5.1 Effectiveness
5.2 Cost-effectiveness
Heyneman & Loxley (1983) studied science achievement in 16 developing and 13 industrialized countries, and found that
Children who attend primary school in countries with low per capita incomes have learned substantially less after similar amounts of time in school than have pupils in high income countries. At the same time, the lower the income of the country, the weaker the influence of pupils' social status on achievement. Conversely, in low-income countries, the effect of school and teacher quality on academic achievement in primary school is comparatively greater. From these data, which are more representative of the world's population of schoolchildren than those used in previous studies, it is possible to conclude that the predominant influence on student learning is the quality of the schools and teachers to which children are exposed. (p 1162)
Fuller (1987) considered more than 50 empirical studies. His review also suggests that
the school institution exerts a greater influence on achievement within developing countries compared to industrialized nations, after accounting for the effect of pupil background. (pp 255-256)
He reports that few studies have emphasized effect sizes or the efficiency with which a particular school factor boosts achievement. Almost always outcomes are measured by tests, often in reading, maths or science, but sometimes by a comprehensive examination. A table summarising Fuller's (1987) findings is reproduced in Lewin with Ross (1992, p 1 87). Some of Fuller's specific conclusions are worth quoting here. Based on the percentage of studies showing positive effects, he found that effective parameters influencing school achievement are length of the instructional programme, pupil feeding programmes, school library activity, years of teacher training, textbooks and instructional materials. Ineffective parameters are pupil grade repetition, reduced class size, teachers' salaries, and science laboratories. For example, "in most situations, lowering class size with the intent of raising achievement is not an efficient strategy".
In the area of materials, Fuller concludes that
A good deal of evidence now suggests that material factors in schools - such as more textbooks or writing materials - exercise more influence on achievement in the Third World than in industrialized countries.(Fuller, 1987, p287)
He finds that the influence of textbooks appears to be stronger within rural schools and among students from lower income families, but that very little research has been conducted on how. and the conditions under which, textbooks shape achievement.
These findings are consistent with those of the earlier review by Schiefelbein & Simmons (1981). They argued that there is a small number of main determinants of school achievement, but also that
it is important to experiment with the suggested changes before policies are endorsed and implemented on a national scale.
Their results relating to teacher characteristics have been stated in section 4. Under the heading of "school resources and processes", Schiefelbein & Simmons (1981, pp 10-12) found that
i) Larger class size was associated with higher performance, or did not affect it, in 9 out of 14 studies.ii) Higher expenditure per student was not associated with higher student achievement in 5 out of 8 studies.
iii) Availability of textbooks was associated with student achievement in 7 out of 10 studies.
iv) The setting of homework was related to higher student achievement in 6 out of 8 studies.
Under the heading "student traits", they found that
v) SES was significant in 10 out of 13 studies.vi) Malnutrition, body weight and health were significant in 8 out of 11 studies (but note that this is highly correlated with SES).
vii) The more repeating the lower the score, in 7 out of 8 studies.
viii) Kindergarten attendance was related to achievement 6 or 12 years later, in 3 out of 4 studies.
An example of research from a single country which is worth strong consideration because of its breadth of methodology and content is that of Vulliamy (1987, p 217), who found persuasive quantitative evidence of the existence of school effects on secondary school examination results in Papua New Guinea, and used qualitative case studies in an attempt to identify significant factors. According to Vulliamy, these factors are
1) quality of teaching
2) style of school administration
3) extra assistance for weak students
4) levels of staff morale
5) the provision of basic facilities (such as water and electricity)
The importance of the headteacher is paramount.
It is now widely accepted that schools do have important effects. The research suggests that such effects are related not to resource-based school input factors, but rather to school-process factors that are more elusively categorised as features of school climate or school culture. (Vulliamy, 1987, p217)
Another single country study was conducted in Zimbabwe by Riddell & Nyagura (1991). Their work was based on a secondary school survey and multi-level analysis. They found that student achievement is higher when schools have a greater availability of textbooks, a larger proportion of trained teachers and teachers who have taught at that school for a longer period of time.
