Policies for early childhood education and assessment

Abstracts

This paper aims to present and discuss tensions in the contemporary debate about the assessment in/of Brazilian early childhood education, based on the distinction between evaluation policy in/of early childhood education and policy evaluation of early childhood education. The text argues that the process of building the social problem "assessment" is beginning in the arena of negotiations in early childhood education policy, which gives rise to clashes concerning bipolar positions in the assessment models proposed and criticized. It emphasizes the importance of ethical considerations and of explicitation of political positions in evaluative research in early childhood education and warns against the danger of transferring to early childhood education hegemonic models adopted in Brazil at other stages of education.

curse evaluation; early childhood education; publics policies; ethics

O artigo tem por objetivo apresentar e discutir tensões no debate contemporâneo sobre avaliação na/da educação infantil brasileira, a partir da distinção entre política de avaliação na/da educação infantil e avaliação da política de educação infantil. O texto argumenta que se está iniciando o processo de construção do problema social "avaliação" na arena de negociações da política de educação infantil, daí embates em torno de posições bipolares nos modelos de avaliação propugnados e criticados. Ressalta a importância de cuidados éticos e da explicitação de posições políticas nas pesquisas avaliativas em educação infantil e alerta para o perigo de transpor para a educação infantil modelos hegemônicos adotados no Brasil para outras etapas de ensino.

avaliação da educação; educação infantil; políticas públicas; ética

El artículo tiene el propósito de presentar y discutir tensiones en el debate contemporáneo sobre evaluación en la/de la educación infantil brasileña, a partir de la distinción entre política de evaluación en la/de la educación infantil y evaluación de la política de educación infantil. El texto argumenta que se está iniciando el proceso de construcción del problema social "evaluación" en la arena de negociaciones de la política de educación infantil, y ello ocasiona embates en torno a posiciones bipolares en los modelos de evaluación propugnados y criticados. Subraya la importancia de cuidados éticos y de la explicitación de posiciones políticas en las investigaciones evaluativas en educación infantil y alerta sobre el peligro de transponer hacia la educación infantil modelos hegemónicos adoptados en Brasil para otras etapas de enseñanza.

evaluación de la educación; educación preescolar; políticas públicas; etica

ISSUE IN FOCUS

TEACHER'S WORK AND TEACHER'S FORMATION

Policies for early childhood education and assessment

Fúlvia Rosemberg

Researcher Consultant at Fundação Carlos Chagas; Titular Professor at Pontifícia Universidade Católica de São Paulo – PUC-SP; Coordinator of Núcleo Gênero, Raça e Idade - Negri. [e-mail]

ABSTRACT

This paper aims to present and discuss tensions in the contemporary debate about assessment in/of Brazilian early childhood education, based on the distinction between evaluation policy in/of early childhood education and policy evaluation of early childhood education. The text argues that the process of building the social problem "assessment" is beginning in the arena of negotiations in early childhood education policy, which gives rise to clashes concerning dichotomies positions in the assessment models proposed and criticized. It emphasizes the importance of ethical considerations and of making explicit the political positions in evaluative research in early childhood education and warns against the danger of transferring these to the hegemonic early childhood education models adopted in Brazil at other stages of education.

Keywords: assessment; early childhood education; public policies; ethics

In recent years we have witnessed a murmuring in early childhood education around the theme/term evaluation, as if up until then the field had been adverse or foreign to the practice and process of evaluation.

In fact, research into the availability of bibliographical references in the area of education that use the descriptor "evaluation" for early childhood education uncovers only a few. For example, Senhorinha de Jesus Pit Paz (2005) found only three articles that associate the descriptors early childhood education and evaluation from among the 137 communications that referred to evaluation presented at the National Association for Graduate Education (ANPED) annual meetings from 1993 to 2003.¹ Complementarily, in the 53 issues that make up the collection of the journal Studies in Educational Evaluation, it was possible to locate only five articles on the "subject" early childhood development.²

This does not mean, however, that the theme of evaluation in early childhood education has failed to mobilize education managers, researchers and activists, but rather that this concern still has not defined a "social problem" that can be included in the agenda on policy evaluation in/of early childhood education. My argument is that we are beginning to build this agenda and thus the need to distinguish between evaluation policy in/of early childhood education and evaluation of early childhood education policy.

The proposed distinction between these two expressions – evaluation of early childhood education policy and evaluation policy for early childhood education – is not to seek a play on words to introduce this article. Rather, it is an analytical perspective to point out the route already taken, the present moment with all its tensions, and to suggest some prospects for the future.

EARLY CHILDHOOD EDUCATION POLICY

The central argument of this topic is that we are undergoing a process of formalization of evaluation policy, even in advance of achieving clarity about whether it is of or in early childhood education. That is, the term/theme evaluation is entering the field of early childhood education defining a new "social problem" for its policy, since early education has not until now been included in the production on evaluation of basic education. In taking on the status of a social problem, evaluation of/in early childhood education appeals for public attention as an issue of social policy. Thus, the theme becomes defined, set up as a problem, enters the arena of social policy negotiation, seeking public visibility, legitimacy, resources and attracts defenders/supporter (stakeholders), as well as opponents.

In a certain sense, this is more a quest by early childhood education, a minor field, to be included in the hegemonic field of basic education. In fact, as has been observed in other areas and moments, the field of research, policy and practices in basic education evaluation practically banned early childhood education from its manifest concerns, although one of the most referenced works in the contemporary Brazilian bibliography on evaluation, that of Jussara Hoffman (1996), treated the theme at the preschool level. The lack of attention to evaluation production has already been noted by Barretto et al. (2001, p. 33) in the exhaustive review "Evaluation of Basic Education in the 1990s in Academic Journals", when they report that, among the few articles that focus on only one educational level, "Those that focus on evaluation of early childhood education" are very rare.

This near silence imposed on early childhood education, observed in the concerns with educational evaluation can also be identified in various other themes, mainly when early childhood education means not just preschool, but also, and above all, daycare. It is enough to recall the resistance to including daycare as a sub-stage in basic education during the debates in the Constituent Assemble (ROSEMBERG, 2008), in the first formulation of the Law for Guidelines and Bases in Education (LDB) (BRASIL, 1996), in the early versions of the Law for Maintenance and Development of Basic Education and the Advancement of Teachers (Fundeb). Added to that, the treatment of early childhood education, especially daycare, as a legitimate sub-stage in basic education has only recently been included on the agenda of the important social movements, such as the black, rural (educational branch) and indigenous movements (ROSEMBERG; ARTES, 2012).

A two-fold movement can thus be seen at the present time: that of incorporating early childhood education in evaluation policy for basic education, perhaps at a slower rate; and another, more intense movement that seeks to incorporate evaluation as a theme/problem evoking specific attention for early childhood education policy.

This does not mean that early childhood education has not processed evaluations until now. As was said, the argument here is that, for a long time early childhood education has been evaluated and has dealt with evaluation, but not in the manner of a cut, named and framed as a social problem demarking a field of knowledge and political action, and consequently, as a disputed territory. That is, we observe here, as Shadish and collaborators report (1995) about the constitution of the professional field of evaluation research in the United States, that many articles on evaluation were not always labeled as such. In Brazilian early childhood education, we would have numerous examples to evoke, however, in the sequence of this article, we will mention only two those related to evaluation of quality.

