Designing a program

Given the right opportunities and the right learning environment, children will develop in similar ways whatever their background. As long as we keep in mind that everything we do is concerned with the development of the whole child, we are all doing the same sorts of things for the same sorts of reasons.

- Dr. Stephen Ngaruiya (as quoted in Bernard van Leer Foundation 1994, p. 9)

Countries everywhere find it difficult to implement programs for preschool-age children on a large scale. Experience has shown the importance of first fitting program approaches to societal needs and sociocultural characteristics and then defining clearly, from the outset, the processes for targeting the children most in need, training staff, and monitoring and evaluating programs.

Targeting. To achieve maximum results with limited funds, interventions must be targeted to reach only those children most in need and most likely to benefit. Proper targeting is an essential part of any social service program. If, for instance, a country has a severe problem with infant malnutrition, it would be inefficient to provide food supplements to all children under the age of fifteen. But by narrowing eligibility on the basis of age group and nutritional status, the program would be able to reach more of those most in need.

Thus one of a preschool program's first tasks is to establish the criteria for eligibility. The North American experience indicates that the poorest children benefit most-in both psychosocial and educational terms-from early child development programs (Campbell and Ramey 1994; Schweinhart 1992). Early child development programs should therefore be designed to reach the largest possible number of children living in poverty or in communities marked by a high prevalence of malnutrition, for these are the children who risk having their mental, social, and emotional development delayed or even permanently stunted.

Programs seeking to achieve different objectives also need to target by age as well as by income. Programs in the United States (WIC, Head Start), the former Soviet states, and India target very young children as the group most vulnerable and most likely to benefit from an intervention program. Children aged three to six, for instance, may need activities to prepare them for school, while for those under the age of two, proper nutrition, health care, and psychosocial stimulation are more critical. The U.S. eight-site Infant Health and Development Program, which targeted low-birth-weight and premature infants, demonstrated how early intervention could enhance these infants' cognitive, behavioral, and physical development.

Programs may also need to define eligibility criteria for services offered to mothers. India, for instance, considers women who are pregnant, lactating, or between the ages of fifteen and forty-five eligible for Integrated Child Development Services (ICDS).

Finally, countries can limit services geographically to make sure that they meet the needs of particular populations or settlement groups. India, for instance, targets areas inhabited predominantly by disadvantaged tribes and scheduled castes (India, Department of Women and Children 1988). Colombia has extended its urban slum programs to reach low-income rural areas as well. And Mauritius's Export Processing Zone (EPZ) Labor Welfare Fund, which covers only the families of EPZ workers, locates its child care centers within or near EPZ industrial estates (table 3).

While limiting eligibility for particularly expensive services is another way to limit costs, targeting is not always easy. India, for instance, has attempted to restrict its expensive supplementary food services to mothers and children who are demonstrably malnourished. But almost half of ICDS supervisors reported that food services were delivered to women and children who did not qualify for them (NIPCCD 1992).

Involving parents and communities. Colombia and India have both noted that rates of use of early child development programs are low among eligible populations-in large part, according to a recent evaluation study in India, because the target population was unaware that such services were offered. In addition, working mothers in India found it inconvenient to attend two- or three-hour sessions during the day. And even Colombia's day long child care was not sufficient where parents worked more than eight hours a day. To be useful to families, therefore, day care programs must be both well advertised and tailored to meet local needs.

Table 3 Targeting characteristics in four countries

Source: Wilson 1995.

Many countries claim to have tried to involve community groups and parents in program development and delivery. But despite the putative emphasis on local involvement, community participation in India's ICDS program, for instance, remains marginal. Even after fifteen years of program operation, most of the community representatives questioned had never been consulted during the initial stages of a preschool education project, and many were unaware that their participation was needed.

One way to involve community members is to use volunteer caregivers. But relying on volunteers can be problematic because it sends the message that preschool programs are of little importance and that care giving is not a viable career. As a result, many volunteers are dissatisfied and seek to be recognized as government workers, and the turnover rate is high. In Colombia, for instance, where the government pays each home day care mother a small stipend for each child in her care and parents pay an additional fee, home day care workers' earrings are capped at 65 percent of the average national income plus social security contributions and access to home improvement loans. Although such loans are granted for up to five years, home day care mothers generally stay in the Colombian program only two years.

Indian caregivers and their helpers receive honorariums set at 75 percent of the national per capita income. The Indian caregivers' union, created in the 1980s, has been pressuring the government to admit them into the civil service. In response, India now proposes to consider preschool caregivers with secondary education and ten years of experience for government appointments as supervisors and project officers.

Defining a curriculum. Some experts on early child development have suggested that, provided a child receives proper care and has interesting activities and other children to play with, the actual type of preschool experiences matters very little (Osborn and Milbank 1987). Increasingly, however, experts favor approaches that are more educational, maintaining that failure to adopt a specific viewpoint or curriculum model leads to poor quality education.

