This classic research is reprinted from set No.1, 1983 because it has lost none of its relevance today. In suburbs where classes were increasing and decreasing in size, the effects of crowding on attending, withdrawing, and aggressive behaviour were discovered.
Three Dozen’s a Crowd
By Jack Campbell and Margaret Robinson,
University of Queensland.
One of the most striking characteristics of classrooms since the Industrial Revolution to the present has been the large number of pupils within them. At no other stage in their lives will young people occupy such socially intimate settings for such extended periods of time. Buses, football pavilions, and cinemas may be more crowded than classrooms, but people rarely stay in them for long, and, perhaps more important, they do not have to meet the demands of individual and social learning within them.
Largely as a result of the studies by Glass and his associates, it is now fairly well accepted that large class size has a harmful effect upon the learning of school children, and it might be argued that the battle for smaller classes has been won. However, before this finding is translated into policies aimed at reducing class size, it is probably neccesary to identify the sequence of events which lies between the physical size of a school class and the achievement of pupils within it.
Speculative Explanations
Most of the speculative explanations relating to the effects of class size involve the teacher - for example, in a smaller class, the teacher is better able to: maintain close supervision over all of the pupils; provide individual instruction; avoid having lines of pupils waiting for attention or resources, and so on. These explanations invite the rejoinder that surely competent teachers should be able to develop strategies to maintain effective management, even though the class is large, and that incompetent ones would be ineffective whether they had four pupils or forty. Without buying into these arguments, it can be noted that large class size can generate a different sequence of events which remains throughout within the mental and emotional states of pupils, and is almost beyond the teacher’s power to control. The sequence goes something like this: when the personal spaces which we establish around our physical selves are threatened or invaded, we are likely to become aggressive, or to withdraw; these responses will result in less energy being available to attend to what we planned to do; diminished attention, in its turn, will adversely affect achievement, especially when the task is a complex one. According to this line of reasoning, effects are likely to flow almost irrespective of teaching strategies.
The Study
During the last eighteen months we have been trying to test part of the ‘crowding’ hypothesis by monitoring the attending, withdrawn, and aggressive behaviour of primary-school pupils, first, when in a class of around 36, and, second, when in a class of around26. A total of 14 classes have been involved - 10 which experienced a reduction in size of the kind just mentioned, 2 which had enrolments of 36 and were not reduced (these were used as ‘controls’), and 2 which experienced an increase in size from 22 to 32 (these were used to check on the effects of change per se).
The Results
This study confirms the judgement of generations of teachers that large school classes lead to unproductive behaviour on the part of pupils:
1. When classes were reduced from 36 pupils to 26, the percentage of time spent on attending increased from 73.62 to 84.76 - during one school year of approximately 1,000 hours, this amounts to a difference of 111 hours, or 22 school days.
2. When classes were reduced from 36 pupils to 26, the percentage of time spent on withdrawn and aggressive behaviour decreased from 15.40 to 5.35 - i.e., approximately 100 hours or 20 school days in a school year.
3. Younger pupils were affected more by crowding than were older ones, but there were no sex-related differences.
4. ‘Highly effective’, ‘moderately effective’, and ‘inefective’ teachers were all able to secure an improvement in the attending of their pupils when their classes were reduced in size.
Discussion
During the last six months we have had the opportunity to discuss the study and its findings with teachers and school administrators throughout Australia, and we thought that readers might find it interesting to listen in, as it were, to the reactions of perceptive critics, and to our responses.
1. Q. How do you know that the effects which you claim to have identified are associated with class size?
A. We were engaged not in a simple survey, but in a naturally-occurring experiment, and this is accepted as a reputable design. The variations in class size were part of the on-going life of schools and were not contrived by us; we acted out a non-interfering role, merely recording the continuous stream of classroom behaviour of the pupils by means of unobtrusive, but sophisticated, event recorders and a microprocessor; there was a substantial measure of control over the two sets of observations (pre- and post-change) in as far as the children were compared with themselves and were being taught the same kind of lesson (usually language-arts or arithmetic), by the same teacher, in the same room, and at the same time of day. In addition, pupils in classes which did not experience a reduction in size revealed no changes in behaviour, and those in classes which went from small to large revealed a decrease in attending behaviour and a corresponding increase in withdrawn and aggressive behaviour. These experimental data look convincing to us, and certainly they are very stable - the same results emerged in class after class.
Finally, the findings are consistent with theory, and, perhaps more important, with the intuitive knowledge of almost all classroom teachers. Teachers process a lot of information in the course of their careers, and we feel reassured when our research findings are in line with their judgements.