It is possible to say that since the first contemporary public demonstrations in favor of daycare – by militants and academics beginning in the 1970s, governmental sometime later – occurred in certain sectors of the country, there has been an intense mobilization for expanding the offer and improving its quality based on evaluations that were not always called that. Wasn't the procedure adopted by the Special Inquiry Commission on Daycare in the Municipality of São Paulo (CEI de Creches, 1983-1984), along the lines professed in the literature on emancipatory evaluation (Saul, 1988), when it heard from different social actors working in daycare or those involved in the struggle for daycare in the city a "participatory evaluation" of the daycare system in the city of São Paulo? To respond to this question, we reproduce the statement of Maria da Pureza, a cook in a city-run daycare center (creche direta) at the time, who in 1983-1984 during one of the public sessions assessed the feeding services provided by the City of São Paulo:

They told us they didn't provide wheat flour because the cake was ready-made [...] the cabbage comes raw. You will use everything from the roots to the leaves. Clean, a one and a half -kilo cabbage is reduced by half, and isn't enough for everyone. And we are in the kitchen performing that Brazilian miracle (CEI/Dossiê I, p. 44, 47, cited in SCAVONE, 2011, p. 4)³

Another example: in making a detailed re-reading of the now historic article by Campos, Füllgraf and Wiggers (2006), "The Quality of Brazilian Early Childhood Education: some results of the study", I learned that, despite its focus on reviewing the literature on "quality and its evaluation", it does not include the descriptor "evaluation" and used the term sparely (only four times in the entire text); equally rare were those from the list of all 50 studies - only two adopted the term evaluation. Thus, in the article, the authors used various expressions to replace the term "evaluation": for example, "diagnosis" ("comparative diagnosis", "diagnosis on care", system diagnosis", etc.), a term historically enshrined in Brazilian early childhood education, the follow expressions that use the term "reality"(the reality studied", "reality that emerges", "reality described", reality studied", etc.), "survey" or "research", all as substitutes for, or alternatives to, the term evaluation".⁴ Can there be any doubt that the study referred to in the excerpt transcribed below made a quality evaluation of the community daycare centers in Fortaleza?

Cruz's (2001) survey of community daycare centers in Fortaleza, Ceará [...] contains information based on observations done at 19 daycare centers that serve 950 children. The daycare centers are badly equipped, and have safety problems. The children are cared for on a full time basis with activities that prioritize feeding [...] There remain long periods of inactivity... (CAMPOS; FÜLLGRAF; WIGGERS, 2006, p. 24)

It is necessary, however, to sound a warning that the inclusion of these old, recurrent concerns in the area of early childhood education is only possible if one adopts a broad framework open to the term/concept – "the process of determining merit, the quality or value of things" (SCRIVEN, 1991, p. 1) – extrapolating restricted concepts focused exclusively on the cost-benefit equation, with an emphasis on results, associated to the "management reform of the government" and to the "letting go of responsibility/turning over/privatization of the provision of goods and social services", and which has provoked reactions "that verge on phobia" from academics and activists (FARIA, 2005, p. 99).

On the other hand, at the present time, when the field has begun to propose early childhood education evaluation policy, there has been a confrontation between more limited frameworks, disputed by the social actors in play, that is, by the various governmental instances, the educational sectors, by the social, and workers' movements, and (not often) users of daycare and preschool. In debating whether an evaluation policy in/of early childhood education would be important, then diverse or divergent models and conceptions arise, which tend to provoke the cited murmurings.

In sum, one can propose an interpretation of the present state of the debate (sometimes a collision) around the issue of evaluation in the field of early childhood education based on studies of the construction of a social problem, especially the study by Bernard Lahire, L'invention de l'"illettrisme" [the Invention of Illiteracy] (1999). Adopting a nominalist concept of the social problem, Lahire and other researchers (for example, BEST; LOWNEY, 2008) point to the importance of the name and of framing in constructing a problem to be included on the public policy agenda at a given moment. A "new" problem is not defined without a new denomination to guide the conceptual political focus that redefines territories. The "new" social problem needs to be "labeled" to raise visibility and thus, compete for a scarce good: specific public attention from among the endless issues, needs, and social problems that compete for inclusion on the public policy agenda.⁵

So it appears to me that including evaluation as an issue for early childhood education, we include in an anachronistic mode what has been discussed and changed in the realm of basic and higher education in Brazil (and in the more general realm of public policy as well) in the last 40 years, at least. Even though not made clearly explicit in the recent debates between defenders and detractors of evaluation in early childhood education, and sometimes inspired in a creed, it is possible to reencounter the old well-known dichotomies of research (Figure 1).

This dualism apprehended in the long-ago defense and attacks on evaluation in/of early childhood education has been problematized, and even overcome, in the contemporary debates over evaluation. It is enough to retake as testimony the already classic Fourth Generation Evaluation by Egon Guba and Yvonna Lincoln (translated to Portuguese by Unicamp Editora in 2011), Foundations of Program Evaluation: Theories of practice, by Shadish, Cook and Leviton (1995), and the Brazilian Fundamentos de um programa de avaliação educacional [Fundamentals of a Program for Educational Evaluation], by Heraldo Marelim Vianna (2005).⁶

In a general way, detractors of evaluation of/in early childhood education tend to conceive it as if it were restricted to a theoretical model considered positivist or of Guba and Lincoln's (2011) first generation, and that it prioritizes the first term in each of the dichotomies: evaluation of product (especially classroom acquisitions by students on the level of knowledge), quantitative, led by technical know-how, however conceiving it as neutral (above good and evil) and objective, valuing especially procedure over theory, ethics and the political dimension.

However, in the contemporary debate over a policy (or system) for evaluation of/in early childhood education, it is not just the detractors of evaluation that tend to sharpen this ancient dichotomy, but defenders of evaluation sometimes support procedures inspired in hegemonic models adopted in Brazil for primary and mid-level education – "large-scale evaluations that take as their main quality indicator the cognitive performance of students as measured by tests" (BRASIL, 2012, p. 6) – which are included under the Brazilian Basic School Assessment System – SAEB.

In order to fill this gap and guided by a specific concept of early childhood education evaluation in consonance with the guidelines of the LDB – which disallow evaluations of children in early childhood education programs for purposes of classification and to hinder their school progress – the Secretariat for Basic Education – SEB – of the Ministry of Education promoted the creation of a working group that wrote a document to guide the policy (there called "systematic") for evaluation of early childhood education (BRASIL, 2012). This "systematic" conceives evaluation as a formative process that, in early childhood education should be directed at institutions, programs and policies. Therefore, it is very far from what was called and interpreted as a positivist model.

...the institutional evaluation covers a set of procedures that range from the organization of students' school records (school progress and profile); of school professionals (training, work week, participation in the school's collegial bodies); the state of the infrastructure (conservation and appropriateness of the facilities; conditions for carrying out teaching (fitness and availability of space and schedules); opinions, perceptions, expectations and suggestions from the entire school community, through the records and critical debate over practices, from the point of view of its coverage, intentionality and relevance. (BRASIL, 2012, p. 21)

In analyzing the dichotomies shown in Figure 1, perhaps it would be admissible that the option to focus on product (and not on process), on procedures (and not on explicit theory) or on quantitative (and not qualitative) techniques could be justified. For example, a strict evaluation of access to daycare could be based on the attendance rate (percentage of children of prescribed age who attend daycare over the number of children in the age group), since until now, there is no other way available to establish goals for access to daycare and evaluate whether the goal is being reached. However, it is not conceivable to subscribe to a concept of evaluation of/in early childhood education that is not always measured by values, that does not attend to the totality of the child's person, and that is not ethical and illuminating for the taking of socially just and democratic decisions. The taking of these positions might not even be explicit, but they are not present in the concepts adopted.

In the realm of early childhood education, two central questions constitute the starting point for any proposals for evaluative research: what and whom does evaluative research into early childhood education serve? What is the reason for the existence of no early childhood education policy?

The first question could find promising clues in the broader contemporary debate over the function of evaluation. Based on the description of the functions attributed to evaluation – information, reallocation and legitimization – Faria (2005), cited by Darien (2001) points out that in Latin America, the "evaluation function" was institutionalized later, only in the 1990s, "the perspective of exploiting evaluative research for the success of reforms [managerial] having prevailed [...] with an emphasis on results and the letting go of responsibility/turning over/privatization of social goods and services".

In early childhood education, the exploitation of evaluation can be gleaned from the texts and reports that define incomplete models of education for children up to 3 years of age, generally called child development, and that redound in the reduction of resources allocated by government, an association especially noted in studies of the impact of evaluation.