This suggests that raising the proportion of trained teachers and, more importantly, improving the provision of textbooks and providing incentives for teachers to remain in the same schools for a reasonable period of time are promising investment options to boost student achievement. (ibid, p51)
In a more wide-ranging study including case studies of effective schools in eight countries, Levin & Lockheed (1991) argue that flexibility appears to be key to effectiveness, and point out the importance of material inputs on achievement in economically impoverished countries.
Resources sufficient to provide even the most rudimentary conditions for success often are lacking.
They argue that creating effective schools in developing countries requires three elements: basic inputs, facilitating conditions and the will to change. The necessary inputs are
* a well-developed curriculum, in terms of both scope and sequence;
* sufficient instructional materials for students;
* adequate time for teaching and learning
* teaching practices that encourage active student learning.
The facilitating conditions are
* community involvement;* school-based professionalism (which includes the crucial role of the principal in school effectiveness, teacher collegiality and commitment, and autonomy balanced with accountability);
* flexibility in curriculum and organisation.
The will to act includes vision and decentralization.
Haddad et al (1990) give an extensive summary of empirical research findings; those which were published in the last decade are reproduced (in abbreviated form) in Lewin with Ross (1992, pp 188-191). A particularly interesting aspect of the summary is that concerned with process factors, especially school management.
We know that well-managed, effective schools share several characteristics: they display an orderly environment, emphasize academic achievement, set high expectations for student achievement, and are run by teachers or principals who expend an enormous amount of effort to produce effective teaching and encourage pupils to learn, no matter what their family background or gender. Few schools in developing countries display these features. (Haddad et al, 1990, p57)
Haddad et al report that
many of those who have observed the schooling process in both developed and developing countries conclude that the most important factor governing how well pupils do in school is school management.... several studies have identified headmaster education and experience as important variables that affect pupils' achievement (ibid)
The issue of school management is discussed further in section 6.
According to Haddad et al (1990, p50)
Educational production function studies have not been able to tell us accurately which school inputs have larger and smaller effects on achievement. Neither have they been particularly useful in identifying which inputs are more cost-effective than others - although "size" effects of the inputs are often a product of such studies, they rarely measure costs of inputs.
A review is given by Lockheed & Hanushek (1988). They point out that only a very few possible educational investments have been subject to analysis containing both effectiveness indicators and cost indicators, and these studies only provide examples of how decisions could be informed by such evidence. The findings of the review are reproduced in Appendix B. According to Lockheed & Hanushek (1988, pp 31-34)
Most research on instructional materials has focused on textbooks, and their cost-effectiveness has been examined in several studies. Textbooks have been found more cost-effective than (a) not having textbooks in Nicaragua and the Philippines, (b) post-secondary teacher education in Thailand and Brazil, and (c) a variety of other educational infrastructure and 'software' inputs in Brazil..... The positive effect of interactive radio education for learning mathematics and foreign or national languages has been demonstrated in Nicaragua, Kenya, and Thailand. Radio has been found more cost-effective than conventional instructional methods relying on teachers alone..... The cost-effectiveness of technical-vocational programs in comparison with general secondary education has been evaluated in Colombia and Tanzania, with somewhat inconsistent results..... The evidence regarding teacher education suggests that training that emphasises basic skills is more cost-effective (vis-a-vis student learning) than training that refines skills at a higher level.
They conclude that
On average, across several countries and a variety of student learning outcome measures, the more cost-effective interventions are textbooks, interactive radio, peer tutoring and cooperative learning. Less cost-effective are teacher training and technical-vocational schools. (ibid, p36)
Although this evidence is very interesting, and aspects of it are convincing, it should again be noted that there are methodological difficulties. The evidence on peer tutoring and cooperative learning comes from USA and Israel. It must be remembered also that there is no evidence on many other potentially cost-effective strategies, for example school management training.