2. Q. You seem to have a hang-up associated with attending behaviour on the part of pupils.
A. Attending, or ‘on-task’ behaviour arises as part of the theoretical argument concerned with the effects of crowding, and this is the reason why it features prominently. In addition, however, -and this is a bonus - several recent studies show that, if a person is compared with himself or herself, the attention which is given to mastering a task is the best single index of how much will be learnt. Thus, if we spend six hours per week trying to master the piano, and only one hour trying to master some other equally difficult task, it could be expected that we would master the piano before the other task. Attending behaviour is a very significant variable in learning.
3. Q. But by focusing on attending behaviour you project a notion of school learning which is very narrow and academic; what about personal and social development?
A. It is true that attending behaviour has an academic, if not industrial, ring about it, but the theory also features other behaviours which are at the very heart of personal and social adjustment - withdrawal and aggression. The effects of class size include affective as well as cognitive outcomes, and it is very likely that the former are the more responsive to crowding; certainly, as the theory proclaims, they have priority.
4. Q. O.K., but what is the optimum time for a pupil to
spend attending a learning task?
A. We have no satisfactory answer to this, and other research literature doesn’t help, either. We accept the implication that the optimum is unlikely to be 100 per cent - there is a need for some tuning out. We could be wrong, but it is our judgement that 85 per cent of time spent attending to the learning task is better than 74 per cent, especially when the difference is spent on withdrawn and aggressive behaviour. Unless a pupil is actually attending to the task set by the teacher, the classroom is not a very stimulating place; in this respect, it differs from the play ground, museum, and the city streets, where children might go ‘off task’ and still learn a lot.
5. Q. If your findings are sound, why not simply increase the size of classrooms - say from 6m x 7m to 7m x 8m and so avoid overcrowding?
A. This sounds plausible - and cheaper! One of our two classes which increased in size from 22 to 32 was able to maintain the space per pupil by moving back flexible walls, whereas the other one wasn’t able to do this, and so we had an opportunity to untangle the two confounding variables - number of pupils, and space. The findings suggest that increasing the size of classrooms helps to overcome the crowding a little, but it is clear that sheer numbers, almost irrespective of space, can generate a sense of crowding. In addition, especially when the pupils are young, the number can be too large for one teacher to manage effectively - too much time will be spent waiting for attention, for resources, for others to finish and so on. There is some, but not much, mileage in making classrooms larger.
6. Q. What is the optimum size of a school class?
A. Our study doesn’t provide the answer to this; we have undertaken research within the special circumstance of classes being reduced by about 10 pupils when their enrolments approximate 37, and we have no evidence beyond this circumstance. The issue is one of what is a comfortable personal space for learning. Our data suggest that when 36 pupils are in a area of 42m2 they are too close for comfort, and that 26 pupils in the same area are more comfortable, but we would not want to argue that 1 pupil, or even 5 pupils, would feel optimally comfortable in those 42m2. Just as they feel threatened when their personal spaces are invaded, so many young people seek the proximity of others and would become anxious if there was too great a separation. It would not surprise us if the optimum size of a class within an area of 42m2 was between 15 and 20 pupils, but this is little more than speculation.
7. Q. Your study has involved only traditionally-designed, one teacher classrooms; what about multiple-area classrooms, team-teaching situations, composite classes and the like?
A. A good question! In this pioneer study, we have taken the simplest set of conditions, and we make no arrogant claim that there are no more worlds to conquer.
8. Q. Your conceptualization of the sequence of events is highly pupil-centred; as you say, it stays within the psychology of the pupil, and the teacher and her control mechanisms are blotted out.
A. This is true, and, as behaviour-setting theorists, we take the implied criticism to heart. The conceptualization presented in the study seems to be strongly supported by the data, but, as in most research investigations, there remains the possiblity that other conceptual frameworks would be equally, or even more effective. We have extensive data on teaching strategies employed by the teachers under the two class-size conditions, and we plan to syncronize these with the behavioural data from the pupils to test the efficiency of alternative viewpoints. The Gordian knot is real and can be unravelled only a few threads at a time. At the risk of mixing metaphors, and with apologies to Lewis Carroll: the further off from England, the nearer is to France - then turn not pale upon the trail, but come and join the dance!
NOTE
The study referred to is presented in detail in a report to the Education Research and Development Committee, Canberra, 1982, under the title ‘ A Social-Density Explanation of Class-Size Effects’.