I observe a kind of slippage in the use of studies [on evaluation] on the impact of early childhood education in basic education. If at first they were done to assess ongoing evaluations (such as Head Start and High Scope in the United States) and their results functioned to argue for an expansion of sources of financing (it is worth investing in early childhood education), at present impact evaluations are being used to define objectives and strategies: improve the indicators in basic education and reduce the cost of programs and projects to a minimum. (ROSEMBERG, 2001, p. 23)

In the Brazilian context, after the Constitution of 1988 and responding to the second question – about the reason for the existence of early childhood education policy – there are two points of consensus, despite their being in tension: the right of children under 5 years of age to education in daycare centers and preschools, and right of working mothers and fathers to have their children cared for in daycare or preschool. Unfortunately, in Brazil we did not have sufficient political mobilization to integrate the two perspectives. Thus some of the tensions we face will be more intense than those faced by countries where this policy integration did occur. For example, some European countries, especially the Scandinavian countries, managed to integrate their policy on early childhood education with a policy of equal opportunity for women and men at the level of work and family life. This is not to say that the integration is complete, that it is immune to adverse political pressures in times of economic crisis, or that it means that their children's rights will always coincide with the right of parents, of the adults (EURYDICE/EACEA, 2009).

In Brazil, the tension arising from the lack of integration between the rights of children to education and the rights of parents to work manifest recurrently, framing policy decisions and subsequent evaluations: to what standard should vacations from early childhood education comply, that of the school system or labor legislation? Should there be nighttime care for children of those who work at night? A period of adaptation for daycare children backed up by labor rights for the mother or the father?⁷

In focusing on these issues, especially the latter, we perceive that if the agenda for early childhood education policies can approximate the evaluation policy for basic education; it cannot, however, be identical to it. In this field of political action, of social practices and knowledge we are dealing with several particularities that generate questions that might or might not be included in the overall policy for basic education evaluation. Or to put it another way, in including early childhood education, the evaluation policy (or system) for basic education must fit the specificities of this stage of education, as well as the children for whom it is destined.

One of the particularities of early childhood education as compared to the later sub-stages of basic education comes from the greater structural vulnerability – not intrinsic – of small children compared to older people who have greater autonomy and social visibility. This structural vulnerability which has been problematized by other authors in the field of social studies of childhood (ROSEMBERG; MARIANO, 2010; SOARES, 1997), meaning lesser visibility, autonomy and power in political negotiation, puts us on high alert in observance of ethical standards.

ETHICS AND EVALUATION

We know that one of the strategies most frequently employed in evaluating the stages and levels of teaching after early childhood education consists in gauging the value added of improved cognitive performance of students – a performance gauged by means of test results. The recent controversy over the use of tests and developmental scales as a strategy to evaluate the performance of young children attending daycare seems to reveal, among other things, an ethical concern over their use, in addition, evidently, to criticisms of the weakness in the precision and the cultural bias of these instruments.⁸

The ethical issue refers to eventual risks of labeling, stigmatization, or the creation of self-fulfilling prophecies for the children evaluated. Those who follow the research of Marília Pinto de Carvalho (2004) on evaluations done by teachers of their students' performance from the point of view of identification of color/race and sex, will be able to perceive the delicacy of the issue. It is enough to observe the passage of color/race to the status of a "variable dependent": teachers' ethnic/racial identification of students seems to depend also on sex, which seems to also depend on their school performance. Carvalho's (2004) hypothesis is that, at least in the school setting, "racial identity of girls and boys is constructed using as reference not just phenotypical features and socio-economic status, but also their school performance" (p. 247).

When transposing these concerns to the level of evaluation of children/infants attending daycare by means of classificatory scales, some issues should be raised: how does family composition, its level of income, the location of the household in richer or poorer areas of the city and ethnic/racial belonging, orient the eye of the "evaluators" and stigmatize the educational trajectory of these children?

The ethical (and technical) concern with using development "tests" appears not just among researchers, managers and activists in favor of qualitative, soft procedures, but also among the economist authors affiliated to the World Bank who are advocates for quality evaluation using the value added of improved student performance in other stages of life. The following citation is long, but pertinent, and was excerpted from the report "Early childhood education: programs for the most important generation of Brazil":

Data have demonstrated the importance of quality in early childhood education for results in early childhood development, but measuring quality for young children is complex [...] At higher levels of teaching, quality is often measured using indicators of value added by student improvement. Nevertheless, although there are many tools available to measure child development at an early age, they are less precise than those for older children and few systems test children universally. Moreover, linking these measures to incentives for daycare centers and preschools can have a perverse effect in leading these centers to exclude the records of children who are behind in development. As a result, the quality of daycare and preschool is usually measured by multidimensional instruments of observation, in which interviewers observe the daycare centers or preschool in action, classifying their quality in a series of areas. (EVANS; KOSEC, 2011, p. 15)⁹

Two other concerns impel the inclusion of ethics on the agenda of evaluation policy of/in early childhood education: the confidentiality of the information and the daycare center as a "cornucopia of subjects" for research in diverse areas of knowledge.

As we know, in Brazil ethical precepts in research involving humans are mainly guided by Resolution 196/96 of the National Council on Health – CNS – Guidelines and Regulatory Norms for Research Involving Human Beings (the 2012 version is the most recent). The resolution is based on four principles related to research subjects: autonomy, beneficence and non-malfeasance, justice and equity. In general terms, the resolution states that "research involving human beings must always treat them with dignity, respect their autonomy and defend them in their vulnerability" making a commitment "to maximum benefit and minimum harm and risk, assuring that predictable harm be avoided" and to the "social relevance of the research with significant advantages for its subjects and minimization of the onus for vulnerable subjects, which guarantees equal consideration of the interests involved, without losing the sense of its social-humanitarian destination".

One of the translations of these principles to evaluation studies is respect for confidentiality, for the privacy of the information provided, that it cannot be divulged without the authorization by subjects or their guardians in certain instances, such as in the case of children. In analyzing the equivalent precept in the context of the North American debate on evaluation Guba and Lincoln (2011, p. 137) point out:

When social science proposed to understand more and more secret spheres of human behavior, it also entered the sphere that Bok describes as intensely personal. Therefore, it becomes sufficiently invasive, to the point of compromising the regulation of privacy, at least in some studies.

The issue we face here is that of subjective consideration of the differences, sometimes in tension, between the levels of intrusion presumed by the researcher/evaluator and those perceived by the person providing the information. A significant example of this mismatch can provide information on the importance of the color/race question on the children's/adolescents' enrollment forms for basic education.

The item declaring race/color introduced in the 2005 School Census in response to a long-standing demand of the black movement¹⁰ is aimed at assessing the differentials in schooling between white and black students. From the time of its introduction until the most recent School Census (2012), a high rate of non-response has been observed, around 25% (ROSEMBERG; ARTES, 2012), although the rate is practically null on the surveys carried out by the Brazilian Institute of Geography and Statistics – IBGE: the Demographic Census and the National Household Sample Survey – PNAD – or the recent study Ethnic-Racial Characteristics of the Population (IBGE, 2011). One hypothesis is that the confidentiality/privacy of informants' responses to the IBGE items is guaranteed by law, contrary to what happens with the procedures used by the schools. A recent study by Cristiane Irinéa Silva (2011) captures the explicit discomfort of a parent who was enrolling his child in a Brazilian school:

A father came to enroll his child and had a lot of difficulty understanding how to fill out the enrollment form; when it asked about the color or race of the child, he said he couldn't answer, for him "Everyone is beautiful, equal, and a child of God", he added that he didn't know the color or race of the child and that only his wife would know that. The employee insisted, reading out each category slowly and clearly, and he restated what he had said earlier. This was the only child who didn't declare color/race on the enrollment form. (Field notebook, Nov. i, 2006, cited in SILVA, 2011, p. 136)

That is, the level of tolerance of public employees, managers, activists, and researchers with respect to what is considered the privacy of fathers/mothers of students can diverge, and it is especially necessary to discern with care the social vulnerability of Brazilian citizen users (dependents?) of public services, especially those that could be called "captives", such as the case of those who use daycare services whose offer is much lower than the demand.

This caution takes on greater importance when attentive to the results of the important IBGE study Ethnic-racial Characteristics of the Population (2011), carried out in 2008 in the Federal District and five other states.¹¹ Among the many questions, two are particularly relevant to this argument: whether a person knows how to report his/her color/race and the influence of color or race on their lives. Interviewing only residents 15 years old and up, the responses to the questionnaire indicate a high percentage (96.0%) of people who know how to declare their own color or race and a significant percentage, over 60%, of those self-declared blacks, browns, negroes, indigenous and yellows who consider that declaring color or race influences one's schooling (IBGE, 2011, Tables 2.6 and 2.30).

