John Oxenham

In addition to its well established national literacy campaign, the Government of Egypt has used support from the British Department for International Development to try out a different strategy for literacy and adult education. Recently completed evaluations of the effort by the General Authority for Literacy and Adult Education, the International Consultant Bureau and the Social Research Center of the American University in Cairo, plus help from the CELL team and British Council office in Cairo, have enabled John Oxenham to present some of the more useful findings here. He is particularly indebted to Prof. Dr. Ramadan Hamed and his team at the university for their careful analyses of the data they collected from their comparative survey and assessments. Dr John Oxenham has worked for many years for the World Bank. He now lives in the United Kingdom and is still working on adult education.

Reinforcements from Egypt: Observations from a Comparison of Literacy Strategies

A recent evaluation in Egypt offers adult educators a number of reinforcements for some current beliefs and recommended practices:

1. The effectiveness of the instructor or facilitator can be more crucial than either teaching method or learning materials.

2. An effective instructor enables larger proportions of learners to achieve their objectives and also narrows the gap between the highest and least achievers.

3. Well trained and supported facilitators can teach literacy from local materials without the aid of primers or texts. However, they tend to be more effective if they have supplementary materials at hand.

4. A combination of community support, a curriculum derived from local interests, ‘active’ learning methods, good training and close support appears more effective than reliance on individual initiative and a standard curriculum embodied in a single set of literacy primers, unsupported by systematic training and regular follow-up supervision.

5. Students who attend more regularly tend to be more successful. Special efforts to promote regular attendance are worth making.

6. Adult learners vary widely in their abilities to learn the skills of literacy. However, the average intelligent non-literate adult learning literacy in Arabic seems to need at least 400 hours of learning and practice to reach a level of skill that could be credibly regarded as ‘lifelong’. As even the most regular learners seem to manage to at- tend only 80 per cent of their classes, programmes should perhaps offer 500 hours of instruction.

7. Long breaks between phases of instruction seem to damage completion rates.

The evaluation compared two current approaches to adult education with literacy, both sponsored by the General Authority for Literacy and Adult Education (GALAE) of the Government of Egypt. The first approach will be termed ‘standard’, while the second is experimental and has the title, ‘Capacity Enhancement for Lifelong Literacy’ (CELL). Fuller descriptions appear below.

The Standard Strategy

In 1994, GALAE launched a 10-year literacy campaign. Although it aimed only to enable adults to learn how to read and write Arabic and to use written arithmetic, GALAE oriented its curriculum to topics and vocabulary that most adults would find useful in daily life. The course has two further aims. One is to enable interested students to gain an equivalency certificate for primary education and thus to qualify for entry to a junior secondary school to continue their education. The second aim is to help people apply literacy skills to a craft or vocation to improve their management and productivity. For this purpose vocational training in a range of crafts is available.

The curriculum includes two primers, one for the first five months of classes, the second for another session of five months. Both primers are standard and used throughout the country. There is an assessment at the conclusion of each phase, based on the actual texts of the primers. Passing the second assessment is the equivalent of completed primary school and qualifies for entry to secondary school.

The campaign relies on a combination of public information and individual initiative to recruit both literacy instructors and students. It does not rely on voluntary service. Instead, the instructors receive a monthly stipend of E£140. This is a modest sum but proportionate to the salary of a full-time primary school teacher. People who have completed secondary school or more are encouraged through radio, TV and newspaper announcements to recruit non-literate relatives and friends as students and to apply to GALAE for recognition as literacy teachers. Provided they mobilise enough learners, the teachers are authorised to start their classes, while GALAE provides them with the necessary manual, primers and other instructional materials.

However, GALAE lacks the resources to offer systematic training. To a large extent, teachers have to rely on their manuals to guide them through the curriculum and to advise them on how to encourage active learning on the part of their students. From time to time, however, the classes and teachers receive supervisory visits from GALAE Branch offices in their districts and governorates.

The strategy worked well for the first years of the campaign. Many governorates attracted their full target of enrolments and some achieved high rates of retention and passes. However, by 1999, enrolments, completions and pass rates had all begun to decline.

