Insufficient effort responding (IER) to surveys, which occurs when participants provide responses in a haphazard, careless, or random fashion, has been identified as a threat to data quality in survey research because it can inflate observed relationships between self-reported measures. Building on this discovery, we propose two mechanisms that lead to IER exerting an unexpected confounding effect between self-reported and informant-rated measures. First, IER can contaminate self-report measures when the means of attentive and inattentive responses differ. Second, IER may share variance with some informant-rated measures, particularly supervisor ratings of participants’ job performance. These two mechanisms operating in tandem would suggest that IER can act as a “third variable” that inflates observed relationships between self-reported predictor scores and informant-rated criteria. We tested this possibility using a multisource dataset ( N = 398) that included incumbent self-reports of five-factor model personality traits and supervisor-ratings of three job performance dimensions—task performance, organizational citizenship behavior (OCB), and counterproductive work behavior (CWB). We observed that the strength of the relationships between self-reported personality traits and supervisor-rated performance significantly decreased after IER was controlled: Across the five personality traits, the average reduction of magnitude from the zero-order to partial correlations was |.08| for task performance, |.07| for OCB, and |.14| for CWB. Because organizational practices are often driven by research linking incumbent-reported predictors to supervisor-rated criteria (e.g., validation of predictors used in various organizational contexts), our findings have important implications for research and practice.

中文翻译：

---

跨调查来源回应不足的混杂效应：性格预测表现的案例

当参与者以随意、粗心或随机的方式提供答复时，就会发生对调查的努力响应不足 (IER)，它已被认为是对调查研究中数据质量的威胁，因为它可能会夸大自我报告措施之间观察到的关系。基于这一发现，我们提出了两种机制，导致 IER 在自我报告和线人评级测量之间产生意想不到的混杂效应。首先，当专心和不专心的反应方式不同时，IER 可能会污染自我报告措施。其次，IER 可能与一些信息提供者评分的指标存在差异，特别是主管对参与者工作绩效的评分。这两种机制协同作用表明，IER 可以充当“第三个变量”，夸大自我报告的预测分数和线人评分标准之间观察到的关系。我们使用多源数据集（N = 398）测试了这种可能性，其中包括五因素模型人格特质的现任自我报告以及三个工作绩效维度的主管评分——任务绩效、组织公民行为（OCB）和适得其反的工作行为（CWB）。我们观察到，在控制 IER 后，自我报告的人格特质与主管评分绩效之间的关系强度显着下降：在五种人格特质中，从零级相关到偏相关的平均降低幅度为 |.08|。对于任务绩效，|.07| 对于 OCB 和 |.14| 为 CWB。由于组织实践通常是由将现任报告的预测因素与主管评级标准联系起来的研究驱动的（例如，在各种组织环境中使用的预测因素的验证），因此我们的研究结果对研究和实践具有重要意义。