Quality is an elusive concept. In an ECCD program there are at least three ways of defining it: by the inputs, by the process followed, and by the results achieved. But quality is relative, and its definition varies from place to place and over time.

Quality as inputs
Too often the quality of a program is defined only by rough measures of the inputs, such as the formal qualifications of a promoter or education agent, the availability of materials, and the kind of facilities in which the program is carried out. These definitions are useful—but superficial.

Formal qualifications usually indicate that education agents have good potential for providing quality attention to children. But more important is what teachers do and whether it has a positive effect on children. A well-trained, duly certified preschool teacher may be authoritarian, insensitive, directive, unable to relate to parents, long on theory, and short on practice. And an uncertified teacher may be democratic, sensitive with children, and good with parents.

So we need to apply standards for quality that go beyond paper qualifications and material inputs. We need to ask what makes for an effective early intervention that produces the desired results.

Quality as process
When quality is defined by process, what is important is that the process meet generally accepted standards of what should occur in an ECCD program, whether it is for children or for parents. A good-quality process might be defined, for example, as one that treats children with affection, emphasizes active learning, respects individual differences, provides a chance for both independent and group experiences, offers a variety of stimulating possibilities for the child, and avoids physical punishment (see the discussion on curriculum in box 4.1). A definition of quality might also emphasize the participation of parents throughout the process, beginning with setting aims and goals.

This view of quality can be limiting if it results in a program being "decontextualized"—that is, if a program responds to general ideas about how a child should develop and how care should be provided, but without making the local adjustments needed to be meaningful and appropriate and without considering conditions in the home and community that may be incongruent with the program. Involving parents should help solve this problem, but it is no guarantee. Consider this evaluation of the nutrition component of Colombia's home day care program.

Observation of the process in the program's day care homes found that children were being fed nutritious food and usually at the level determined sufficient to provide them the nourishment needed for proper growth. But the evaluation showed that the program did not affect the children's nutritional status. The reason? Parents fed their children less at home instead of viewing the school feeding as supplementary. Thus children did not receive supplements on weekends and diarrhea took its toll. In this case the definition of a quality process needed to extend well beyond the day care home.

Quality as results
Some people believe that an ECCD program cannot be good quality if it does not produce the results it is expected to produce. Clearly, a central problem is deciding what the results should be. Parents, "experts," and government bureaucrats may have very different perspectives. Western-trained experts might set their sights on increasing the expression of independence by children. Parents might be more concerned about instilling obedience, adherence to group norms, and socialization to solidarity.

An attempt at synthesis
A review of the literature on the quality of early education programs provides a basis for a rough attempt to describe elements of quality that can be seen as one perspective of what is desirable (see box 4.1). Each of these elements deserves discussion beyond what this guide can provide (for further discussion see the references on quality listed in tab 8). These dimensions of quality have all been linked by research to effectiveness, but include elements of both process and inputs. An attempt has been made to avoid specific quantitative definitions (such as a particular ratio of adults to children, or certified training). But it will be immediately clear that the elements imply a set of cultural and social values that while widely shared are by no means universal. Nevertheless, the set of elements can provide a useful point of departure for a discussion of quality with stakeholders.

The relative nature of quality
If the standards set out in box 4.1 for quality ECCD programs were applied systematically in evaluating large-scale programs of early education and care operating in Latin America (or anywhere else), most if not all programs would be judged low quality. Programs often set out clear aims, but these goals, imposed by a central bureaucracy, are not necessarily shared. Education agents often lack motivation, good training, and sensitivity, and turnover among them is often high. Curricula are often integrative, active, and meaningful in name but not in application. The ratio of children to adults in Latin American preschool programs generally exceeds 20 and is often much higher, making frequent interaction and personal attention difficult. The learning environments are frequently makeshift, lacking adequate sanitary facilities and far from secure. Supervision is usually equivalent with inspection-and rarely associated with continued on-the-job training for the adults who work directly with children. There is virtually no systematic application of validated evaluation methods to adjust educational activities. In large-scale preschool programs parental participation is at best weak and at worst missing.

