There is a serious shortage of research on the inter-relationship between adult literacy and children's education (see, for example, Barton 1994). Since Jomtien many governments have invested in primary education and have not invested equally in adult literacy. Since it is parents who have to pay for their children's education (even when it is supposedly free there are indirect costs, and increasingly there are also direct costs) the attitude of parents to education must be important. Parents are also key to creating a literate home environment (and wider community) to reinforce learning in school - and the involvement of parents in school committees is increasingly seen as a key to improving the quality and local accountability of primary schools.

The lack of research on the impact of adult literacy programmes on children's education is thus both surprising and worrying. The evaluation of the REFLECT pilots was seen as an opportunity to contribute to this debate - particularly in Uganda where parents are expected to pay many of the direct costs of their children's education.

UGANDA

In Bundibugyo the formal education sector has been poorly resourced by central government (even in comparison to other parts of Uganda) and there is some resistance to schools by local people who dislike the fact that their children are taught in Rutoro. The REFLECT programme was the first adult literacy programme in Bundibugyo of any significant scale and the evaluators were therefore keen to determine the impact it had on attitudes to children's education.

In practice the evaluators found that there had been a significant impact. Not only had attitudes to education significantly changed, but parents, within the space of one year were taking practical steps to educate their children. This was seen by significant changes in three respects:

• an increase in enrolment of children in government primary schools,
• increased attendance at other pre-existing schools, and
• the establishment of new non formal education centres.

1. Government Primary Schools

The table below shows the changes in enrolment in government schools fed by families where some of the adult members are REFLECT learners. The five schools near REFLECT classes have been compared with fourteen schools in the same sub-county with no local REFLECT literacy groups. It reveals a 22% increase in enrolment in schools within the catchment area of REFLECT circles compared to just a 4% increase in schools with no REFLECT circles but with a similar catchment area in other respects.

Comparison of the Increase in Enrolment in Government Primary Schools by REFLECT, and those not Fed by REFLECT.

No factors were identified to account for the difference between these schools other than the presence of the REFLECT circles. The rate of increase summarised on this table was uniform from Primary I to Primary VII, (although the numbers enrolled are much greater in Primary I - IV) and this was true for all schools, not just the REFLECT five. Despite the very positive influence on increasing overall enrolment, it should be noted that there was not a specifically positive effect on the enrolment of more girls (whose enrolment increased more or less in the same proportion as boys).

2. Other Pre-existing Schools

Faced with inadequate government provision some communities have established primary schools for themselves, without Ministry of Education recognition (though this was often being sought). These schools received a fresh impetus from the REFLECT programme with parents sending their children in large numbers. Cases where participation has increased dramatically were Budikuliya: 256 - 360; Mutogo: 103 - 260, and Hakitara: 35 -120. There appeared to be a particular increase in girls' attendance at these schools. The REFLECT facilitators had often become active as leaders on the Management Committees of these schools (and in some cases of the government schools).

3 New Nursery Schools/ Non-Formal Education Centres.

Another development was the starting of new NFE Centres (or "nursery schools" as they were called locally) for younger children using the REFLECT literacy shelter, the same blackboard and some of the same teaching materials. Usually parents paid the REFLECT facilitator to teach their children. Although called Nursery Schools, they contained all younger children up to 9 or 10 who had no previous chance to go to school. Nine classes in the sample of 24 (ie more than one third) had started nursery schools, having as many as 60 or 70 children in each. It was interesting to note that facilitators were trying out REFLECT methods with the children as an alternative to their more dimly remembered primary education model.

The combination of these three outcomes represents a substantial impact, in the short term, on children's education in Bundibugyo. The creation of new schools in one third of sampled communities, in the space of just over one year, is particularly notable.

Although such changes are almost certainly the product of the whole REFLECT process many circles said that the new initiatives emerged specifically from discussions surrounding the construction of mobility maps. These maps involved identifying the different places people went to and the reasons for going (eg for markets, employment, health care, visiting relatives etc). They were designed to promote discussion of isolation which had been identified as a key problem in the initial research by ACTIONAID in Bundibugyo. The discussions which emerged often focussed on the education of children which was felt to be the only real way to lessen the isolation of the area - because children could migrate to work (and be visited by family members) or get their education and then come back to take salaried jobs in the area. This would mean communication of the most useful kind with the rest of Uganda. The Bwamba have always experienced people from the Toro ethnic group in positions of responsibility in offices and schools, and in positions of political authority. The REFLECT circles enabled them to address how this could be overcome through investing in the future.

UGANDA Mobility map - showing different places that people go and the reasons for going. Discussion often focussed on the education of children as a key means to reduce isolation.

It remains to be seen how sustained the increased enrolment will be and whether it will lead to improvements in achievement. One positive sign is that, although the language of instruction is still officially Rutoro the Ministry of Education are now considering the recognition of Lubwisi in the early years and informally it is already widely accepted. This is likely to make parents feel more positive about school education for their children. Moreover, many REFLECT learners are already taking an active interest eg through participating in parent teachers associations. Their experience of the REFLECT methodology is likely to give them a sense of good teaching practice and they will look very closely at the teacher's behaviour, having the confidence to criticise when they think something is wrong.

BANGLADESH

In Bhola most women identified helping their children as one of the primary motivations to learn literacy. The only Unit directly relevant here was the Education Matrix which led to some discussions though in most cases children were already enrolled in school. In a handful of cases children have been enrolled for the first time in school since the REFLECT circles. However, a better indicator might be greater regularity of attendance at school and this was indicated verbally by many women though it proved too complex to measure statistically.

Interestingly the most common thing which women have actually read since the REFLECT programme is their children's school books (71% of women have done so). Many women commented that they try to help their children who are attending school. Although unable to help children who have reached higher grades they now feel able to help those in the first three grades.

When asked how their husbands felt about their participation in the circle the responses from one community (CU115) were particularly revealing. Their husbands are actively supportive of their wives in the REFLECT Centre because they feel that their wives will then be able to help their children. This was commented as something that was becoming an essential part of the "role of being a good wife."

EL SALVADOR

The evaluation of REFLECT in El Salvador did not address the impact on the education of children in any detail. However, in some communities it did emerge as an issue for local organisation, with many learners assuming positions of responsibility on PTAs. In other communities like El Carmen, there is no existing school and the REFLECT circles acted as a focal point for discussion about how to establish one. Although no new schools were actually started up, some steps were taken to put pressure on the local Ministry of Education to increase their coverage. Sadly this failed to solicit a response as the education budget was said to be already over-stretched just to cover existing provision.