Further, the strategy was less effective in the poorer and more rural districts. Efforts to launch classes seemed to fail at a very high rate. This meant that the poorest people were losing out on the benefits of the campaign. Even more important, it meant that rural women, who had the highest rates of illiteracy, were once again missing out on the opportunity of an education.

Second Thoughts

Some quarters in GALAE had anticipated such problems. These educators had felt that the strategy selected had not capitalised fully on recent thinking and practice in adult education for very poor and conservative communities. In this light, even at that early stage, GALAE sought to try out alternative approaches. In 1995, with the support of the British government through its Department for International Development (DFID), an Adult Literacy Training Project (ALTP) developed an approach very similar to the increasingly popular REFLECT strategy.

The cardinal principles were that (1) communities – not just individuals – should encourage and support non-literate members to learn reading, writing and written arithmetic; (2) the learning should proceed through local, immediate conditions and materials, not through standard texts; (3) the learning and the way it was organised and supported should aim to generate community development in social, economic, environmental and political dimensions. The learning should be independent of prescriptions by external authorities, whether through pre-determined curricula leading to standard assessments for standard certification or through standard primers and other instructional materials. ALTP modified these principles to fit the environment in Egypt and opened experimental classes in two governorates, which it ran until late 1999.

The experience was promising but further testing was necessary. Also, GALAE needed support in strengthening its capacity to mainstream the new approach.

The CELL Strategy

In late 1998, GALAE and DFID began to build on ALTP with a project that would aim (a) to teach reading, writing and written arithmetic at least as efficiently and effectively as the standard programme; (b) to promote community, social and economic development and poverty reduction especially for women and especially in particularly poor and disadvantaged rural communities; and (c) in anticipation of success, to prepare GALAE officers to mainstream the new pattern of operations across the country. In the light of these aims, the new effort would have the title ‘Capacity Enhancement for Lifelong Literacy’ (CELL).

The final design embodied the following elements:

Pedagogy: Unlike the standard strategy, CELL would do without set primers or text-books. It would rely instead on training class facilitators to utilise local issues to teach reading, writing and calculating in written form. There would be a broad framework on which to build instruction, but the actual content would arise from a particular known locality. For example, the introductory sessions would consider ‘My Family and I’ and lead into ‘My Village and I’ and then on to ‘My Country and I’. To underpin the training, class facilitators would receive manuals with detailed hints on the sorts of materials and occasions that would raise the right kinds of opportunities to initiate local development. In addition, they would enjoy close and friendly support from community facilitators to help them develop their courses.

‘Basic’ and ‘Pathways’: The course would have two phases, which together would last for ten months, like the standard course. The first would be common to all classes and be known as the ‘basic’ phase. It would take about five months to complete and would be assessed in the same way as the completion of Book 1 of the standard course. The second phase would branch into three ‘pathways’ or areas of interest. Pathway 1 would aim deliberately to reach the equivalent of a complete primary schooling, as assessed by GALAE. This would mean that its curriculum would be a little more fixed than the other two, but would still leave the class facilitator with plenty of room for matters of local interest.

Pathway 2 would be vocationally oriented, not to any particular vocations or crafts, but rather towards general business management. Pathway 3 would be oriented to the situations, issues and problems of daily life in the family, community and country.

All three ‘pathways’ would be assessed in the same way as the completion of Book 2 of the standard course. On the assumption that classes would meet on five days per week for an average of two hours per day for 20 weeks for each of the two phases, the total curriculum would involve some 400 hours of learning and practice in class.

The CELL course would encourage active learning on the part of the students. Rather than simply listen to their class facilitators, they would contribute their own views, experiences and knowledge to the development of the course and undertake projects in group-work. To model such a pattern of facilitation and learning, the training courses for the class facilitators would themselves put active learning into practice.

As part of sharing responsibility with the students, class venues and timings would be negotiated with them. In effect, the course would fit the convenience of the students, not the other way round.

Community Support: The standard campaign relied on individuals to recruit students from among their families, friends and neighbours. CELL in contrast would rely on community effort to identify potential students, potential class facilitators and potential community supporters for the longer term.

