Results showed a general increase in number of conceptual items (and ...... only, as shown in research by the National Cumculum Council, 1991 in England).
British Journal of Educational Psychology (1993). 63,400-413 0 1993 The British Psychological Society
Printed in Great Britain
Pupils’ perceptions of a good teacher: a developmental perspective from Trinidad and Tobago Peter Kutnick* ICAPE, Universiv of Sussex, Brighton
Vena Jules Facully of Education, Universily of the West Indies, St Augustine, Trinidad
The study reports on a large-scale survey of pupils’ perceptions of a good teacher in the Caribbean republic of Trinidad and Tobago. An essay-based, interpretive mode of research was used to elicit and identify constructs used by pupils between ages 7 and 17. The sample was a proportional, stratified, clustered, yet randomly selected representation of primary and secondary schools across the two islands. 1633 essays were content analysed and coded for age developmental comparison. Factor analysis showed no consistent underlying groupings of the 166 conceptual items by age, thus analysis was undertaken within logically constructed sets of items that are described: physical and personal characteristics of the teacher, quality of the relationship between teacher and pupil, control of behaviour by teacher, descriptions of the teaching process, and expected educational and other outcomes obtained by pupils due to teacher efforts. Results showed a general increase in number of conceptual items (and words per essay) by age, but this confuses four further criteria of items used. The four criteria were: age constancy, characteristics of younger pupils, characteristics of 1113year-olds, and characteristics of the oldest pupils. All ages perceived good teachers by physical presentation (clothing and appearance), teachers’ care for pupils, descriptive teaching actions, and trustworthiness. Younger pupils focused on appearance, subjects taught, and assertion of physical punishment. Mid-aged pupils focused on the range of classroom control used by teachers (including distributive and retributive punishments), actions involved in the teaching process, and a growing awareness of the individual needs of pupils. Oldest pupils understood that good teachers must be well trained and highly motivated, should be sensitive and responsive to the needs of pupils, draw the pupils into the learning process, and have a major responsibility in preparing the pupil for the world of work and further education. Important aspects of the study show the link betweenpupils’ conceptual reality and classroom environment, that pupils are very concerned about the relationship between themselves and teachers, and that the curriculum dominated approach to teaching does not meet pupil expectations.
*Requests for reprints should be addressed to Peter Kutnick, Education Development Building, University of Sussex, Falmer, Brighton BNl 9RG.
Perceptions of a good teacher
40 1
Knowledge of pupils’ perceptions of a good teacher is important from at least two standpoints. Firstly, this knowledge helps teachers in their practice. Such knowledge gives teachers a clearer picture of pupil expectations and can, if seen as relevant, guide teachers’ behaviours in the classroom. Secondly, and more important, perceptions are the individual’s private interpretation of interaction. Knowledge of pupil perceptions therefore allows others into the world of pupils’ realities - realities which are frequently the source of their beliefs, and in turn, their behaviours. Any individual’s perceptions may be unique and subjective. On the other hand, there is also a likelihood of commonality of perceptions among individuals within school situations; what Woods (1976) calls the ‘group perspective’ which is ‘modes of thought and action developed by a group which faces the same problematic situation’. An understanding of pupil perceptions of the good teacher must therefore explore individual pupil perceptions for their commonality in order to identify constructs by which pupils understand and define the good teacher. Differences in this definition, however, will exist based on pupils’ age and experience in school, gender, and ability. In this article on pupils’ perceptions of a good teacher the focus will be on differences that exist based on age and length of school experience. A main requirement of studies concerning pupil perceptions of a good teacher is to elicit constructs which are both valid and relevant. Depending on the method of obtaining and analysing data, Cohen & Manion (1981) point out that such studies can be either normative or interpretive. Normative studies identify preset criteria, as defined by others, for pupils to affectively rate. Interpretive studies allow pupils to express knowledge of a good teacher based upon their own classroom experience. The methodology involved in normative studies therefore delimits the investigation to preset criteria and may not account for all constructs of a good teacher that would be generated in the pool of pupils’ classroom experience. The methodology of interpretive studies, however, draws upon pupils’ direct perceptions of classroom events which provide both