Thus the importance of paying attention to the contexts in which the investigations aiming at evaluation are carried out: they are not just those managed by the state/government, but also those under inspiration of the social movements, such as the case of including the color/race item on the student enrollment form for the Brazilian education system. To the benefits of knowing ethnic-racial belonging of students in order to democratize education, one can juxtapose the negativity of discomfort, and the occasional apprehension of the stigma on the part of parents.

As we have seen, if on the one hand, the term evaluation associated to early childhood education has made a late and sparse entry into the field of education; on the other, daycare centers and preschools have been the locus for evaluations in other fields of knowledge, such as psychology, medicine (pediatrics), nursing, food science, and social service. For example, a quick search of the Scielo Brasil data located 58 titles of articles that have the descriptors "evaluation: daycare".¹² Some examples: "ingestion of nutrients by children at a philanthropic daycare center"; "evaluation of food consumption of children belonging to a philanthropic daycare center"; "evaluation of the nutritional state of psychomotor development in children who attend daycare"; "evaluation of global motor performance and axial and appendicular motor abilities in nursing children who attend daycare"; "children's abilities"; "evaluation of the child development benchmarks"; "evaluation of global motor development".

These examples are already sufficient to point out that Brazilian daycare centers, especially public daycare centers, have been "offering" researchers from various fields of knowledge, especially from the field of health and related disciplines, opportunities for research, on themes, problems, but especially of subjects: children, infants, and sometimes their families and workers.

Renata Ishida (2013), analyzing theses and dissertations from the Coordination for Personnel in Higher Education (Capes) that uses the descriptor daycare, found the evaluation descriptor associated to 16 titles in diverse areas of health (but in only four for education). Flávio Urra (2011) found an important number of articles in Brazilian pediatrics journals where research is reported that evaluates various dimensions of the children who attend Brazilian public daycare centers. Urra (2011) points out two important aspects that deserve attention: lack of care in the description of the reactions of children and their families when using the devices adopted for the research, even when they employ invasive procedures; public daycare is evaluated, explicitly or implicitly, as a place of risk for children, "mobilizing feeling of heaviness and negativity" (p. 1), however, the information broadcast on precise strategies to correct the problems observed at the establishments studied are evasive and generalist.

Both aspects elicit inquietude with respect to ethical standards in evaluation research carried out in public daycare: on the one hand, the need for a clear explanation of the ethics adopted, of the description of eventual reactions when faced with the invasive stimulus and the responses of the researchers; on the other, the clear weighing of the maleficence or beneficence of the research, not just for children and their families, but also for the daycare systems and policy. In reading the articles analyzed by Urra (2011), it is hard to know up to which point these studies are concerned with providing a return, with feeding information back to the daycare centers studied in order to improve their quality. In this case, one can see that the transition from evaluation in early childhood education to evaluation of early childhood education, or better said, daycare, which is being judged as an institution in perennial crisis, could strengthen the social stigma.

From these examples it is possible to point out two "tabs" for this reflection. The first is the need for an evaluation policy (system) in/of early childhood education to consider the ethical dimensions, when the institutions (especially public institutions) are used as the locus or subject of evaluations for other disciplines outside education: what regulations do we have in Brazil for this situation other than Resolution 196/96/2012 of the National Council on Health? The second is the need to guard over the dissemination of the evaluation results.

The evaluation results must not be used solely and exclusively to translate a certain school performance. Their use implies being used in a positive way in defining new public policies, projects to implement and modify curricula, continuing education programs for teachers, in a decisive manner, in defining the elements for decision making in order to have an impact, i.e., changes in the thinking and action of the members of the system. (VIANNA, 2005, p. 17 italics in the original)

That is, a policy for evaluation in/of early childhood education gains a new meaning and appeal for greater care when including also (and being framed as) evaluation of the early childhood education policy.

EVALUATION OF EARLY CHILDHOOD EDUCATION POLICY

There was a time when it was considered that research in evaluation could dispense with taking a position, with values, that it was objective, technical and distanced from politics. This position has long been questioned on the national and international scenario (HOUSE, 1980; FIGUEIREDO; FIGUEIREDO, 1986; FARIA, 2005). Perhaps this position was one of those responsible for the ill will or phobia observed (FARIA, 2005) among Brazilian researchers when faced with evaluation, even in education, beyond the justifying function mentioned earlier.

The assertion that evaluation is a political activity was emphatically stated by House (1980, p. 121):

[...] evaluation is by nature a political activity. It serves the decision makers, results in allocation and reallocation of resources and legitimizes who gets what. It is intimately involved in the distribution of social goods. It is more than a reaffirmation of ideas, it is a mechanism for social distribution [...] Evaluation must not only be true, it must be just [...] and justice presumes an important standard by which evaluation must be judged.

In adopting this perspective, evaluative research establishes and declares it loyalties in terms of its function, of its objectives, methods, procedures, ethics and dissemination of results. This also implies including in the evaluation the objectives of the policy, program or project under analysis, asking exemplary questions: for whose benefit were these policies, programs and projects in early childhood education proposed? Are the objectives proposed in the projects, programs and goals are in consonance with the consensual and those legally instituted? Thus, an evaluation of the National Education Plan 2011-2020 must start with the initial distinction of expanding the offer of daycare (50%) and preschool (universalization), before evaluating whether the goals were met. Why were these goals established? In response to which needs, interests or conceptions? From which social sectors/actors?.

This perspective recognizes that concepts and interests with respect to early childhood education can diverge and even conflict. For example, the Constitution of 1988 established that children from 0 to 3 years of age have a legitimate right to education and to care provided by a daycare institution included under the educational system. However, the United Nations Fund for Children – Unicef – does not translate this right into its evaluation of Diversity and Equity in Brazil (UNICEF, 2003, p. 52) using data collected and processed by the IBGE.

Daycare – Unicef considers it important that children have good living conditions and to accomplish this, it defends the position that up to 3 years of age children can use the family experience and care by parents. In this document, the analysis of data on early childhood education does not include educational indicators for the 0-3 age group, since despite being recognized by Unicef as a right, daycare is not the only opportunity for education at this stage of life. It is essential that care offered by parents or students responsible for guarding over the development of children up to 3 years of age be valued. (Italics mine)

That is, at the same time that it reaffirms the right, the evaluation procedure denies it by excluding indicators relative to this age range. Here, the political or ideological principle of valuing education of infants in home and family space prevailed over the rights of young children and their fathers and mothers. In this case, the Unicef evaluation carried out using data collected and processed by government agencies can be politically evaluated as unjust by the defenders of the constitutional principles that confer the right to daycare on children up to 3 years of age.

Figueiredo and Figueiredo (1986, p. 108) had already argued for the opportunity for a political evaluation of the policy as:

...analysis and elucidation of the criterion or criteria on which a given policy is based: the reasons that they opted for one over another [...] these reasons have to be relevant, i.e. they have to be referenced in principles whose application would presumably contribute to a desirable amount and distribution of well-being [...].

Here we are getting in to a level of analysis that, although not exclusively, is affected by political philosophy, in which debates have become a complex thicket in recent decades with the contemporary emphasis of social movement on demands for practices that redound in policies of identity valuation. That is why Nancy Fraser (2001, 2007), a feminist philosopher, despite associating the principle of justice of recognition to the principle of distributive justice, takes a stand against the contemporary overestimation of identity claims. Fraser points out that in these decades when there was sharp economic inequality; demands for recognition were superimposed over demands for redistribution of social goods. Parodying the famous Hartmann article, "The Unhappy Marriage between Marxism and Feminism" from 1981, Fraser (2007) develops her argument as "The Unhappy Marriage of Culturalism and Neoliberalism" when recognition was raised to the status of a central demand of feminism:

...a venerable category of Hegelian philosophy resuscitated by political scientists, this notion captured the distinctive nature of the post-socialist struggles, which frequently took the form of identity politics, aimed more at a higher valuation of difference than the promotion of equality. (p. 296)

This situation can be pointed to in contemporary Brazilian early childhood education. Based on two examples:

The first refers to early childhood education in the context of race relations. Here, the perspective of evaluating distributive justice or injustice of social policy has been assessed using an indicator called "race bias or discrepancy", which means gauging how close or near the social indicators (employment, education, basic sanitation, etc.) are with respect to blacks and whites. This has been an indicator widely used to evaluate access of blacks and indigenous to higher education. This indicator has also been used to establish quotas to reserve places with the idea of reaching goals: generally equivalent to the percentage of these groups in each of the regions defined. It is not my purpose here to discuss the merit (greater) or lack of merit in this perspective of evaluating indicators of access to higher education and for the correction of inequalities. The issue is the transposition of their use to measure access to daycare, an educational stage that shows a lower rate of attendance for children from 0 to 3, blacks as well as whites. Here what is most decisive would not be the discrepancy, the difference between blacks and whites, but the inequality that affects both. As can be seen in Figure 2, in the percentage curves for black and white children attending daycare which overlap, indicating near equality in their destiny!