In the first step, the GALAE branch offices would identify villages which were very poor, had high rates of illiteracy, especially among their women, and had proved ‘closed’ to earlier efforts by the national campaign to open literacy classes. In the second step, GALAE branch and district officers, with CELL staff, would consult with the official village leaders about the desirability of literacy classes and about the possibility of a survey to assess demand and interest. The third step would be a Participatory Rapid Appraisal (PRA) of the state of the village. It would identify informally influential people, who would be willing to act as voluntary supporters of literacy groups, as well as people who might be suitable for appointment as community and class facilitators, and people who might be potential students.

The potential supporters would be invited to form a community team to take responsibility for helping to find suitable venues for classes, ensure good rates of attendance, retention and completion and solve problems that might arise for individual students or whole classes. The added hope was that the community team would serve as the incubator of initiatives for various forms of social and community development growing out of their work with the literacy classes.

Local Recruitment: Following the PRA, GALAE branch officers and CELL staff in consultation with community leaders and the community team would recruit a community facilitator from the village itself by means of an examination in literacy, an interview and an assessment of knowledge, understanding and involvement in the community (most of the community facilitators actually recruited were university graduates – 15 of 23 – while the remaining eight had at least a full secondary education of 12 years). The community facilitator would be a GALAE employee under contract, but seconded to CELL for the purposes of the project. Her or his role would be to assist class facilitators with the social relations of their classes and to act as an intermediary with the community team, GALAE officers and CELL staff.

Then, depending on the numbers of potential students, a number of class facilitators would be recruited, again from the community. They would need to be full secondary school graduates, but would also have to take an examination and be interviewed. Such recruitment was of course possible only because the class facilitators were not volunteers offering their services free, but, like the teachers in the national literacy campaign, would be paid the usual fee of E£140 per month.

Improving the Status of Women: In many rural and agrarian commu- nities in Egypt, some men would not agree to let their wives or daughters either study or teach in literacy classes. The CELL design therefore explicitly aimed to enrol a majority of women as students and to encourage large proportions of women to act as class facilitators.

Social Development: The manuals for class facilitators contain sug- gestions that of course need to be adapted to local conditions.

Capacity Enhancement: The CELL design built in specific measures to assist GALAE to develop the capacity to adapt the new strategy for national dissemination. The measures included strong GALAE representation in the management of the project, steps to associate the central cadres of curriculum development, training of trainers and monitoring and evaluation in the work of CELL and the inclusion of GALAE officers in the CELL Coordinating Unit, which was itself housed in GALAE headquarters close to the Executive Director’s office.

For building capacity in the GALAE branch offices in the governorates, each Branch Director was to nominate a Liaison Officer, who would act as intermediary between the branch and CELL, as well as two Training Cadres, who would work with CELL in selecting communities for the project, implementing the PRA, forming the community teams, selecting and training community and class facilitators and supporting the classes and community teams for the duration of the project. Through association with all these processes, the GALAE officers, together with the community facilitators, class facilitators and community teams would form a core of new capacity on which GALAE could found the expansion of the CELL strategy.

Implementing CELL

The CELL project was to open in mid-2000 and close in mid-2003, a span of three years. The full span would have allowed a year for developing the curriculum, producing the materials required, identifying the governorates, districts and villages which would participate in the experiment, and carrying out the necessary training. The following two years would have allowed two full cycles of ‘basic’ and ‘pathways’ phases to take place, undergo proper monitoring and evaluation and enable GALAE to take a fully informed decision on whether or not to adopt the CELL strategy for the national literacy campaign.

In the event, a change of leadership delayed implementation for two years, but an extension of funding enabled the project to run one full cycle of the CELL curriculum between August 2003 and February 2005.

The PRA were carried out and by the end of 2003 a total of 5,305 students enrolled in 200 classes. By February 2004, 4,355 students remained in 189 classes. In August 2004 and subsequent months, 4,066 students enrolled in 173 ‘pathway’ classes.