relevant and valid constructs at the levels of the individual and the group. These perceptions of pupils are valid because they are specifically unearthed in the research and are not derived from what others outside the immediate pupil experience believe. Perceptions are relevant because they pertain to the processes which define teacher-pupil interactions at first hand. An interpretive approach also allows for an evaluative framework through the frequency with which pupils mention constructs of importance to them; it provides for the validity and relevance necessary for research attempting to probe the depth of pupil perceptions because it is the one method which allows for the widest possible range of pupils’ freely expressed views and insights. Of the research studies exploring pupil perceptions (Allen, 1959; Blishen, 1969; Brecklemans, Wubbels & Creton, 1990; Dale, 1976a; Darmon & Rich, 1988; Entwistle, Kozeki & Tait, 1989; Hollis, 1935; Lomax, 1978; Meighan, 1977; Nash, 1976; Payne, 1987; Sandford, 1984; Taylor, 1962) the majority follow a normative mode. Among the findings from normative studies two consistent aspects of the good teacher emerge. The first aspect is work related, and has to do with the teacher’s competence to teach. Good teachers must explain well, have strong control over the lesson content and behaviour of the class, and be well organised. The second aspect is relational between teacher and pupils. The good teacher is helpful, understanding, and patient. Some variations have been found within these aspects as they relate to age and gender, with older children and females
402
Peter Kutnick and Vena Jules
(generally) rating relational aspects most highly. The few studies that follow the interpretive mode use classroom observations (Sandford, 1984), interviews (Nash, 1976; Woods, 1976), and descriptive essays (Blishen, 1969; Payne, 1987) to obtain data. Only two of these interpretive studies explored specifically for the good teacher. These interpretive studies show that good teachers are acknowledged for their relational qualities. Relational qualities include understanding, patience, humility, encouragement (of all pupils) to participate, praise, and controlling behaviour democratically. Many of these interpretive studies have built-in limitations with respect to their claims for the in-depth nature of pupils questioned. Limitations include: whose perception of a good teacher is being recorded - the observer or the learner, and interview strategies allow for in-depth probes which reveal areas of pupil insight (Woods, 1976) while being delimited by the interviewer’s preset core questions. There is a need, therefore, for a valid and relevant strategy that will allow pupils to articulate their views freely without the above limitations. We conclude that only a free response such as an essay-based strategy is able to elicit the widest possible range of pupil concepts because it allows the child to reflect on experience before responding and because it does not inhibit various articulations within the face-to-face interaction between child and adult interviewer or teacher. An essay writing strategy assumes, however, that the child can effectively communicate hisher thoughts, feelings and ideas using this medium. One study using the essay approach has been undertaken in the Caribbean region by Payne (1987) who explored the popular and unpopular teacher. Payne’s research shows that pupils can communicate well through essays within the Caribbean region and provides a useful backdrop, popularity or unpopularity of teachers. Yet Payne’s subject matter does not get at the heart of pupils’ understanding of a good teacher. The concern of the study reported here is to identify pupil generated concepts of the characteristics of a good teacher from the perspective of the various age and length of school experience of primary and secondary children. The present study will avoid the use of normative methods and will undertake an interpretive study amongst pupils in a developing country. The present study uses age or length of school experience as the main variables to provide developmental insights into pupil perceptions of a good teacher. Data analysis will take into account the following related findings: (a) Cognitive developmental arguments question whether development is simply the addition of more items about a particular subject or whether advancement and conceptual development are dependent on actual experience (Kohlberg, 1969). (b) Livesley & Bromley (1973) found that development of person perception moves from a peripheral to a central view, in other words younger children will note physical characteristics of a person while older children will note personal and interpersonalattributes of an individual. (c) In social and relational development, Youniss (1978) points out, the child moves from a realisation of a dependent (through hierarchically reciprocated) relationship to a realisation of interdependence and mutual reciprocation in relationships. Thus any age developmental study of children’s perceptions of their teacher must take into consideration that cognitive, social and relational criteria will affect the perception of a pupil.