The other example comes from analyzing what the 2012 School Census calls "differentiated school location", i.e., the fact of the school being located in an agrarian reform settlement area, a community of quilombo descendants or on indigenous land. These categories correspond to social identity movements which, among other things, claimed and obtained inclusion in the School Census questionnaire of an item about availability of "differentiated teaching materials", i.e., teaching materials that welcome differences. This information was made available in the 2010 School Census for the 4,739 establishments in "differentiated locations". However, they did not include a general item on the availability of teaching materials in general for the 108,967 establishments that make up the universe of schools that offer early childhood education (ROSEMBERG; ARTES, 2012).

In this example I highlight two aspects: the political dimension of the decision of whether to include items that unfold into indicators for evaluation of early childhood educaiton policies, and the prioritization of identity indicators over indicators focused on distributive justice.

This raises a warning about proximity and distance between the agenda of the social movements – black, indigenous, MST/rural education, women – and the agenda of the movements around children's rights, and early childhood education with evident repercussions for the political focus of early childhood education evaluation policies.

I think it is unnecessary to demonstrate any further that decisions for formatting models for evaluation of early childhood education policy are always political and have predictable consequences for the instruments selected for such evaluations.

EVALUATION OF ACCESS POLICY: INDICATORS AND DATA

One of the basic instruments for the evaluation of public policies, social indicators are used in all stages of the cycle – from constructing the problem, through the agenda, to the evaluation of results – as well as for the diverse social sectors: health (for example, the infant mortality rate), labor (the rate of female activity), income distribution (GINI index), education (rates of daycare attendance), among others. Created in the United States at the end of the 1960s as a tool for planning and evaluation of public policy, the social indicators were named and used in Brazil in 1975, under the auspices of the Brazilian Institute for Geography and Statistics (SANTAGADA, 1993). Let's begin with their conceptualization.

...the social indicators are measures used to make an abstract concept or a demand of pragmatic interest operational. The indicators, point out, indicate, approximate, translate into operational terms the social dimensions of interest defined based on theoretical or political choices made earlier. (JANNUZZI, 2005, p. 138)

Jannuzzi attests to the growing interest in Brazil in the development and dissemination of social indicators that has mobilized governments, political parties, social movements, the media, churches, unions, academics and the business class, i.e., the multiplicity of social actors who involve themselves in implementing or accompanying public policy. He further notes that this interest has stimulated and continues to stimulate the production of extensive better-quality information (such as that provided by demographic and school censuses) and greater public transparency. Moreover, computerized technology has expanded access to information, formerly restricted to the technicians at the institutions that produced the data.

In this context, the national statistical systems are precious sources for developing social indicators to evaluate public policy. Like any aggregated data, the social indicators are worth only as much as the concepts, definitions and procedures adopted to construct and to collect the data.

Good practice in social research recommends that the procedures for constructing indicators be clear and transparent, that the methodological decisions be justified, and that the subjective choices, invariably many, be explained in an objective manner. (JANNUZZI, 2005, p. 141)

The construction of indicators in early childhood education is in its infancy, since this stage of education was a latecomer to inclusion in the main national statistics systems: I could not find a precise date that reports the beginning of the inclusion of kindergarten (or nursery or preschool) in the systematized education statistical systems of Inep. However, we know precisely that the IBGE included daycare and the age group up to 6 years of age in its surveys beginning with the National Household Sample Survey (PNAD) of 1994 and that the first Demographic Census to collect this information took place in 2000.

Therefore, to arrive at a critical reflection about the indicators used to evaluate early childhood education policies, especially the rate of attendance or enrollment in daycare or preschool, and the information used to evaluate access, it is necessary to go back to the concepts and procedures adopted for data collection, which will be done in the following.

Statistical information on education of the Brazilian population including early childhood education is collected mainly by the IBGE or by the Inep.¹³ At the IBGE, the main studies that deal with the population and define the household as units for collection are the demographic censuses and the PNAD.

The questionnaires for the demographic censuses and the PNAD locate daycare among the "programs" that people attend alongside others on the list: preschool, literacy classes, literacy training for youth and adults, basic, middle, higher education, specialization in higher education, master's and doctoral degrees. In the technical notes for the 2010 Demographic Census the concepts transcribed below appear.

Program attended. Programs people attend were classified into:

• Daycare – for a program destined to daytime care for infants in establishments, whether regulated or not.

• Preschool – for programs (nursery school or kindergarten) whose purpose is comprehensive development of the child in their physical, psychological, intellectual and social dimensions, complementing family and community activities.

• Literacy Class – for literacy programs for children.

(IBGE, 2012, s/p, Technical Notes)

In other words, the 2010 Demographic Census incorporated a configuration not legally recognized by the LDB (1996), since the "program" literacy classes was not contemplated in the Brazilian education system.

The question asked by the 2010 Demographic Census was whether the person "attended daycare or school".

This considered those who attended daycare to be children enrolled and attending establishments, whether or not legally regulated, intended to provide daycare for infants.

Those who attended school, i.e., students, were those who were enrolled or attended programs: preschool (nursery school or kindergarten); literacy classes – CA [...]. (IBGE, 2012, s/p, Technical Notes)

On the other hand, the Inep, responsible for organizing, consolidating and disseminating the School Census adopted another concept, as we shall see. First, however, it is necessary to reiterate the centrality of the information systematized by Inep, beginning with Law 9.424/96 the Fund for Maintenance and Development of Elementary School and the Advancement of Teachers – Fundef – changed the calculation of the amount of resources allocated to basic public education in the states and federal district, and it became associated to the number of enrollments shown by the School Census. This system was improved after approval by the Fund for the Maintenance and Development of Basic Education – Fundeb – in 2006.¹⁴

The instrument for collecting data for the School Census is a form required to be filled out by the teaching establishment that appear on the Inep list. The form is signed by the person responsible for filling it out (principal or secretary) and verification of the information is expected by means of in loco research (BRASIL, 2005, p. 3).

Initiated in 2005 and implemented in 2007, the records for students and teachers were an advance (formerly the information was on enrollment and teaching functions) and was included in the School Censuses, whether in an aggregated version disseminated on the site of the Ministry of Education (2007 and 2008), or just in the micro data version. These records contain variables on the people, in the case of students and teachers, in addition to the information on enrollments and teaching functions. Thus, the information on students because it refers referring to individuals could be equal to information on those who reside and attend daycare or school as surveyed by the Demographic Censuses despite some particularities or divergences from that collected in each study. According to Kappel (2008) the student records were still not well consolidated and still need adjustments.

Notwithstanding the advances, research on the status of the supply of early childhood education in rural areas (ROSEMBERG; ARTES, 2012) pointed to the persistence of important discrepancies in the data for 2010 and disseminated by both institutions: while the Inep reported 8,179,685 enrollments in daycare, preschool and primary school for children from 0 to 6 years of age, the IBGE indicated that 9,969,352 children up to 6 years old were attending daycare, preschool, literacy programs and primary education, also in 2010. The major difference occurred in comparing children up to 3 years old, always in the sense of underreporting in the data from the Inep, as shown in Table 1.