Early child development programs need to have a curriculum that is well defined-one that is designed to promote children's development and based on validated models, and that emphasizes children's choice, decisionmaking, and active learning. As illustrated in table 4. approaches are likely to differ on the basis of people's beliefs about how children learn and the role of the teacher. Teachers and caregivers need to understand the curriculum selected and be trained to make decisions about what to do in the classroom based on their understanding of the curriculum's theoretical base. They will therefore need ongoing training and supervision from someone trained in that educational technique. The underlying theory will govern staff training and the methods used to deliver the service.

Table 4 Three approaches to the education of young children based on three different theories of learning

a. Based on Jean Piaget's constructivist theory.

b Traditional center-based method, based on the British Infant School movement

c. Uses behaviorist model of didactic instruction reinforced by systematic awards.

d. Not considered an educational approach

Source Based on Roopnarine and Johnson 1993.

Structured programs have the added advantage that they can be evaluated on the basis of existing standards for interaction and activities for young children, and methods proved to be successful in the field can be widely applied. By contrast, unstructured collections of ideas— even good ideas cannot be evaluated or replicated, and the quality of the outcomes cannot, therefore, be maintained. Adopting a specific, valid, theoretical approach is therefore the first step toward instituting an effective early child development program.

While it is not yet possible to rate the effectiveness of different approaches on the basis of evidence from developing countries, data from the United States and England suggest that which active learning curriculum is chosen is less important than whether active learning techniques are used consistently and well.

The following guidelines are offered for the implementation of any early child development program:

· Design a program that meets children's educational and psychosocial needs as well as their physical needs. Children most need a supportive environment in which they are free to choose their own learning activities, to take responsibility for completing them, and to talk about what they have accomplished.

· Train in-service staff to carry out this program.

· Provide supportive supervision.

· Keep groups to no more than sixteen to twenty three- to five-year-olds for every two adults, twelve two-year-olds for every two adults, or eight one year-olds for every two adults. Most important is not necessarily the size of the group of children but that there are two adults.

· Emphasize the need for staff to be sensitive to children's physical, health, and nutrition needs and their families' child care and social service needs.

· Evaluate procedures for their appropriateness to different stages of the children's development.

Low-income developing countries that wish to begin focusing on hygiene, nourishment, health, and mental stimulation could set up integrated early child development programs that:

· Provide children with a secure space for movement and play.

· Provide flexible multifunctional furniture that can be handled by children.

· Provide adequate food and clean water for drinking and washing.

· Provide children with material especially adapted for handling, experimenting, and playing.

Selecting and training staff. The quality and consistency of child care staff have been identified as among the major determinants of a program's efficacy. The U.S. National Child Care Staffing Study (Childcare Employee Project 1992) documented that children in centers with rapid staff turnover spent less time engaged in social activities with peers and more time wandering about aimlessly, and showed disturbing lags in their social and language development. A study commissioned by the California General Assembly found that in classes in which the ratio of children to teachers was too high, children were less involved in classroom activities. Better outcomes for the children were significantly correlated with higher staff wages and benefits (Childcare Employee Project 1992).

Despite these findings, however, the fact remains that most preschool programs in developing countries rely heavily on volunteer caregivers. And because they are volunteers, the criteria for their selection have been relatively flexible. In India, for instance, caregivers are expected to be literate, but illiteracy has been cited as a serious problem (NIPCCD 1992).

Colombia's home-based program requires only that caregivers be twenty to forty-five years of age, have no more than two young children, and own their homes-although women with writing, reading, and basic mathematical skills are preferred. Those interested in becoming caregivers must attend a forty-hour workshop offered by national child development representatives. The most promising candidates are then selected by the national representative in concert with the parents association managing the program locally. In Kenya preschool teachers are selected by parents and local authorities and are expected to receive six training sessions, totaling eighteen weeks, over a two-year period. Yet in 1993 less than 40 percent had completed their training. A major factor was lack of funding. In India, too, trained government program officers and supervisors have had difficulties passing along what they have learned to local providers. The reason they cite is that sites are inaccessible and too numerous for trainers to be able to maintain the schedule. In Mauritius, by contrast, where the preschool programs do not rely on volunteers, staff have at least a secondary education and are recruited locally.

To upgrade staff training and improve the quality of early childhood education programs, policymakers must therefore:

· Define and adopt effective training practices.

· Set up a system for training teachers throughout the country.

· Schedule pre-service and in-service teacher training.

· Identify professional and paraprofessional caregivers who should receive training.

· Make training requirements flexible enough to suit a variety of early childhood program settings.