The timing of each class was the outcome of negotiation between the students and class facilitator. As a result, CELL classes began as early as noon, or as late as 19.00 hours, rather than all being bunched in the evening. Prima facie, the CELL strategy had succeeded in providing and sustaining educational opportunities for rural communities.

Outcomes of the CELL Approach

Enrolments: CELL succeeded in its aim of enrolling a majority of females: the ratio in the ‘pathway’ phase was 75.6 per cent female, 24.4 per cent male. About half the female enrolees were under the age of 20 years and unmarried. Although some 13 per cent had some schooling, most had never been to school. In addition, class and community facilitators and community teams reported several instances where they had joined together to persuade husbands to allow their wives either to attend classes or to act as class facilitators. In one of the latter cases, the husband agreed on condition that the class should be held in his house. These reports suggest that CELL may have succeeded in opening access to education to larger numbers of females in such communities than the standard classes would have managed. They also help underline the value of winning the close and practical support of influential members of the community for efforts in adult education.

Attendance rates: No systematic data are available on attendance rates in either the standard or CELL classes. What is known relies on the impressions of GALAE and CELL personnel, including class facilitators. GALAE officers and class facilitators who had previously taught standard classes were confident that attendance rates in the CELL classes were higher. If the estimates of a number of class facilitators are reliable, attendance rates in the CELL classes seemed to range on average between 75 and 80 per cent. The importance of this observation lies in the positive correlation found between regularity of attendance – as reported by students themselves – and probability of passing the GALAE assessment.

Completion Rates: The tables which follow will compare the completion rates of the standard and CELL classes. The first table gives a summary of the evolution of the CELL classes.

The first notable point about the figures is that, despite the poverty and difficult circumstances of the experimental communities, the CELL classes were able to hold more than 80 per cent of their students during their ‘basic’ phase and a similar proportion during their ‘pathways’ phase. However, the corollary is that more than a third of the students ceased attending during the two phases of the course – despite the efforts of the community facilitators and community teams – and that only half of them were able to pass the GALAE assessment.

Table 1: Statistical Summary of the CELL Project, 2003-2005

Stage of CELL Project
Original enrolments, August 2003 – January 2004
Completed and tested ‘basic’ phase (5 months)
Enrolments in ‘pathways’ phase, August –September 2004
Completed and tested ‘pathways’ phase (5 months)
Pass GALAE assessment


Similar statistics are unfortunately not available for the standard classes. However, Table 2 summarises the statistics obtained. They show that in some of the governorates where CELL operates, the CELL retention rate is approximately 10 per cent better than the standard programme’s. They also show that both programmes succeed in holding the majority of their students during the second half of their courses. As with all educational programmes, including schools, these overall figures mask a considerable range of variability. For example, the CELL retention rates for individual classes range from 53.0 per cent all the way up to 100.0 per cent, while the standard programme’s retention rates for 2004/5 range between 56.7 per cent and 81.4 per cent. The reasons for this variability have not been investigated.

Table 2: Comparison of Retention Rates, Standard: CELL

Source of Statistics Retention Rate (Students tested as percentage of students originally enrolled)
GALAE statistics for 2002/3 for standard classes in the governorates of Aswan, Beni Suef, Fayoum and Sohag 71.7
GALAE statistics for 2004/5 for standard classes in parts of the governorates of Aswan, Beni Suef, Qena and Sohag 71.6
CELL statistics for 2004/5 for students in the second or ‘Pathways’ phase only 82.3


Graduation or Pass Rates: Pass rates are available on three differ- ent and independent measures. The first is of course the assessment conducted by GALAE. The second is a test applied to samples of GALAE and CELL students by the Social Research Centre of the American University in Cairo (SRC/AUC). The third is a test applied by the International Consultants Bureau (ICB) to additional and different samples of GALAE and CELL students. All three measures suggest that CELL is more effective than the standard programme in helping its students to learn how to read and write. However, it appears no more effective in helping them to do written arithmetic.