This discrepancy between the data from both sources in not new and neither is its dissemination. In an analysis of the national data available at the end of the 1990s (ROSEMBERG, 1999) in which the information from the 1989 School Census is compared to that from the National Health and Nutrition Study, also in 1989 – the only study of national coverage done to date by the IBGE with data for the entire age range from 0 to 6 years of age attending daycare or preschool – a notable different is already observable, especially between the number of children/enrollments for the 0 to 3 age group.

The underreporting of enrollments verifiable at present is more resistant to interpretation after the implementation of the Fundeb, by which resources are allocated according to the number of enrollments and, according to Inep itself, a child can correspond to more than one enrollment. Several hypotheses have been raised to explain this discrepancy: differences in the collection units (household or school), of the reference (the person or the enrollment), the informant (parents/head of household and school employee), definition of age (age completed at the time of the census/birth certificate/information from parents/guardian), reference date for information (October or May), as well as the construction and updating of the school records by Inep. That is, it can be supposed that the population uses "daycare or schools" that are not included in the Inep records because they are not formally characterized as daycare centers or schools, or that these units are included or excluded from records in successive years.

Given that these discrepancies have been observed for over a decade, perhaps it is the time to carry out specific studies to understand them. What might explain, for example, the big difference with respect to 2010 (24% more enrollments in the IBGE data)? One possible explanation for underreporting the enrollments might be in the non-recording of low-quality schools. If this were true, the present data on the quality of supply collected by the Inep would tend to present a more favorable picture of daily life at daycare centers, preschools and schools. Local studies of records from city schools and of the understanding of the questionnaires and records by managers, employees and parents could provide clues to correcting the information provided, if it were considered relevant, thus achieving greater trustworthiness, overcoming this mismatch of educational statistics.

Despite the legal and formal conceptualization in Brazil of daycare, preschool and elementary school, recent research found a large number of "out of place" children (ROSEMBERG; ARTES, 2012),where either the level of school they attend or at which they were enrolled does not coincide with the prescribed age (Table 2).¹⁵

Of all children from 0 to 3 years old, 797,794 would be out of place according to the 2010 Demographic Census: 41.5% of children under 3 attend daycare and 58.5% attend preschool. Nevertheless, at this age, more than half the children would be "out of place", a situation which is more frequent in rural areas, where 60.5% of children under 3 who attend some teaching establishment were in preschool and not in daycare.

Data collected by the Inep via the 2010 School Census also show a significant important number of enrollments of children from 0 to 6 years old (Table 3). The discrepancy is greater for daycare than for preschool, as well as for enrollments in the rural area.

A complementary analysis relating the incidence of the "out of place" category to the variables that could indicate structural power relationships among the social segments (sex, color/race, region, public or private educational system) reveal a worrisome picture (Table 4).

Among the trends, one notes that except for the sex variable (for which the "out of age" enrollments are very close), the largest percentage of "out of age" is found among children with fewer social opportunities: residents of rural areas, the North and Northeast regions, declared black, indigenous or yellow for color/race. One also observes a higher percentage of discrepancies in enrollments from public establishments.

Two initial comments are possible: the first is that the level-age discrepancy should be reported in the Statistical Synopsis of Basic Education, so that the reader can be clear that the numbers that the tables report for daycare do not correspond to the formal conception. The second refers to the complexity of comprehending the symbolic routes that lead to this hierarchy in constructing the means of daycare as a legitimate educational level, a right of all children from 0 to 3 years of age. To accomplish this, we would need deeper research such as is being carried out in in other contexts.¹⁶ It is hard to know, just in the light of the data presented here, whether the meaning of daycare is still associated to the traditional social service clientele; or whether in early childhood education establishment called and registered as daycare, enrollment of older children take away places from children 0 to 3 years old.

Yet the analysis of the data present allows us to conclude that the information from the 2010 School Census by level of school inflate the number of daycare enrollments when only the explicit legal concept is considered, in the same way that was observed with the data from the 2010 Demographic Census.¹⁷ That is, evaluation of the policy for access to early childhood education, especially to daycare, should pay attention to the age group to which the attendance rates refer. The indicator become particularly inadequate when a relationship is established between the numbers of daycare enrollments given by the School Census and the number of children from 0 to 3 years old according to data from the Demographic Census or the PNAD: many daycare enrollments are from children over 3 years and 11 months old. In sum, for to effectively evaluate the policy of access to Brazilian early childhood education, especially daycare, we still do not have precise, reliable indicators.

An equivalent error derives from the exclusive use of daycare or school attendance rates for comparative purposes, whether between historical periods (for example, fulfillment of the goals of the National Education Plans) or among geographic or territorial units (regions, states, and municipalities, for example). The already cited report by Evans and Kosec points out:

Regional trends in the expansion of early childhood education are not limited to income groups. The Northeast – one of the poorest regions of Brazil – had higher enrollment rates in daycare and preschool than the relatively wealthy Central-west throughout nearly all of the last decade and has some of the highest rates of preschool enrollment. (2011, p. 57)

Now, this conclusion might be cause for rejoicing if the high rate of enrollment or attendance for early childhood education in the Northeast Region were not accompanied by a lower presence of fulltime care. That is, the high rates of early childhood education attendance in the Northeast seem to be explained by the low rate of fulltime care, giving form to the popular saying, "robbing Peter to pay Paul" (Table 5).

In sum, there is still a long way to go in order to have solid macro data to construct reliable indicators of access, given that evaluation is only one of the dimensions of national early childhood education policy.

FINAL SUGGESTIONS

Despite pointing out many tensions, the perspective in the text was not to cause turmoil ruckus, on the contrary, it was to sound a warning and draw the attention of activists and researchers to the cause of early childhood education with equity and quality for the new theme/ problem of evaluation without phobias, ill will or prejudice, but with care. The functions of evaluative research can be multiple, including and above all that of informing social actors more directly involved with the usufruct of daycare centers and preschools.

Alongside the concerns and tensions at the time of defining the "new" thematic field and problematic, exists the expectation of being able to participate, create and influence. At the end of this article I could not avoid this impulse and dared to develop suggestions that will sound ingenuous to some, but to others might seem as obvious as the "Emperor's new clothes". They are:

1. Develop a state-of-the-art for evaluative studies on early childhood education;

2. Strengthen the channels of communication and mutual formation between the fields of basic education and early childhood education;

3. Create more efficient and proactive dissemination and monitoring of the implementation of guidelines, orientations, resolutions and other normative documents related to early childhood education;

4. Plan for the implementation of legal alterations before their discussion and approval (prior analysis of feasibility and the implementation timetable);

5. Elucidate the mismatched information and concepts referring to early childhood education coming from the federal agencies that produce statistics, municipal education departments, teaching establishments, fathers/mothers;

6. Carry out research based on the IBGE and Inep microdata adopting the states, federal district and municipalities as the unit of analysis, aimed at planning and creating monitoring policies;

7. Review the school census questionnaire/form in light of the eventual correspondences among its items and scales for evaluation of quality, aimed at monitoring the quality of the offer;

8. Analyze the location of daycare centers and other schools in the national territory (rural and urban) according to the population density of children and income distribution and plan for the location of new buildings for daycare centers and preschools;

9. Improve the dissemination of data, including indicators on young children especially residents of rural areas;

10. Include evaluation of early childhood education programs, especially those that change constitutional amendments, such as occurred with Constitutional Amendment 59/09;

11. Develop strategies to expand access to daycare and improve its quality;

12 Implement training experiences for managers and legislators related to early childhood education, especially for daycare.

REFERENCES

ÁVILA, Ivany Souza (Coord.) et al. Avaliação da qualidade do atendimento oferecido em creches e pré-escolas no Estado do Rio Grande do Sul: relatório de pesquisa. Porto Alegre: UFRGS, 1994. (Consultora: Maria Isabel Edelwiss Bujes).

BALL, Stephen J.; BOWE, Richard. Subject departments and the "implementation" of National Curriculum Policy: an overview of the issues. Journal of Curriculum Studies, London, v. 24, n. 2, p. 97-115, 1992.

BARBOSA, Maria Carmen Silveira. Avaolhiando: olhando as avaliações infantis: relatório de pesquisa. Porto Alegre: Universidade Federal do Rio Grande do Sul, 1995. (Consultora: Maria Isabel Edelwiss Bujes).

