Introduction: The “stable marriage” algorithm underlying the National Residency Match Program (NRMP) has been shown to create optimal outcomes when students submit true preference lists. Previous research has shown students may allow external information to affect their rank lists. The objective of this study was to determine whether medical students consistently make rank lists that reflect their true preferences. Methods: A voluntary online survey was sent to third-year students at a single midwestern medical school. Students were given hypothetical scenarios that either should or should not affect their true residency preferences and rated the importance of six factors to their final rank list. The survey was edited by a group of education scholars and revised based on feedback from a pilot with current postgraduate year 1 residents. Results: Of 175 students surveyed, 140 (80%) responded; 63% (88/140) reported that their “perceived competitiveness” would influence their rank list at least a “moderate amount. Of 135 students, 31 (23%) moved a program lower on their list if they learned they were ranked “low” by that program, while 6% (8/135) of respondents moved a program higher if they learned they were ranked “at the top of the list.” Participants responded similarly (κ = 0.71) when presented with scenarios asking what they would do vs what a classmate should do. Conclusion: Students’ hypothetical rank lists did not consistently match their true residency preferences. These results may stem from a misunderstanding of the Match algorithm. Medical schools should consider augmenting explicit education related to the NRMP Match algorithm to ensure optimal outcomes for students.

中文翻译：

---

误解匹配：学生是否根据真实偏好创建排名列表？

简介：事实证明，当学生提交真实的偏好列表时，国家居住匹配计划 (NRMP) 所依据的“稳定婚姻”算法可以创造最佳结果。先前的研究表明，学生可能会允许外部信息影响他们的排名。这项研究的目的是确定医学生是否始终如一地列出反映他们真实偏好的排名列表。方法：对中西部一所医学院的三年级学生进行了自愿在线调查。学生们被给予假设的场景，这些场景应该或不应该影响他们真正的居住偏好，并评估六个因素对他们最终排名的重要性。该调查由一组教育学者编辑，并根据当前研究生一年级住院医师试点的反馈进行了修订。结果：在 175 名接受调查的学生中，140 名（80%）做出了回应；63% (88/140) 表示，他们的“感知竞争力”至少会“适度影响”他们的排名。在 135 名学生中，31 名 (23%) 的受访者如果得知自己在该项目中排名为“低”，则将其在列表中的排名降低，而 6% (8/135) 的受访者如果得知自己在该项目中排名为“低”，则将其在列表中的排名调高。在列表的顶部。” 当询问他们会做什么和同学应该做什么的场景时，参与者的反应相似（κ = 0.71）。结论：学生的假设排名列表与他们真实的居住偏好并不相符。这些结果可能源于对匹配算法的误解。医学院应考虑加强与 NRMP 匹配算法相关的显性教育，以确保学生获得最佳结果。