Box 4.1   Elements of quality in early education programs Elements that define quality and have been associated with effectiveness in early education programs include the following: Aims and objectives. Clear aims and objectives set and shared by teachers and parents, understood by children, and subject to modification through a process involving all interested parties. The process for agreeing on the aims and objectives may be more important than the outcomes. Education agents. The continuous presence of sensitive, healthy, committed, loving, and responsible adults who, as a result of experience and training, are knowledgeable about how children develop and who interact with children in a consistent, respectful, supportive, and unthreatening way. Curriculum. A proven curriculum that takes a holistic view of a child's development; provides a variety of relevant, stimulating, and enjoyable learning experiences for both setting roots and learning to fly; encourages children to play, explore, and initiate their own learning activities; and respects and attends to individual differences. A quality curriculum integrates education and care, attending to children's physical, social, and emotional needs as well as to their cognitive and intellectual needs. And it fosters sound relationships of the child with self, with others, and with the environment. Physical environment. A clean, ventilated, stimulating, secure, and healthy environment providing enough space for children to play. Evaluation. Use of systematic and validated evaluation methods by education agents and parents to adjust teaching to children's needs. Ratio of children to adults. A ratio low enough to permit frequent interaction and personal attention when needed. Training and supervision. Meaningful training on the job and supervisory support fostering continued professional and personal growth. Program leadership. Strong leadership that devotes much time to coordinating and managing yet stays close to the daily process of educating and socializing children. Parental and community participation. Real involvement and participation of families and communities as partners in the program, helping the program to set appropriate standards, to function well, and to adjust to local conditions and needs at the same time that they learn to improve their attention to young children. Resources. A consistent and permanent financial and material resource base sufficient to support working in an appropriate way with children and to sustain educational activities so that education agents need not be distracted from their immediate task of educating children. Source: Ball 1994; Moss and Pence 1995; Schweinhart 1997; NAEYC 1986; Basili 1994.

If correct, does this reading of quality mean that ECCD programs (or large-scale ECCD programs) are inevitably poor quality and ineffective? Certainly not. In a review of 25 ECCD programs, including several large-scale ones, Lira (1994) found that programs were often judged to be effective. Further, she concluded that the educational effectiveness of preschool programs is not tied closely to the locale, or the kinds of materials used, or the curriculum theory followed. Effectiveness was instead found to be closely related to the quality of the education agent (p. 216). That does not mean that the education agent must be a professional with a formal certificate. Neither is it sufficient to be simply a person from the community, with no training151;no matter how motivated and loving. But people with varied levels of education can become effective education agents if they acquire the abilities that characterize an effective educator. A community member can be transformed into an effective education agent through training and supervision combined with experience.

Many of the programs found to be effective by the evaluation results in Lira's study would fare poorly against the canon of quality, a difference that can be attributed at least in part to different definitions of quality. Programs often do not stand up to a demanding ideal. But if we compare the environments offered by most early education programs-the physical conditions and instructional methods and content-with the alternative conditions in which children learn and develop, using the criteria in box 4.1, the comparison will usually favor the program environment, with all its imperfections. This is particularly true for some outcomes (such as the production of language) and for children living in conditions of intense poverty. Imagine, at the extreme, an environment in which parents struggling to survive must tie a two-year-old in a crib for several hours a day while they go out to work. A comparison of that environment with the environment of almost any child care program is likely to favor the program, imperfect though it might be in its physical location, training of caregivers, and ratio of caregivers to children.

Thus as programs expand, the challenge is not only to attain or maintain quality, but also to choose an appropriate standard of quality against which to evaluate them. In choosing that standard, it is well to remember that different cultures set different criteria and standards for quality, reflecting their conditions and their values, customs, and beliefs. Also important to recognize is that the standards for judging quality may be different at different times. Understanding the relative nature of quality helps program developers understand that the best can be the enemy of the good, particularly if the best is very costly and limits participation to those who can afford to pay or to a very few of the many who could benefit from a lesser but still adequate program.

This view may cause problems when good rather than best solutions to the challenge of achieving quality are set beside the challenge of achieving equity in programs' content and results, particularly if costly private systems provide the best and subsidized public systems the good. Many of the nonformal early education alternatives aimed at reaching the unreached in Latin America are criticized precisely because they are thought to be of low quality relative to private or even formal public programs. This criticism is often based on the fact that teachers are not certified and do not know how to apply the approved preschool curriculum, or on the failure of facilities to meet governments' minimum standards for physical conditions. This criticism has its place, but it can easily be exaggerated. Many alternative programs are more integral, more active, more participatory, and more relevant to local needs than the formal preschool programs with which they are compared. Again, the issue is differences in the standard chosen.

What we need to know is whether or not the outcomes of alternative preschool programs, whatever the standards chosen, are on a par with those of the formal and conventional programs provided to similar children. This information is rarely available, however.

Summing up

From the above, it is reasonable to conclude that:

• When developing an ECCD program, it is important to involve stakeholders, including potential beneficiaries, in defining quality.

• When judging quality, it is wise to go beyond a definition based on inputs or processes to look at effects. It is wise, too, to compare program environments with alternative learning environments, not just ideal standards.

• Ideal standards should be sought and should be used as a basis for providing incentives to improve. But the best can be the enemy of the good, particularly in the early stages of a program.

• Formal ECCD programs may or may not be judged better than nonformal alternatives when the standard is results and when all dimensions of quality-including relevance, participation, and active learning-are taken into account. But the evidence is limited, and the jury is still out.

• We should not assume that a preschool program is of lower than average quality because it is inexpensive, because it is nonformal, or because it is carried out by people who have experience and training but no professional certification.