Measuring Political Knowledge and Not Search Proficiency in Online Surveys
Measuring Political Knowledge and Not Search Proficiency in Online Surveys
复制标题
在在线调查中衡量政治知识而不是搜索能力
DOI：
10.1093/ijpor/edac002
复制
发表时间：
2022
影响因子：
1.8
通讯作者：
DeBell, Matthew
中科院分区：
文献类型：
--
作者：
DeBell, Matthew
大多数测试政治知识的在线调查问题在参与者查阅答案时都容易出现测量误差。本文报告了五种检测和防止这种常见误差源的方法的研究。为了检测查找，“捕捉问题”比自我报告更可靠，因为许多参与者撒谎，而不是承认查找答案。措辞强硬的指令减少了约三分之二的查找次数，而指令的三重组合--要求承诺不查找答案，以及适应性反馈(要求查找答案的参与者停止这样做)--将查找答案的受访者比例进一步降低了一半，至3%。对于办公室回忆知识项目，基于照片的开放式问题消除了查找，并具有与传统的基于文本的版本相似的有效性，使其在视觉格式可行时成为一个很好的选择。
Most online survey questions testing political knowledge are susceptible to measurement error when participants look up the answers. This article reports five studies of methods to detect and prevent this common source of error. To detect lookups, “catch questions” are more reliable than self-reports, because many participants lie rather than admit looking up answers. Strongly worded instructions reduced lookups by about two-thirds, while the triple combination of instructions, requesting a promise not to look up answers, and adaptive feedback (asking participants who look up an answer to stop doing so) reduced the percentage of respondents looking up an answer by a further half, to 3%. For office recall knowledge items, photo-based open-ended questions eliminated lookups and had similar validity to traditional text-based versions, making them a good choice when a visual format is viable.