Method and sample The intention of this study is to explore pupil perception of a good teacher with respect to pupil age and class level (length of experience) in school in a developing country. To obtain
Perceptions of a good teacher
403
as much pupil insight concerning a good teacher as possible it was decided not to prescribe preset criteria for pupils to select amongst, rate, or analyse. Children would be asked to write a general essay on ‘a good teacher’. The country selected for study was Trinidad and Tobago, a twin island republic in the southern Caribbean. Schooling there is compulsory from age 5 to age 15 (with additional schooling available for those working to complete A-level examination courses); 82 per cent of the 11- to 12 year-old age cohort who have completed primary schooling have access to secondary education. A cross-sectional design was used to obtain a representative sample of pupils within the age range 7 to 17. Educational districts were chosen to represent a geographic transect of the country, covering three of the seven educational districts of Trinidad and the educational district of Tobago. The sample was selected to be proportional (a ratio of the total enrolled school population to the chosen school district), stratified (covering class levels in both primary and secondary schools, types of primary and secondary schools, private and state schools, and the state schools under various religious and secular managing authorities), random (schools and classes within schools were randomly selected once proportional and stratified criteria were identified) and clustered (whole classes were used to generate information rather than selected individuals). In all, the expected sample was to be 1977 pupils, but due to absences essays were written by 1633 students. The sample represented 5 per cent of each of the following class levels; standard 2 (7 to 8 years of age) and 5 (10 to 11 years) in primary schools and forms 2 (12 to 13 years) and 5 (16 to 17 years) in secondary schools. Upon identifying appropriate schools and classes, arrangements were made so that the class teacher who would normally request an essay provided all pupils with sheets of lined paper and explained to their pupils that they were to write an essay on ‘what is a good teacher’. To avoid focusing on a particular teacher or teacher-gender, pupils were asked not to identify or name any particular teacher, but to write about good teachers in general. Pupils were also asked to include their name and other biographic information with their essays. Coding of essays was undertaken only after a full content analysis of a representative 10 per cent of all essays was completed. The representative 10 per cent was selected as a random choice of 10 per cent of pupils in each class of the full sample (which was already selected to be proportional and stratified). The content analysis identified each conceptual item of each essay. Content coders then analysed conceptual items for replication and overlap, and withdrew any items that were clear replications of a previous item. In total, 166 separate conceptual items were identified. Essays selected for the content analysis were then returned to the complete set of essays and an analysis by content item, total number of concepts and total number of words was undertaken for each essay in the full sample of 1633 essays. By using a representative sample to obtain conceptual items, the study draws upon its interpretive method by having established grounded items for analysis. In the event that other conceptual items might appear in the complete content analysis, coding was undertaken to allow additional conceptual items to be added to the original items. In the event, no new items were found in the full content analysis. Further, to ascertain that essays provided all of the conceptual information for the selected pupils a small-scale reliability check was simultaneously undertaken in which five children from each class of 30 were interviewed by the authors instead of writing the essay. At each class level, there were no significant differences in number of concepts identified between interviewed and essayed pupils, except for the 5th formers where the interviewed pupils identified an average of 13
404
Peter Kutnick and Vena Jules
concepts while the essayed pupils identified an average of 10 concepts. Thus, reliability of the essay-based outcome was deemed satisfactory for further analysis to proceed. Both the biographic data and the concepts identified in the essays were analysed in the search to find what relationships existed between pupils’ age and their perception of a good teacher.
Results Having a large number of conceptual items and a large sample to draw upon, qualitative and quantitative analyses were possible. A logical ‘first-step’ in a developmental analysis of perceptions of a good teacher was to factor analyse all of the 166 conceptual items by class level, searching for significant combinations of items that would vary (develop or regress). Scores were obtained from each individual pupil on each conceptual item, and from this analysis class level (or year in school) was chosen as the basic unit of analysis as it represented age development; pupils generally progressed as a whole year grouping from one class level to the next. Results indicated that a few of the concept items were mentioned by large numbers of pupils and the corresponding factor analysis did not show significant item groupings. Initial gross quantitative analysis found: total number of words used ranged from 0 to 444 and number of words was significantly correlated to class level; total number of concept items ranged from 0 to 27 and number of items was significantly correlated to class level (see Table 1). An initial class level analysis was undertaken to ascertain whether the number of words and concepts identified were related to the gender of teacher andor gender of pupil (in effect a check on bias due to gender). Table 2 shows that only a proportion of the essays identified the teacher as male or female and this proportion decreased by class level. Only in Standard 2 were number of words and concepts related to gender, such that girls used more words and concepts when writing about female teachers and boys used more words and concepts when writing about male teachers. These significant interactions did not occur in any of the other class levels. Form 5 females generally used more words and concepts than male pupils, although these results did not relate to gender of the teacher.
Table 1. Average number of words and concepts per class level (standard deviations in brackets) Class Level Standard 2 Standard 5 Form 2 Form 5
Number 425 478 430 303
Correlationto class level (Pearson corr.)
Average words
Average concepts
per essay
per essay
62.66 (43.08) 125.08 (60.52) 133.50 (73.76) 166.04 (108.89)
6.2 (3.47) 8.40 (3.41) 9.06 (4.23) 9.68 (5.99)
0.621
0.468
405
Perceptions of a good teacher
Table 2. Gender identification of teachers as represented in total words and total concepts by pupils (significant differences and interactions) Total words Difference due to: Gender of teacher
Total concepts
Gender Interaction Gender of pupil of teacher
Standard2
NS
NS
Standard5 Form 2 Form 5
NS NS NS
NS NS F1,77=8.475
F1,208=8.055 p
