The COVID-19 pandemic and the consequent introduction of virtual collaboration has introduced educators to unexpected situations and challenges. One of these challenges is social distance, which minimizes knowledge on another person’s character, and leaves room to misconceptions. Perceptions on a person’s personality are also referred to as dispositional attributions, and when misplaced, lead to alternative educator-student dynamics. This paper studies dispositional attributions exhibited by software engineering educators in higher education and aims at raising awareness of potential misconceptions affecting the educator-student relationship caused by the virtual setting. We performed a case study in a practical university course with twelve distributed software engineering teams, each led by one or two educators. The course was conducted entirely virtually during the COVID-19 pandemic. The research process included discovering, categorising, and modelling attribution-based personas preceded by qualitative and quantitative research methods of semi-structured interviews and survey questionnaires. These personas represent the subjects of potential misconceptions, and encapsulate typical behaviours and attributions. Our research resulted in the creation of seven personas: the Unprofessional, Ego is the Enemy, L’Etranger, the Loner, the Underperformer, Hiding but not Seeking and Distraction Monster. These personas differ primarily in terms of character traits and motivation attributed to them. The results provide evidence that the virtual setting of the course can lead to several dispositional attributions. Educators in virtual software engineering settings should be aware of these attributions and their potential impact on the educator-student relation.