Marcus Bernd, Schütz Astrid
Psychologisches Institut, Chemnitz University of Technology, Germany.
J Pers. 2005 Aug;73(4):959-84. doi: 10.1111/j.1467-6494.2005.00335.x.
This article outlines a typology of nonresponse biases that may occur in research based on voluntary participation and presents empirical evidence on the relationships between nonresponse and personality traits. We describe four different types of nonresponse occurring at the recruitment stage, during data collection, and at the stage of recruitment for further research. Personality correlates of all types of nonresponse were investigated using data from an online survey e-mailed to owners of personal Web sites, and observer ratings on the targets' personality were obtained. After reviewing the target persons' Web sites, observers judged complete nonrespondents as less agreeable and less open to experience than respondents. Furthermore, higher extraversion, higher openness to experience, and higher narcissism differentiated volunteers for follow-up research from nonvolunteers in both self- and observer ratings. We found only weak evidence of personality correlates of careless and selective item nonresponse. Findings show that nonresponse biases may have significant implications for representativeness in surveys and personality test norms.
本文概述了基于自愿参与的研究中可能出现的无应答偏差类型，并给出了关于无应答与人格特质之间关系的实证证据。我们描述了在招募阶段、数据收集期间以及进一步研究的招募阶段出现的四种不同类型的无应答。使用发送给个人网站所有者的在线调查数据，研究了所有类型无应答的人格相关性，并获得了对目标对象人格的观察者评分。在浏览目标对象的网站后，观察者判断完全无应答者比应答者更不随和、更不愿接受新体验。此外，在自我评分和观察者评分中，较高的外向性、较高的接受新体验能力以及较高的自恋倾向使后续研究的志愿者与非志愿者区分开来。我们仅发现粗心和选择性项目无应答与人格相关的微弱证据。研究结果表明，无应答偏差可能对调查的代表性和人格测试规范具有重大影响。
