stav - Studierendenauswahlverbund
Assessing and Developing Social Skills
At the end of 2017, the German Federal Constitutional Court issued an unexpected ruling: Aspiring medical students should no longer receive a university acceptance based solely on their final school grades. Instead, other criteria, such as social factors, should also play a prominent role in the selection process. Many argued that this ruling could be implemented by assessing applicants’ social skills, such as empathy or persuasiveness. But how well can different social skills among applicants really be assessed?
We investigate this question in a project funded by the Federal Ministry of Education and Research: the stav (student selection network) project. Here, six German universities have bundled their resources to analyze various aspects of the selection process as well as the further development and training of medical students. Together with the medical faculty of the University of Münster, we lead the subproject on the assessment of social skills.
Here, we consider and evaluate existing practices aimed at assessing social skills. This includes short interview exercises, short role-play exercises (similar to classic Assessment Centers), as well as Situational Judgment Tests. Initial analyses highlighted that existing procedures were not particularly suited to measure distinct social skills. To further investigate if and how social skills can be assessed we videotaped the Münster selection procedure (i.e., interpersonal role-plays) and coded actual expressed behaviors. This made it possible to differentiate between behavioral expression and interpersonal perception of social skills.
Results highlighted the need to adapt social skill assessment. We have since developed a behavioral-focused selection procedure (based on the Multiple Speed Assessment framework) that directly aims to assess candidates’ distinct social skills. This procedure has been successfully piloted at multiple German universities and is currently (in an adapted form) used for the Landarztauswahl in NRW.
As a next step we are exploring ways of reducing cost and time constraints of classic behavioral assessments. For this, we are building on recent development in artificial intelligence and machine learning approaches that may help in providing significantly less time-consuming and more reliable selection decisions that depend less on the preferences of certain assessors.
Furthermore, we are currently working on moving beyond the pure selection context and aim to implement social skill assessment as part of the complete course of study for medical students. This would allow for a differentiated training and development of specific social skills with real and simulated patients.
For more information concerning the conceptualization, assessment, and development of social skills / social competencies see here.