· Make training requirements flexible enough to be adaptable to the needs of different cultures.

Assessing children's progress. Because young children need different inputs at different stages of their intellectual and emotional development, preschool caregivers must screen their progress regularly. But this is no easy task.

Developmental measures for young children are notoriously unstable, and no single indicator can accurately capture a child's well-being or readiness to learn. School performance is affected by many factors other than cognitive ability-such as physical well-being, emotional security, social confidence, language fluency, learning disabilities, interests, engagement, approach to learning, the presence of a parent or guardian to explain things, and general knowledge and skills. The danger always exists that assessment tools will be used to exclude or track certain classes of children, making it undesirable to use standardized tests to signal readiness for school. The only defensible criterion for school entry is attainment of the legal chronological age.

But in early childhood development programs continual screening is a necessity. A recent review of instruments for assessing children's psychosocial development recommended that children be tested and regularly reevaluated for specific indicators, determined by each program's objectives (Landers and Kagitcibasi 1990). Working together, the Tufts University School of Nutrition in Boston and the Diponegoro University School of Medicine in Java devised a "cultural-specific inventory of child development milestones," derived from focus group discussions, concept testing, and formative evaluations (table 5). Now thoroughly tested and refined, the Tufts-Diponegoro chart is being used to study the progress of children under three as part of Indonesia's National Growth Monitoring Program (Landers and Kagitcibasi 1990).

Table 5 Child development chart used in Indonesia

Source: Landers and Kagitcibasi 1990.

Assessing a program's success. Early child development researchers today can confidently assert that enriched early childhood experiences permanently enhance children's competence in everyday life-that is, their ability to meet social expectations, advance appropriately in school, stay out of trouble, and have high aspirations for themselves (Schweinhart and Weikart 1980). But contrary to earlier theories, this benefit-as the High/Scope Perry Preschool study has shown-does not derive from improved intellectual performance (Schweinhart, Barnes, and Weikart 1993). The success of an early child development program therefore cannot be measured by whether or not participants' IQ scores rise five or ten points.

Katz (1992) proposes that preschool programs be judged in large part by children's long-term satisfaction with them, as reflected in the degree to which they feel intellectually engaged and respected. High/Scope researchers, again in the High/Scope Perry Preschool study, further maintain that children from successful early childhood interventions experience greater success in their first weeks in school, which then increases their motivation and leads to better school performance and higher regard from teachers and classmates. More successful school careers in turn increase the chance that children will graduate from high school, hold jobs, and not turn to crime.

Since early child development programs ideally integrate a variety of services, assessment indicators are required in the areas of health, nutrition, psychosocial and cognitive development, and parental and community factors. Specific indicators relevant to each program's objectives must be identified at the outset so that mechanisms for monitoring and evaluation can be put in place (table 6).

Preschool programs in Bolivia, Indonesia, and Turkey, for instance, are being judged on such outcome variables for participating children as anthropometric data, immunization status, psychosocial development (Bolivia uses a specially adapted development scale), and schooling (age at entry, dropout and repetition rates, academic performance, behavior). Outcome variables for participating mothers include personal health, participation in the labor force, education level, knowledge of child-rearing practices, and self-confidence. A valuational instrument sensitive to the many and varied effects of early child development programs is badly needed, both to help standardize the evaluation of such programs and to make supervision of them more effective.

Most early child development projects have been able to do little more in the way of self-evaluation than to list the number of their participants and describe the extent and content of services offered. Impact evaluations tend to be limited. They are by nature difficult because the time lag between the intervention and the expected benefits is far longer than the duration of projects, often five to seven years.

Colombia's Hogares Comunitarios de Bienestar (Community Child Care and Nutrition Project, or HCB), for instance, certainly helped many children (UNICEF 1993). But attempts to assess the project's impact more precisely were skewed because they compared groups of children from the same project (that is, those who had attended the program for three months or less with those who had attended nine months or more), underestimating the fact that the groups would reflect the same selection bias. Yet the evaluation has proved useful in further improving program implementation. Children who had attended for a year, it was found, still suffered nutritional deficiencies at a rate above the national average suggesting a need to improve community food purchasing and delivery systems. A later census confirmed that fully a fifth of HCB homes had no program food supplements on hand and that nearly a third had not received the complete nutritional benefits offered under the program.

Table 6 Indicators for assessing the success of early child development programs based on potential benefits

Source: Myers 1995.

Box 6 Indicators used in Bolivia

Bolivia's Integrated Child Development Project-supported by the World Bank, U.S. Agency for International Development (USSAID), and public and private donors-targets poor households in urban areas of more than 10,000 inhabitants. The project proposes to install 9,000 day care center between 1993 and 1999. Every year these centers will provide informal, home-based, integrated child development services to more than 200,000 children aged six months to six years-roughly 16 percent of all Bolivian children living in urban areas. An estimated 16,000 pregnant women will also benefit directly from the project, either as caregivers or as the mothers of children in project day care centers.