BARRETTO, Elba S. de Sá et al. Avaliação na educação básica nos anos 90 segundo os periódicos acadêmicos. Cadernos de Pesquisa, São Paulo, n. 114, p. 49-88, nov. 2001.

BARRETTO, Elba S. de Sá; PINTO, Regina Pahim (Coord.). Avaliação na educação básica: 1990-1998. Brasília: MEC/Inep, 2001. 219 p.

BEST, Joel; LOWNEY, Kathleen S. Teaching social problems from a constructionist perspective: Manual to Accompany Joel Best's Social Problems. New York: Norton & Company, 2008.

BRASIL. Lei de diretrizes e bases da educação nacional, n. 9.394, de 20 de dezembro de 1996. Disponível em: <http://portal.mec.gov.br>. Acesso em: 2 maio 2013.

BRASIL. Ministério da Educação. Instituto Nacional de Estudos e Pesquisas Educacionais Anísio Teixeira. A qualidade da educação: conceitos e definições. In: REUNIÃO DO FÓRUM HEMISFÉRIO EDUCACIONAL QUALIDADE DA EDUCAÇÃO, 2., maio 2005, Brasília. Artigo apresentado... Disponível em: <http://escoladegestores.mec.gov.br/site/8-biblioteca/pdf/qualidade_da_educacao.pdf>. Acesso em: maio 2013.

______. Censo escolar 2010. Disponível em: <http://portal.inep.gov.br/>. Acesso em: 15 jul. 2012.

______. Censo escolar 2012. Disponível em: <http://portal.inep.gov.br/>. Acesso em: 21 fev. 2013.

BRASIL. Ministério da Educação. Secretaria de Educação Básica. Educação infantil: subsídios para construção de uma sistemática de avaliação. Documento produzido pelo Grupo de Trabalho instituído pela Portaria n. 1.147/2011. Brasília, MEC, 2012.

BRASIL. Resolução n. 196/96, de 10 de outubro de 1996. Dispõe sobre as diretrizes e normas regulamentadoras de pesquisas envolvendo seres humanos. Conselho Nacional de Saúde, Brasília, DF, 10 out. de 1996. Disponível em: <https://conselho.saude.gov.br/docs/Reso196.doc>. Acesso em: mar. 2013.

CAMPOS, Maria Malta; FÜLLGRAF, Jodete; WIGGERS, Verena. A qualidade da educação infantil brasileira: alguns resultados de pesquisa. Cadernos de Pesquisa, v. 36, n. 127, p. 87-128, jan./abr. 2006.

CARVALHO, Marília Pinto de. O fracasso escolar de meninos e meninas: articulações entre gênero e cor/raça. Cadernos Pagú, n. 22, p. 247-290, 2004.

CIASCA, Maria Isabel Filgueiras Lima; MENDES, Débora Lúcia Lima Leite. Estudos de avaliação na educação infantil. Estudos em Avaliação Educacional, v. 20, n. 43, p. 293-304, ago. 2009.

DAMIANI, Magda Floriana et al. Educação infantil e longevidade escolar: dados de um estudo longitudinal. Estudos em Avaliação Educacional, São Paulo, v. 22, n. 50, p. 515-533, dez. 2011.

ELTINK, Caroline Francisca. Indícios utilizados por educadores para avaliar o processo de inserção de bebês em uma creche. Ribeirão Preto: USP/FFCLRP, 2000. p. 93-100. (Seminário de pesquisa, 3).

EURYDICE/EACEA. Agence Exécutive Éducation, Audiovisuel et Culture. L'éducation et l'accueil des jeunes enfants en Europe: réduire les inégalités sociales et culturelles. Bruxelles: Union Europeenne, 2009.

EVANS, David K.; KOSEC, Katrina. Educação infantil: programas para a geração mais importante do Brasil. São Paulo: Fundação Maria Cecília Souto Vidigal, 2011.

FARIA, Carlos Aurélio Pimenta de. A política da avaliação de políticas públicas. Revista Brasileira de Ciências Sociais, São Paulo, v. 20, n. 59, p. 97-110, out. 2005.

FIGUEIREDO, Marcus Faria; FIGUEIREDO, Argelina Maria C. Avaliação política e avaliação de políticas: um quadro de referência teórica. Análise e Conjuntura, Belo Horizonte, v. 1, n. 3, p. 107-127, set./dez. 1986.

FRANCO, Maria Laura P. B. Porque o conflito entre tendências metodológicas não é falso. Cadernos de Pesquisa, São Paulo, n. 66, p. 75-80, ago. 1988.

FRASER, Nancy. New School for Social Research. Mapeando a imaginação feminista: da redistribuição ao reconhecimento e à representação. Estudos Feministas, Florianópolis, v. 15, n. 2, p. 291-308, maio-ago. 2007.

GUBA, Egon G.; LINCOLN, Yvonna S. Avaliação de quarta geração. Campinas: Unicamp, 2011. 320 p.

HARTMAN, Heidi. The unhappy marriage of marxism and feminism: towards a more progressive union. In: SARGENT, Lydia. Women and revolution. Boston: South End, 1981. p. 1-4.

HOFFMANN, Jussara. Avaliação na pré-escola: um olhar sensível e reflexivo sobre a criança. Porto Alegre: Mediação, 1996.

HOUSE, Ernst R. Evaluating With Validity. Beverly Hills: Sage Publications, 1980.

INSTITUTO BRASILEIRO DE GEOGRAFIA E ESTATÍTICA - IBGE. Características étnico-raciais da população: um estudo das categorias de classificação de cor ou raça 2008. 2011.

______. Notas Técnicas. Disponível em: <http:www.ibge.gov.br/home/estatística/população/defauttab_amostra.shtm>. Acesso em: 20 jul. 2012.

______. Censo Demográfico 2010. Disponível em: <http://www.ibge.gov.br/>. Acesso em: 10 jan. 2012.

ISHIDA, Renata. Discursos de agentes comunitárias de saúde paulistanas sobre creche e cuidado de bebês. Dissertação (Mestrado) – Programa de Estudos Pós-Graduados em Psicologia Social, Pontifícia Universidade Católica, São Paulo, 2013.

JANNUZZI, Paulo de Martino. Indicadores para diagnóstico, monitoramento e avaliação de programas sociais no Brasil. Revista do serviço Social Público, Brasília, v. 56, n. 2, p. 137-160, abr./jun. 2005.

KAPPEL, Dolores Bombardelli. Educação infantil nas estatísticas do Censo escolar do INEP. Rio de Janeiro, 2008. (Documento técnico para o MEC)

LAHIRE, Bernard. L'invention de l'"illettrisme": rhétorique publique, éthique et stigmates. Paris: La Découverte, 1999.

MAINARDES, Jefferson. Abordagem do ciclo de políticas: uma contribuição para a análise de políticas educacionais. Educação e Sociedade, Campinas, v. 27, n. 94, p. 47-69, jan./abr. 2006.

MAZZOTTI, Alda Judith Alves. O debate atual sobre os paradigmas de pesquisa em educação. Cadernos de Pesquisa, São Paulo, n. 96, p. 15-23, fev. 1996.

NAZARETH, Leila. Discursos sobre creche na revista Pais e Filhos: análise da ideologia. Tese (Doutorado em Psicologia Social) – Pontifícia Universidade Católica de São Paulo, São Paulo, 2011.

PAZ, Senhorinha de Jesus Pit. A avaliação na Educação Infantil análise da produção acadêmica brasileira presente nas reuniões anuais da ANPEd entre 1993 e 2003. 2005. 132 f. Dissertação (Mestrado) – Universidade Federal de Santa Catarina, Florianópolis, 2005.

RAMOS, Marília Patta; SCHABBACH, Letícia Maria. O estado da arte da avaliação de políticas públicas: conceituação e exemplos de avaliação no Brasil. Rev. Adm. Pública, Rio de Janeiro, v. 46, n. 5, p. 1271-1294, set./out. 2012.