Table 3: Pass Rates as a Percentage of Enrolments for Phase 2 of Both Programmes

Source of Statistics

Pass Rate (Students passing assessment as percentage of number enrolled)

GALAE statistics for 2002/3 for standard classes in the governorates of Aswan, Beni Suef, Fayoum and Sohag 49.1
GALAE statistics for 2004/5 for standard classes in parts of the governorates of Aswan, Beni Suef, Qena and Sohag 43.4
GALAE statistics for 2004/5 for CELL ‘Pathways’ students on the GALAE assessment 64.7


Tables 3 and 4 express these statistics in two forms. Table 3 states the statistics on passes as a percentage of the students who enrolled for the second phase of each programme. Table 4 states the passes as a percentage of the second phase students who were actually assessed. Table 5 provides a summary of the data available on each of the three skills assessed, namely, reading, writing and written arithmetic.

Table 4: Pass Rates as Percentage of Students Actually Assessed

Source of Statistics

Pass Rate (Students passing assessment as percentage of students assessed)


GALAE statistics for 2002/3 for standard students in the governorates of Aswan, Beni Suef, Fayoum and Sohag 71.7
GALAE statistics for 2004/5 for standard students in parts of the governorates of Aswan, Beni Suef, Qena and Sohag 70.5
GALAE statistics on CELL ‘Path- ways’ students 2004/5 on GALAE assessment in Aswan, Beni Suef, Fayoum, Luxor, Qena and Sohag governorates 82.3
SRC/AUC statistics for 2004/5 for 10 standard classes (156 students) in Fayoum on SRC assessment 60.9
SRC/AUC statistics for 20 CELL ‘Pathways’ classes (288 students) in Aswan, Beni Suef, Fayoum, Luxor, Qena and Sohag governorates on SRC assessment 75.3


The data in Table 3 reflect the higher retention rate of the CELL classes as well as their higher pass rate on the GALAE assessment. They also suggest that there may have been some deterioration in the performance of standard classes on the GALAE assessment, since the data in

Table 2 show that the retention rate for the year 2004/05 was almost identical with that for 2003/03, whereas the data in Table 3 suggest a decline of nearly six per cent in the pass rate.

Table 4 documents that on both the GALAE and SRC/AUC assessments the CELL students had higher overall pass rates than the standard students by margins of 10 to 15 per cent.

The GALAE statistics on assessment are published only as consolidated pass rates. They do not show performances separately on the three skills of reading, writing and written arithmetic. However, both SRC/AUC and ICB do so: Table 5 summarises the outcomes of their assessments.

Table 5: Comparison of Pass Rates in Reading, Writing and Arithmetic – SRC/AUC and ICB Assessments (Number of Students passing as percentage of number of students tested)

Standard n = 64
CELL n = 91
Standard n = 156
(Fayoum only)
CELL n = 2881







Note 1: A T-test of means in reading, writing and written arithmetic for the two samples shows that the differences in means between the CELL and standard samples for reading and writing are statistically significant at the .001 level, but that the difference in the mean for written arithmetic is not statistically significant.


On both independent assessments, the CELL classes have higher pass rates in two of the three skills by margins that are much the same

as in the assessments by GALAE. The agreement of the three modes of assessment permits the inference that the CELL programme has indeed been more effective than the standard programme in enabling its students to learn the skills of reading and writing.

However, the two assessments differ substantially in their assessments of skills in written arithmetic. On the one hand, the ICB measure gives both samples of students much higher pass rates than they attained in reading and writing, with the CELL sample attaining a rate 14 per cent higher than its standard counterpart – indeed, almost attaining perfection. On the other hand, the SRC/AUC assessment finds that the standard sample has done about as well on arithmetic as it has on writing and, more importantly, that it has done slightly better than the CELL sample, which has a much lower pass rate than it achieved for reading and writing (the differences of performance on the three different assessments – GALAE, ICB and SRC/AUC – obviously raise questions about the assessments themselves).

The important inference is that there is a prima facie case that an approach to adult education and literacy that is based on community support and driven by the expressed interests and needs of the students is more efficient in terms of retaining its students and more effective in enabling them to achieve their learning goals than the standard approach.

