Department of Psychology, McGill University, 2001 McGill College Avenue, Montreal, QC, H3A 1G1, Canada.
Department of Psychology, University of Toronto, Toronto, Canada.
Cogn Res Princ Implic. 2023 Mar 9;8(1):17. doi: 10.1186/s41235-023-00472-3.
Relating learned information to similar yet new scenarios, transfer of learning, is a key characteristic of expert reasoning in many fields including medicine. Psychological research indicates that transfer of learning is enhanced via active retrieval strategies. For diagnostic reasoning, this finding suggests that actively retrieving diagnostic information about patient cases could improve the ability to engage in transfer of learning to later diagnostic decisions. To test this hypothesis, we conducted an experiment in which two groups of undergraduate student participants learned symptom lists of simplified psychiatric diagnoses (e.g., Schizophrenia; Mania). Next, one group received written patient cases and actively retrieved the cases from memory and the other group read these written cases twice, engaging in a passive rehearsal learning strategy. Both groups then diagnosed test cases that had two equally valid diagnoses-one supported by "familiar" symptoms described in learned patient cases, and one by novel symptom descriptions. While all participants were more likely to assign higher diagnostic probability to those supported by the familiar symptoms, this effect was significantly larger for participants that engaged in active retrieval compared to passive rehearsal. There were also significant differences in performance across the given diagnoses, potentially due to differences in established knowledge of the disorders. To test this prediction, Experiment 2 compared performance on the described experiment between a participant group that received the standard diagnostic labels to a group that received fictional diagnostic labels, nonsense words designed to remove prior knowledge with each diagnosis. As predicted, there was no effect of diagnosis on task performance for the fictional label group. These results provide new insight on the impact of learning strategy and prior knowledge in fostering transfer of learning, potentially contributing to expert development in medicine.
将所学知识应用于类似但全新的场景中,即学习迁移,是医学等许多领域专家推理的一个关键特征。心理学研究表明,通过主动检索策略可以增强学习迁移。对于诊断推理,这一发现表明,主动检索关于患者病例的诊断信息可以提高将学习迁移应用于后续诊断决策的能力。为了验证这一假设,我们进行了一项实验,两组本科学生参与者学习了简化精神病诊断的症状列表(例如,精神分裂症;躁狂症)。接下来,一组参与者收到书面病例并主动从记忆中检索病例,另一组参与者则将这些书面病例阅读两遍,采用被动复述学习策略。然后,两组参与者都对测试病例进行诊断,这些病例有两个同样有效的诊断——一个由学习过的病例中描述的“熟悉”症状支持,另一个由新的症状描述支持。虽然所有参与者更有可能将更高的诊断概率分配给那些由熟悉症状支持的病例,但与采用被动复述相比,采用主动检索的参与者的这种效果明显更大。在给定的诊断中也存在显著的性能差异,这可能是由于对疾病的已有知识的差异。为了验证这一预测,实验 2 比较了接受标准诊断标签的参与者组和接受虚构诊断标签(旨在去除每个诊断的先验知识的无意义单词)的参与者组在描述性实验中的表现。正如预测的那样,对于虚构标签组,诊断对任务表现没有影响。这些结果为学习策略和先验知识在促进学习迁移方面的影响提供了新的见解,可能有助于医学领域的专家发展。