 100875,China)

机构地区：[1]北京师范大学中国基础教育质量监测协同创新中心,北京100875

出　　处：《心理科学进展》2021年第2期225-237,共13页Advances in Psychological Science

基　　金：国家社会科学基金一般项目“基于大数据循证的学校治理现代化研究”(20BGL234)。

摘　　要：问卷调查是心理与教育领域十分常见的数据收集方法,而被试的不认真作答可能导致问卷数据失真。回顾已有研究发现:(a)不认真作答可以从外在作答模式和内在产生原因两个方向进行定义;(b)不认真作答的常见事前控制方法主要包括降低任务难度以及提高被试作答动机两大类;(c)事后识别方法主要包括嵌入识别量表、作答模式识别、反应时识别三大类。今后的研究中应基于作答机制的研究优化与开发控制方法,检验作答识别方法的跨情境适用性并开发新方法,并对局部不认真的识别与处理进行更深入的探讨。Surveys are commonly used in psychological and educational research.Insufficient effort response(IER),as one source of invalid response data,is somewhat prevalent due to the low-stakes nature of the majority of surveys,which often leads to statistically significantly biased estimates and invalid inferences.The current literature shows:(a)IER is commonly believed to be caused by some inner causes,(e.g.,low motivation),showing as specific patterns,(e.g.,random responding);(b)The most common methods to prevent IER include reducing task difficulty and increasing respondents’motivation;(c)Current detection methods fall into three main categories,which are proactive approaches/direct screening methods,response patterns analysis,and response time analysis.Recommendations for future research directions and practitioners are(a)deepening the investigation on IER mechanism and improving the preventing methods,(b)examining the effectiveness of IER identification methods’applicability of cross-situation and developing new approaches,and(c)delving into the identification and treatment of partial IER.

关 键 词：不认真作答 数据清理 无效数据 调查问卷设计 

分 类 号：B841[哲学宗教—基础心理学]