Lifelong Literacy: While the CELL students did better than their standard fellows in reading and writing, it is not certain that majorities of them had achieved a self-sustainable level of skill that could be called ‘lifelong’ literacy. The reasons for the uncertainty lie in the nature of the assessments and in the way the assessments were marked. The SRC/AUC assessment asked students to read just one sentence, to write just one sentence dictated to them and to do one addition and one subtraction sum. As can be seen above in Table 5, at least a quarter of the CELL students did not achieve a pass mark, i.e. they even read and write with some difficulty. If a more stringent criterion is applied, that a student should read a simple sentence without difficulty and write a dictated sentence without difficulty, only 58 per cent would pass in reading and 64 per cent in writing, while only 37 per cent would pass in written arithmetic.

This tentative conclusion raises an issue for planning adult education with literacy. It is the question of ‘time on task’. Estimates vary on how many hours of guided learning and practice are needed to enable the average intelligent adult to develop adequate skill in reading, writing and written arithmetic. They range between 150 and 400 hours. In the case of CELL, classes reported meeting on five or six days every week in two phases, each of about 20 weeks. If that were so, the maximum number of hours a student would have spent in class would be 400 hours. However, virtually all the students sampled by SRC/AUC acknowledged that they had not been able to attend all their classes, and informal estimates by class facilitators suggested that the average attendance rate was about 75-80 per cent. If that information is used as a base for calculation, it would suggest that the average student attended class for some 300-320 hours.

That amount of learning led to a result in CELL – although superior to the result in the standard programme – that could not be deemed ‘lifelong’ literacy with certainty or even high probability. However, the average mark for reading and writing on the SRC/AUC assessment raises the possibility that an additional 100 hours of systematic learning and practice might enable the average student to improve her attainments in reading and writing to a level that would be credibly permanent. Should literacy programmes for Arabic-speaking rural adults then plan for programmes of 500 hours of class time on the assumptions (a) that the average adult student, however willing and eager, will not be able to attend more than 80 per cent of class sessions, and (b) that 400 hours of learning and practice will suffice to enable the average student attain permanent or lifelong literacy? The issue of written arithmetic needs more radical consideration. On the relatively simple SRC/AUC assessment, neither the standard programme nor the CELL project was able to assist more than a third of their students to do the two sums easily. It seems evident that many of the class facilitators in both programmes were not able to teach the matter, even though virtually all of them had completed at least a secondary education.

Correlates of Learning: Apart from levels of skills, the SRC/AUC samples of 288 CELL and 156 standard students offer support for a number of observations in adult education and literacy in other contexts.

The first is the finding of strong consistency among class facilitators in their effectiveness in helping their students learn how to read, write and do written arithmetic. Class facilitators who are effective in teaching one skill will almost certainly be equally effective in teaching the other two: the Pearson correlations are as follows (the statistical significance of all these correlations is at the .000 level, i.e. there is less than one chance in 10,000 that they occur by chance):

Reading : Writing .85
Reading : Arithmetic .69
Writing : Arithmetic .73

A facilitator whose class does well in reading has an 85 per cent probability of having it do well in writing also and a 70 per cent probability of its doing well in written arithmetic. Equally important is that fact that the variability of students’ performances on the skills is narrowed. That is to say the less able students are likely to do better with an effective class facilitator than they are with less effective facilitators.

The SRC/AUC sample reinforces this observation: seven of the ten ‘standard’ class facilitators saw their students do better than the students of six of the 20 CELL facilitators in reading. The importance of that statistic lies in the facts that the CELL facilitators had taken more training, enjoyed more technical and moral support, and had more freedom in applying their curriculum than their counterparts in the standard programme. It appears that an effective facilitator can do better with an ‘inferior’ curriculum than a less effective facilitator with an ‘improved’ curriculum.

The clear inferences, familiar to educators, are first that a good teacher will get good results in every subject that she or he teaches. She will also help students to make more of their abilities than a less effective teacher. Second, it is worth making special efforts to ensure that every teacher knows and practises how to teach well: good preparatory training, close support and good quality refresher training will likely pay off in terms of enhanced retention, completion and pass rates.

