Class size
This article is about a parameter associated with the learning context {{{1}}}Property "Defining ingredient" (as page type) with input value "{{{1}}}" contains invalid characters or is incomplete and therefore can cause unexpected results during a query or annotation process.. The value of the parameter influences how well the learning context works and what techniques are most suited for the learning context.
View list of learning context parameters
Definition
Class size refers to the number of learners present (synchronously) in a class. It generally excludes the explainer and other observers who may be present, as well as learners who consume the class content asynchronously. The reason for focusing on learners is that the learners are the ones whose learning experience is the typical goal of the classroom, and increasing the number of learners generally makes this more challenging (though it also offers some advantages).
Related metrics
- Student–teacher ratio measures the ratio of the total number of students to the total number of teachers at an educational institution.
Heuristics
- Maimonides' rule states that a class size must not exceed 40; any class with more than 40 students should be split into two. This is not to be interpreted as saying that a size of 40 is optimal in and of itself, but rather, that it is the maximum above which it is justified to have an overhead of an additional class.
Mechanisms of effect of class size and other confounding variables
Summary
Consideration | Are larger class sizes better or worse? |
---|---|
Ability of the explainer to choose a pace and method of explanation that works for the learning experiences of all learners | worse |
Ability of the explainer to spend individualized time with learners in the classroom | worse |
Cold calling | better/ambiguous for increases from low or medium, but worse beyond a point |
Polling | better in the sense of diversity of information collected, but possibly more challenging to conduct (can be overcome with technology) |
Ability of the explainer to choose a pace and method of explanation that works for the learning experiences of all learners
In theory, the larger the class size (i.e., the larger the number of learners), the harder it is for the explainer to choose a pace and method of explanation that is satisfactory to all learners.
The extreme case of a single learner is illustrative: in this case, the explainer may be able to entirely skip parts the learner is already familiar with, and go into depth into parts the learner is struggling with, even if those are parts that are not typically dwelled upon. Even with two learners, trade-offs need to be made: some parts may be easy for one learner and hard for others, and some parts may be interesting to one learner but boring to others. Different learners may learn better from different types of examples.
In practice, outside of the extreme cases of really small sizes (three or less), class size is not the main factor determining how easy it is to pace and create instruction for all learners. Rather, the bigger factors are the explainer's own understanding of learners, the gap in prerequisites and other background between learners. Large and relatively homogeneous classes may pose less of a challenge to explainers than small but diverse classes.
Ability of the explainer to spend individualized time with learners in the classroom
Explainers can spend a little bit of time with learners one-on-one through methods such as in-class desk work: the explainer assigns some work to all learners, then checks on the work of some of them (and may choose different learners for different pieces of work). The amount of time the explainer can spend with each individual learner is inversely proportional to the class size (i.e., number of learners).
The operation of cold calling
Further information: Cold calling
Cold calling is a practice where the explainer randomly selects a learner and asks the learner a question. Cold calling can assist with continuous assessment by both the explainer and the learners, and help prevent the illusion of transparency.
The dynamics of cold calling change dramatically with a change in the class size, which is the number of learners between whom the explainer can select. When the class sizee is very low, then the cold calling has less of an element of surprise, and each learner gets called fairly often. This allows for more continuous assessment of each learner, but can also be more fatiguing to individual learners. As the class size increases, each individual learner is called upon less to support the same overall cold calling frequency. Up to a certain point, this increase might be worthwhile because it reduces fatigue for individual learners, and also gives both the explainer and learners more diverse, varied feedback on how learners are processing information.
However, beyond a point, an individual learner's likelihood of being cold called drops low enough that cold calling is no longer a way of systematically keeping the individual learner engaged.
A related challenge the explainer faces is being able to juggle all the learners' names and faces, and being able to effectively remember how often a given learner has been cold called to make sure everybody is being cold called equally.
Maimonides' rule (which sets 40 as an upper limit on classroom size) is a reasonable rule-of-thumb: for a class duration of about 50 minutes, 40 is approximately the largest size at which we can be reasonably sure that everybody will be cold called at least once per class, for a reasonably high cold calling frequency of once per minute.
The operation of polling
Further information: Polling
Polling for yes/no or multiple-choice questions generally benefits from larger classroom sizes, because we can collect more interesting data on the performance of each option, and drill down more deeply. Beyond a certain classroom size, efficient use of polling may require the use of technology like clickers (since a show of hands becomes too difficult to aggregate). In fact, lecture halls of hundreds of students can support and benefit from polling.