It is common to think that examination results are based on many factors, both in the framework conditions for the student and in relation to the close meeting between student, teacher, and the teaching material. The relationship is complex and is the basis for why the student's achievements can be explained by conditions of the framework, e.g., conditions in the administrative follow-up of students (so-called framework quality) and by conditions in the teaching program the students have encountered, the so-called teaching quality. In the study quality model shown in Fig. 2 (Handal, 1990), we see that the examination result should be seen in relation to the admission quality, the student’s prior knowledge, the quality of the results, the relationship between learning outcomes in the courses and the student’s prerequisites for succeeding in exams.
Several factors described above can and should be seen together with the number of students taking a course, and it is common to think that large courses provide weaker conditions for following up each individual student, which the result in our study underlines. According to research by Iglesias et al. (2020), larger student-to-class ratios result in a less favorable social climate in MBA programs' academic and social environments. Additionally, they discovered that having a large class size raises the dropout rate. Acknowledging that a less favorable social atmosphere may impact the group's learning atmosphere and, consequently, the student's academic performance is imperative. Thus, the program quality will be a function of the number of students taking a course – also indirectly in that, on courses with many students, there will be greater challenges in clarifying and communicating the relationship between learning outcomes and teaching- and assessment methods for the student (Raaheim et al., 2019; (Iglesias, Entrialgo, & Müller, 2020). We may think that, regardless of the quality level of admitted students, larger courses make it more difficult to follow up each individual student and thus compensate for weak prerequisites; adapt the teaching to the individual; and follow up administratively where necessary. Having many students on a course makes it more demanding to facilitate student active learning and relational follow-up than on courses with fewer students. In the study quality model, we find this relationship made particularly clear in the center of the figure, ref. the teaching quality, and in the meeting between teacher, student, and course content. The meeting itself as a metaphor for the link between teacher, student, and course material is particularly important (Bolnow, 1969; Nespor, 1994). This meeting will have a different form, depending on the number of students taking the course. If we assume that the actual meeting between teacher, student, and the teaching material will have a different framing and thus provide different prerequisites for learning in courses with many or few students, we see an explanation for differences in examination results. But such relationships are complex. High quality level of admittance can help to compensate for such effects, while weaker quality can make high program and teaching quality important for students' learning. We may then assume that poor quality level of admittance will lead to poorer grades, while high teaching quality can compensate for poor quality. At the same time, we may assume that it will be easier to maintain high teaching quality and high quality on courses with fewer students. More info is required. Further, it is reasonable to assume that high program quality will to some extent compensate for less satisfactory teaching quality, especially for students who fall into the category of high-quality level of admittance. This could be independent students who, within the framework of a clearly structured program (high program quality), where the link between the didactic categories is clear, will be able to get good grades even if subject teachers' follow-up of students and other study social conditions are weak. With reference to the framework quality, we may also think that good social and non-academic support, both that provided by the study support system and that which the students themselves experience in relation to their own study environment, could be of importance for the examination results. It is difficult to see clear connections between such support and the number of students taking a course, but they could exist.
Another condition under what is referred to as frame factor quality in the study quality model is the teaching competence of the teachers. This aspect is found in the bottom blue frame in Fig. 1. In much of the literature regarding, e.g., the problem of dropping out of higher education, it seems that this aspect, for instance, the question of teaching methodology, has a bit of "touchy" about examining it critically (Iglesias, Entrialgo, & Müller, 2020). Such a perspective would then mean turning attention from external factors towards one's own and colleagues' practice. Such practice is vulnerable to criticism and questioning, since it is so closely related to both the professional and personal identity of being a university teacher (Mortensson, & Roxås, 2016; Barbarà-i-Molinero et al. 2017). Being a competent university employee is closely linked to being successful as a teacher, although many university employees experience a conflict between the role of teacher and the role of researcher (Linna Xu, 2019). However, there is reason to assume that the educational competence of staff will also be able to contribute to a high quality of study in courses with many students and will thus be able to help compensate for both weaker quality level of admitted students and weak program quality. Skilled and experienced university teachers can, e.g., have taken over weakly justified courses, in which they themselves have not had the opportunity to carry out the course- or program planning, but can still succeed well in working with the students' learning. We must also ask whether experienced university teachers, to a greater extent than their inexperienced colleagues, are in a position to choose which courses they want to teach. They may then have a tendency to skip courses with many students, at the same time as there being a competence factor. Courses with many students are usually introductory courses on a basic level, while courses with fewer students are usually at master's level. That said, this will vary between faculty’s. There will be a natural distribution of courses between highly formally qualified university teachers, where experienced teachers receive or are assigned courses at master's level with few students, while new and less experienced university teachers have to take responsibility for introductory courses with many students. Such conditions can be one of the explanations why courses with many students seem to have weaker grades than courses with fewer students. In addition to what is mentioned above, it is also important to highlight the “relationship effects”, which have been given less attention in the literature. These effects contribute to complicating the matters we have raised above. We think that a close link between the student and the university teacher, which in the model falls under teaching quality, in itself can contribute to a feeling of responsibility for the subject (ownership relationship) among the teachers (Antunes Scartezini, & Monereo, 2018). In this way, one “blurs” the connection between candidate and performance. It is reasonable to assume that such an effect will be strongest for courses with few students, since this connection will in most cases be closer. Courses with few students create a form of shared ownership of the work, which can create unclear boundaries in the relationship between the student and university teacher and between the university teacher and the course, on the one hand, and between the assessment of the student's performance, on the other.
We implicitly think that large courses provide weaker prerequisites for learning, which in turn affects the examination results. Courses with many students may contribute to less student contact, more alienation, weaker prerequisites for identifying and following up students who are performing poorly. An important aspect, especially in connection with the points highlighted above, is whether the university teacher is an internal examiner (if there is a two-examiner system) or an examiner. If so, one explanation for the fact that censorship can be stricter on large courses could be that there is a greater distance between the student and the teacher. On small courses, where the contact or meeting between teacher and student is close, a relational link is possibly created, as well as a direct and more implicit responsibility of the teacher. There is a closeness between student and teacher in courses with few students which, on the condition that the subject teacher is also a (co)examiner, can contribute to better grades due to the closeness and not just as a result of the qualities of the tasks or the teaching program itself.
Despite the results in our study showing that courses with fewer students give better grades than courses with more students, it is still important to highlight that, even if we run our analysis based on courses at the same level, there is still a challenge in that large courses with many students are often compulsory, to a greater extent than courses with few students. The students taking the large courses are then not necessarily as motivated as those taking the more specialized smaller courses which students have chosen and are particularly interested in and/or good at. This may influence students’ motivation, which in turn affects the examination results.
It is also common to think that large courses contribute to less teaching resources per student, leading to poorer examination results. However, such a series of arguments is too hasty. It is not necessarily the case that the teacher/student ratio is worse in courses with more students. In some education programs, one can have a situation where large courses are divided into many small classes. We may also have situations in the larger courses where many student assistants are used to provide individual help and feedback. This can then contribute to more help and feedback in the largest courses, compared with more moderately sized courses where the university teachers are often working alone.
In addition, large courses, combined with few lecturers, place some limitations on what can be done with student active learning methods: methods that have been documented to contribute to good examination results. The question is, however, whether student active learning methods are used more often in courses with fewer students, despite the opportunity for doing so being at its greatest in such courses.
Based on the results from our analysis, it is particularly interesting to see that the difference in examination results for the different class sizes decreases for higher course levels. Such results are not surprising. Students' experience of being students could be one reason. More time spent in the routine of being a student could be an important explanatory variable. It is reasonable that students with several years of experience better handle being a student than students without such an experience.