To improve information on Bolivia's young children, the project will conduct an integrated household survey, which will allow it to compare outcomes for participants and nonparticipants. The project's management information system will include detailed information on day care administration and financing, the number of day care centers available, the number of caregivers and of children enrolled, the duration of children's attendance, and children's health and developmental status (as defined by the presence of health cards' nutritional status, vaccinations, number of health visits, weight, and other developmental indicators).

By contrast, Bolivia's Integrated Child Development Project started off with a project impact evaluation scheme in place (box 6). From the outset to the end of the project, data are being gleaned from household surveys of both participants and nonparticipants. Evaluators will use this information to assess the project's success at targeting high-risk children, women, and families; its effect on the development and well-being of both participants and communities; and its effect on children's later performance in school.

It is somewhat easier to evaluate early childhood interventions whose major thrust is improving the diet and health of mothers and children (see description of WIC program in part II). India's massive.

Integrated Child Development Services, for instance, has been evaluated frequently over the years. A recent review of nearly thirty nutrition impact studies confirmed that the ICDS program-even operating at a minimal level of quality-has had a positive impact on children under six. A 1984-86 comparative study found higher infant mortality rates.

Box 7 India's first Tamil Nadu Integrated Nutrition Project

Twenty years ago 85 percent of children under six in Tamil Nadu were underweight, and malnutrition contributed to three-quarters of all infant and toddler deaths. India's first Tamil Nadu Integrated Nutrition Project (TINP I) was designed to reach the populations most at risk-weaning children' pregnant and nursing mothers, and adolescent girls. One of the largest nutrition projects in the world, it reached more than 13 million people from 1980 to 1989. Its US$81 million cost was funded in part by a US$32 million International Development Association approved loan in 1980.

The project's goals were to halve the protein-energy malnutrition rate (roughly 60 percent for mothers and children in the state's worst-hit rural areas), to reduce by a quarter the infant mortality rate (then at about 125 per thousand), to reduce the rate of vitamin A deficiency in children under five from about 27 percent to 5 percent, and to lower the rate of nutritional anemia in pregnant and nursing women from about 55 percent to 20 percent.

Program design

The project provided a package of services in nutrition education, primary health care, immunization, and supplementary on-site feeding for severely malnourished children. In an innovative move, the project offered low-weight infants and toddlers-identified through monthly weighing-supplementary food for short periods to help them recover their growth, rather than offering prolonged feeding programs for older children. Closely supervised women, locally recruited and trained, delivered education and other services to 1,500 people at each of the project's 9,000 community nutrition centers. TINP I's high supervisor-to-worker ratio and intensive on-the-job training proved critical to Its success.

What worked, what didn't

Identifying inappropriate child care practices as the root cause of most malnutrition, the project sought to change these practices through education. Problem cases were identified through monthly weightings, where mothers and nutrition workers could also meet to discuss concerns and exchange information. World Bank evaluators found that frequency of weighing, lower age at enrollment, and immunization were all significantly associated with improved nutritional status, while deworming and vitamin A supplements were not. TINP I recurrent costs have been estimated at US$ 9.41 per beneficiary for 1984 85, of which about 47 percent went for nutrition (30 percent for food) and 43 percent for health care (Berg 1987).

Under TINP I, weight-for-age improved steadily in Tamil Nadu, and malnutrition rates dropped for all ages served by the program. Children aged six months to three years gained more than 6 percent in weight, and these improvements lasted. Half to three-quarters of these gains are credited to TINP. Monitoring indicators suggest that the procedures for beginning and ending supplementary feeding, providing educational inputs, encouraging community participation, and keeping accurate records were carefully followed and that leakage to nonparticipants was small. Since scheduled-caste children's participation increased from 37 percent to 42 percent and the average age at enrollment declined, it also appears that the program became progressively more successful at reaching the neediest.

Lessons learned non-lCDS areas, and a study on the effects of primary education found that ICDS participants had higher primary school enrollment rates and better attendance and performance records. Among children from low and middle castes, primary school dropout rates were significantly lower for those served by ICDS (Consultative Group on Early Childhood Care and Development 1993). Another extensively evaluated program is the Tamil Nadu Integrated Nutrition Project (box 7).

TINP I has shown that education can change unhealthy child-rearing practices and that narrowly focused targeting is cost-effective. The project's success rested on:

· The careful selection of community nutrition workers.

· Detailed but simple work routines that emphasized outreach.

· Training to develop skills.

· Intensive and supportive supervision.

· Efforts to gain community support and generate enthusiasm for the project.

Source. World Bank 1995.