ROCHA, Edmar José da; ROSEMBERG, Fúlvia; Autodeclaração de cor e/ou raça entre escolares paulistanos(as). Cadernos de Pesquisa, São Paulo, v. 37, n. 132, p. 759-799, set./dez. 2007. (Artigo elaborado com base em pesquisa para a dissertação de mestrado de Edmar José da Rocha, 2005)

RODRIGUES, Gisele Sonsini; LARA, Angela Mara de Barros. Avaliação das propostas do Banco Mundial para a educação infantil: influências e consequências nos países periféricos. Est. Aval. Educ., São Paulo, v. 17, n. 33, p. 89-104, abr. 2006.

ROSEMBERG, Fúlvia. O estado dos dados para avaliar políticas de educação infantil. Estudos em Avaliação Educacional, São Paulo, n. 20, p. 5-58, dez. 1999.

______. Crianças e adolescentes na sociedade brasileira e a Constituição de 1988. In: OLIVEN, Ruben G.; RIDENTI, Marcelo; BRANDÃO, Gildo M. (Org.). A Constituição de 1988 na vida brasileira. São Paulo: Hucitec, 2008. p. 296-334.

______. Bebês e creche: discursos e políticas. In: REUNIÃO ANUAL DA ANPED, 35., 2012, Porto das Galinhas, PE. Anais... Porto das Galinhas: ANPED, 2012. GT7: Educação de crianças de 0 a 6 anos.

ROSEMBERG, Fúlvia; ARTES, Amélia. O rural e o urbano na oferta de educação para crianças de até 6 anos. In: BARBOSA, Maria Carmen Silveira et al. (Org.). Oferta e demanda de educação infantil no campo. Porto Alegre: Evangraf, 2012. p. 13-69.

ROSEMBERG, Fúlvia; MARIANO, Carmem Lúcia Sussel. A Convenção Internacional sobre os Direitos da Criança: debates e tensões. Cadernos de Pesquisa, São Paulo, v. 40, n. 141, p. 693-728, set./dez. 2010.

ROSEMBERG, Fúlvia; PINTO, Regina Pahim. Criança pequena e raça na PNAD 87. Textos FCC, n. 13, 1997.

SANTAGADA, Salvatore. Indicadores sociais: Contexto Social e breve histórico. Indicadores Econômicos FEE, Porto Alegre, v. 20, n. 4, p. 245-255, 1993.

SANTOS, C. P. dos. Discursos sobre o bebê e a creche no jornal on line Folha de S. Paulo. Dissertação (Mestrado) – Pontifícia Universidade de São Paulo, 2012.

SAUL, Ana Maria. Avaliação emancipatória: desafio à teoria e à prática de avaliação e reformulação de currículo. São Paulo: Cortez; Campinas: Autores Associados, 1988.

SCAVONE, Darci Terezinha de Luca. Marcas da história da creche na cidade de São Paulo: as lutas no cotidiano (1976-1984). Dissertação (Mestrado) – Universidade São Francisco, Itatiba SP, 2011.

SCRIVEN, Michael. Evaluation Thesaurus. Califórnia: Sage, 1991.

SHADISH, William R.; COOK, Thomas D.; LEVITON, Laura C. Foundations of program evaluation: theories of practice. Califórnia: Sage, 1995.

SILVA, Cristiane Irinéa. Acesso de crianças negras à educação infantil. In: ROCHA, Eloisa Candal; KRAMER, Sonia (Org.). Educação Infantil enfoques em diálogo. Campinas: Papirus, 2011. p. 131-138. (Prática Pedagógica).

SILVA, Juliana Bezzon da; SOUZA, Tatiana Noronha de. Análise da utilização de uma escala para avaliação da qualidade de creches. Estudos em Avaliação Educacional, São Paulo, v. 22, n. 48, p. 137-158, abr. 2011.

SOARES, Natália Fernandes. Direitos da Criança: utopia ou realidade? Infância e juventude, Lisboa, n. 4, p. 101-126, 1997.

FUNDO DAS NAÇÕES UNIDAS PARA A INFÂNCIA - UNICEF. Dados sobre a infância e a adolescência: diversidade e equidade no Brasil. In: ______. Relatório da situação da infância e adolescência brasileiras: diversidade e equidade. Brasília: UNICEF, 2003. Parte 2.

URRA, Flávio. Concepção de creche em revistas brasileiras de pediatria: uma interpretação a partir da ideologia. Dissertação (Mestrado) – Pontifícia Universidade Católica de São Paulo, São Paulo, 2011.

VIANNA, Heraldo Marelim. Fundamentos de um programa de avaliação educacional. Brasília: Liber Livro, 2005.

Received in: JANUARY 2013

Aproved for publication in: MARCH 2013

TRANSLATED BY Ann Puntch

Version reviewed, updated and expanded of a communication of the same title presented at the Seminar Evaluation of Early Childhood Education: trends and perspectives. [I Amélia Artes for developing the tables and Marcia Caxeta for formatting]

1

The work of Avila and collaborators (1994), Barbosa (1995) and Eltink (2000).

2

They were Rosemberg (1999), Rodrigues and Lara (2006), Ciasca and Mendes (2009), Damiani and collaborators (2011), and Silva and Sousa (2011).

3

The citation was excerpted from the master's thesis of Darci Terezinha de Luca Scavone, which in its version of the history of daycare in São Paulo for the period 1976-1984, offers a very special evaluation of the daycare CEI: "an immersion that seemed to want to embrace the issue of daycare in all its dimensions, diluting the focus that motived the investigation - the turning over of the daycare centers constructed by the municipal authorities to private entities by means of contracts" (SCAVONE, 2011, p. 142). Rosemberg (2008) in reviewing the same event points out however, that its conclusions included the first formulation in the country of the proposal for the Constituent Assembly to recognize daycare as extending to young children the universal right to education.

4

According to Ramos and Shabbach (1012. p. 1, 277) the diagnosis, or study of the situation, is a kind of ex ante evaluation of public policies, programs or projects "when the needs are mapped and feasibility studies are done to guide the formulation of the program". In the field of Brazilian early childhood education, the term diagnosis was used more broadly as evaluation of a given situation of offer.

5

In educational literature, the bibliography most often referenced on the policy cycle was authored by Ball and Boswe (1992). See Mainardes (2006).

6

For the younger generations, I still recall the seminal article by Maria Laura P. Franco published in

Cadernos de Pesquisa (FRANCO, 1988) on qualitative and quantitative approaches (controversy with Sérgio Luna), the summary of the war between theoretical and meta-theoretical paradigms so well elaborated by Alda Judity Alves Mazzotti (1996) in summing up one of Guba and Lincoln's books, and the already mentioned article by Ana Maria Saul (1988).

7

This is perhaps one of the greatest forms of discrimination against young children whose fathers/mothers do not have the resources to take leave from their jobs during their children's period of adaptation to daycare/preschool. It should be noted that availability and strategies for the period of adaptation are not included in the criteria for evaluations of the quality of offer, nor on the agenda of the contemporary social movements in Brazil.

8

See the report in the document produced by the MEC/COEDI orking group (BRASIL, 2012) on the controversy over the use of developmental scales as an instrument to evaluate children who attend daycare.

9

Katrina Kosec is a Ph.D in Political Economy from Stanford University. David K. Evans holds a Ph.D in Economics from Harvard University (

www.ifpri.org). Accessed on: Dec.12, 2012.

10

The information should be provided by the students' themselves beginning at age 16, and before this age, by the parents (ROCHA; ROSEMBERG, 2007).

11

Amazonas, Paraíba, São Paulo, Rio Grande do Sul, Mato Grosso.

12

Not all refer to daycare in early childhood education. The term is also used in veterinary medicine. Research done on March 20, 2013.

13

In reality, Inep systematizes information collected by the schools.

14

At present, the information collected on the School Census provides support for several other programs in addition to FUNDEB to allocate resources, such as the National Textbook Program (PNLD), school transportation, and school feeding programs, among others.

15

This information had already been collected by Rosemberg and Pinto (1997) and Rosemberg (1999)

16

Among them Urra (2100) and Nazareth (2100).

17

The other.

Publication Dates

• Publication in this collection
  25 June 2013
• Date of issue
  Apr 2013

History

• Received
  Jan 2013
• Accepted
  Mar 2013