The second point is of special concern to curriculum specialists. A particular emphasis of the CELL curriculum is independence from standard texts and, instead, strong use of local themes and materials. In accordance with this principle, CELL did not supply its class facilitators with any material beyond the class facilitator’s manual. Many class facilitators and students complained about the lack of a primer, even though they grew accustomed to learning without one. Others simply obtained GALAE primers from sources of their own and wove them into their CELL teaching. The justification of this violation of their training is that in the SRC/AUC sample of 288 CELL students the strongest correlate – at the .000 level of statistical significance – of performance on the assessment was the use of textbooks in a class. The signal from this statistic may be that the disobedient class facilitators were able to use the materials to engage their students more effectively in reading and writing, to get them to practise more and so to help them make the most of the freedom that the CELL design offered. They made the best of both worlds.

The next strong correlate after the availability of textbooks will surprise no educator. It is regularity of attendance at class. In the absence of systematic statistics on attendance, the SRC/AUC sample observed it indirectly and crudely: it simply asked students if they had been regular in attendance. Those who said that they had been fairly regular tended to score better on the assessment – the statistical significance of the correlation was, however, slightly weaker at .005 rather than .000. This finding signals that the efforts to sustain regular attendance through the community facilitators and community support teams were soundly conceived.

Other possible correlates of performance on the SRC/AUC assessment, such as age, marital status, number of children, occupation, or distance of residence from venue of class, had such slight influences as to be statistically insignificant.

The preceding paragraphs have suggested:

  • Whether the CELL project or the standard programme enabled their students to attain ‘lifelong’ literacy is uncertain.

  • The average adult student learning literacy in Arabic may need a full 400 hours of learning and practice to achieve fluent reading and writing skills.

  • More work is needed on the effective teaching of written arithmetic.

  • The competence of the class facilitator is key to the learning outcomes of her or his class: facilitators who are not ‘natural’ teachers should benefit from better preparation, support and refresher training.

  • The CELL policy of not supplying supporting material for reading and writing needs to be reviewed.

  • Making special efforts to ensure regular attendance by students is likely to pay off in higher retention rates and better overall performances. However, assuring 100 per cent attendance by even the most willing students is probably infeasible, as is assuring 100 per cent retention. However, rates of 80 per cent and better on both indicators should be within reach.


The next question is whether the CELL programme achieved these advantages in efficiency and effectiveness at a cost roughly equal to the cost of the standard programme, or at least in line with the margins of its greater efficiency and effectiveness (10-15 per cent for retention, perhaps 20 per cent for pass rates).

A cost effectiveness study of the CELL Project by Adam Rorris (British Council, Cairo, March 2005) calculates that the total unit costs per enrolled student amount to E£146 for the standard programme and E£435 for CELL; that is, the latter is 2.98 times or 298 per cent more expensive than the former. (This estimate set aside the development costs and considered only the recurrent costs common to both programmes.) Using the pass rates given in Table 2 for 2004/05 for both programmes would yield the following unit costs per student passing the GALAE assessment (per graduate):

Cost per passing standard student       £E146/.434 = £E336.40

Cost per passing CELL student             £E435/.647 = £E672.33

Difference of cost per pass                  £E335.93

Difference of cost CELL/Standard        = 1.99

The data available thus indicate that in overall terms a pass by a CELL student is virtually twice as expensive as a pass by a standard student under current conditions.

However, the cost-effectiveness study declares that CELL need not be so expensive. Indeed, it proposes ways to reduce the unit costs of CELL even below those of the standard programme. It points out that, whereas the chief costs of the standard programme lie in teacher salaries and instructional materials, those of CELL arise through training, community development, monitoring and support; and particularly through the very rich ratios of support given by the community facilitators and field coordinators. Table 6 summarises the calculations of the study.

