ABSTRACT
There is now a substantial body of research about school effectiveness and school improvement and in reading this you need to read critically, being aware of ways in which to judge the relevance of the research to your own situation. Anyone reading this research needs to ask questions about the text:
What is the sample of schools/teachers chosen? Does it include different types of schools? What age groups does it deal with? What generalisations can be made from it?
Many studies have been made in areas which are socially disadvantaged because these schools give more cause for concern. This needs to be remembered in generalising from the results. A large number of studies have been made in other countries and while we can learn from them differences in the culture need to be taken into account.
What criteria of effectiveness have been chosen?
A number of studies have simply used tests in the basic skills and these may be lower order skills rather than skills requiring creativity and higher level of thinking. The more useful studies have also been concerned with affective and social development and matters such as children’s self-esteem and self-confidence.
Romberg and Carpenter (1986: 862) make the following comment about the use of standardised tests in research about teacher effectiveness: ‘As long as teachers choose the contents they emphasise in their classroom, such tests can never be used as fair measures of teacher effectiveness because they evaluate students in areas that the teacher did not cover or emphasise.’
It is also important that a range of criteria are used. For example, Rutter et al. (1979) used examination results, attendance, behaviour and delinquency in their study of secondary schools. Mortimore et al. (1988) used reading, mathematics (written), mathematics (practical), writing, speaking, attendance, behaviour and a 7range of attitudes in their study of primary schools. They also used measures of the socio-economic status of the parents.
What measures of the children’s background have been used?
Many of the studies of school effectiveness have arisen from concern that schools do not start equal because children come from different backgrounds, some of which are more supportive of school work than others. Reynolds and Reid (1985: 195) sum this up as follows: ‘Outcomes cannot merely be seen as being produced by effective or ineffective schools without an assessment of which pupils, from which families and for which communities they are held to be effective or ineffective.’
What is the time element in the study?
Some studies are simply measures at one point in time. Others look at progress over a period. Effective schools need to be effective over time. A school may obtain good results one year but poor results the next. Year groups vary and it is the ability to obtain good results over time that is important.
Some researches are concerned with progress as well as achievement, for example Mortimore et al. (1988). This gives a different point of view and is a better measure of a school’s work than achievement alone.
What research techniques have been employed?
There will be some statistical information obtainable from the school, for example, attendance figures, Standard Attainment Tasks (SATs) results and teacher assessment. Studies may also include observation of classroom and management practice, examination of children’s work, testing of various kinds, questionnaires to children, teachers and parents, interviews with teachers, children and parents, action research with and by teachers, case studies and any other ideas the researchers may have. It is important that more than one method is used so that information from one source of information reinforces another. For example, studies relying on questionnaires may fail to ask questions which turn out to be important and those responding may fail to give information which might be important. Questionnaires need to be backed up with other information.
It is also important that studies are valid and reliable. Validity is the extent to which a test or observation measures what it is intended to measure. Reliability is the extent to which the same results will be obtained if the measure is repeated.
What methods of analysis have been used?
The methods of analysis should make it possible to provide information about the ways that measures are related. However, it is important not to assume that when two measures are highly correlated that one is the cause of the other. For 8example, there is usually a correlation between the number of free school meals in a school and children’s achievement, but it would be ridiculous to claim that reducing the number of free school meals would improve performance.
Most modern research uses highly sophisticated statistical techniques which allow different factors to be isolated.
Do several studies draw similar conclusions?
Evidence is more convincing if it is confirmed by more than one researcher.
Is there any contrary evidence?
Different researches will come up with slightly different conclusions and it is useful to ask whether several studies come to similar conclusions. It is also worth considering whether the findings accord with your experience and that of other teachers.
