Department of Emergency Medicine, Inselspital University Hospital Bern, Bern, Switzerland.
Centre for Educational Measurement, Faculty of Educational Sciences, University of Oslo, Oslo, Norway.
Med Educ. 2019 Jul;53(7):735-744. doi: 10.1111/medu.13801. Epub 2019 Feb 14.
The ability to self-monitor one's performance in clinical settings is a critical determinant of safe and effective practice. Various studies have shown this form of self-regulation to be more trustworthy than aggregate judgements (i.e. self-assessments) of one's capacity in a given domain. However, little is known regarding what cues inform learners' self-monitoring, which limits an informed exploration of interventions that might facilitate improvements in self-monitoring capacity. The purpose of this study is to understand the influence of characteristics of the individual (e.g. ability) and characteristics of the problem (e.g. case difficulty) on the accuracy of self-monitoring by medical students.
In a cross-sectional study, 283 medical students from 5 years of study completed a computer-based clinical reasoning exercise. Confidence ratings were collected after completing each of six cases and the accuracy of self-monitoring was considered to be a function of confidence when the eventual answer was correct relative to when the eventual answer was incorrect. The magnitude of that difference was then explored as a function of year of seniority, gender, case difficulty and overall aptitude.
Students demonstrated accurate self-monitoring by virtue of giving higher confidence ratings (57.3%) and taking a shorter time to work through cases (25.6 seconds) when their answers were correct relative to when they were wrong (41.8% and 52.0 seconds, respectively; p< 0.001 and d > 0.5 in both instances). Self-monitoring indices were related to student seniority and case difficulty, but not to overall ability or student gender.
This study suggests that the accuracy of self-monitoring is context specific, being heavily influenced by the struggles students experience with a particular case rather than reflecting a generic ability to know when one is right or wrong. That said, the apparent capacity to self-monitor increases developmentally because increasing experience provides a greater likelihood of success with presented problems.
在临床环境中自我监测表现的能力是安全有效实践的关键决定因素。多项研究表明，这种自我调节形式比在给定领域对自己能力的综合判断（即自我评估）更值得信赖。然而，人们对学习者自我监测的信息提示知之甚少，这限制了对可能有助于提高自我监测能力的干预措施的深入探讨。本研究的目的是了解个体特征（例如能力）和问题特征（例如案例难度）对医学生自我监测准确性的影响。
在一项横断面研究中，来自 5 个年级的 283 名医学生完成了一项基于计算机的临床推理练习。在完成每个案例后收集信心评分，并且当最终答案正确时，自我监测的准确性被认为是信心的函数，相对于最终答案错误时的信心。然后，探讨了这种差异的幅度作为年级、性别、案例难度和整体能力的函数。
学生通过在答案正确时给予更高的信心评分（57.3%）和在解决案例时花费更少的时间（25.6 秒）来展示准确的自我监测，而在答案错误时分别为 41.8%和 52.0 秒（p<0.001，在两种情况下 d>0.5）。自我监测指标与学生的年级和案例难度有关，但与整体能力或学生性别无关。
这项研究表明，自我监测的准确性是特定于情境的，受学生在特定案例中遇到的困难的严重影响，而不是反映出普遍的知道自己正确或错误的能力。也就是说，自我监测的能力随着时间的推移而发展，因为经验的增加增加了在呈现的问题上取得成功的可能性。