Table 6: Comparing Unit Costs per Enrolled Student, Standard:


Cost Item
E£ – Standard
% – Standard
% – CELL
Teacher salaries
Instructional materials
Monitoring and support
Teacher training
Community development

Source: Rorris, 2005, pp. 26-27

On the basis of these calculations, the study proposes steps that would reduce the costs per CELL pass to levels well below the current costs per pass of the standard programme. The study is of course unable to predict whether raising the ratios and reducing the supports that CELL currently provides would affect either efficiency or effectiveness; it necessarily assumes steady levels of performance.

Enhancing Capacity

One of the crucial capacities that should be available to an experimental project that aims to transform national policy and practice is the ability to monitor and evaluate processes and outcomes reliably. It was unfortunately not available to CELL. Also, the efforts to enhance what capacity existed in the CELL Coordinating Unit and in the private market fell short of what was needed.

In contrast, the capacities of personnel in the GALAE branch offices and in the 22 experimental villages appear to have been enhanced considerably. Intensive interviews with GALAE Branch Directors, Liaison Officers and Training Cadres, and focus group discussions with community facilitators, community teams and class facilitators revealed an almost unanimous conviction that the CELL strategy was more effective than the standard and should be adopted on a national scale. Further, everyone who had participated in helping to launch it in the experimental villages had experienced changes in perspective and attitude to adult education and social transformation, quite apart from gaining skills in:

  • Participatory Rapid Appraisal
  • Negotiating with the leaders and people of poor rural communities
  • Designing and delivering training courses that promote active learning through modelling behaviour rather than lecturing about it
  • Creating learning programmes from local issues and materials
  • Learning how to capitalise on the experiences and views of women and men who had never been to school
  • Developing supportive rather than inspectorial approaches to supervision
  • Teaming with community members to sustain the momentum of a project

Signals from the CELL Experience

This final section aims to draw together the signals for adult educators.

The experience of the CELL project confirmed that:

  • Basing a literacy curriculum on the interests, surroundings and convenience of a particular local group of students is effective in helping to sustain good attendance rates, retain higher proportions of students and achieve satisfactory results on official assessments.
  • Using the techniques of Participatory Rapid Appraisal can be very helpful in identifying the core interests of potential local students and in identifying those members of the community who are likely to be most effective in mobilising support for adult education.
  • Subject to the requirements of quality and effectiveness, recruiting members of the community as class facilitators and community facilitators can make an educational initiative more acceptable to communities.
  • Negotiating the timing and venue of classes with students and class facilitator can be helpful in encouraging higher attendance and completion rates.
  • Successfully conducting a curriculum in reading and writing on the basis of a framework of core local interests, without the aid of a set syllabus of content and primers, is practicable. However, it is advisable to assist classes to obtain reading and writing materials to support their learning and practice.
  • The experience of CELL suggests that doing without a text is less practicable for teaching written arithmetic.
  • Such a curriculum is within the competence of facilitators who have eight or nine years of schooling (as found in Egypt) augmented by good quality training in facilitating adult education and by close support in dealing with unexpected problems.
  • Class facilitators will vary among themselves in effectiveness; those who are successful in teaching one subject will likely be equally successful in teaching other subjects; those who appear to be less effective will need to be helped with additional support.
  • Raising the effectiveness of class facilitators will raise the average attainment of students and diminish the variability between them.
  • Community support can be very helpful in recruiting students, forming classes, overcoming cultural resistance to the education of females, dealing with issues that arise in classes, maintaining attendance and resolving particular local social issues.
  • Personnel like community facilitators and field coordinators are necessary to sustaining the engagement of community support, on the one hand, and on the other, the morale and effectiveness of class facilitators.
  • Permitting extended breaks between phases of courses leads to losses of students and higher drop out rates.
  • Approximately 500 hours of classroom instruction and practice may need to be scheduled to ensure that most students do take at least 400 hours of them to achieve permanent or ‘lifelong’ literacy in terms of fluency in reading and writing, at least in Arabic.
  • Relying on the voluntary principle seems to be sound in the case of people like the members of the Community Support Teams, whose services were needed mainly on an occasional basis and to solve specific problems. The Egyptian practice of not relying on the voluntary principle for long term, regular and routine matters like class facilitation and community facilitation seems prudent.