Research on the Field of Education Policy: Exploring Different Levels of Approach and Abstraction

This paper, of theoretical nature, explores the levels of approach and abstraction of research in the field of education policy: description, analysis and understanding. Such categories were developed based on concepts of Bourdieu’s theory and on the grounds of epistemological studies focused on education policy and meta-research. This paper highlights the importance of different kinds of studies on education policy and claims the need to expand the number of papers at understanding level, which can contribute more effectively to the strengthening of education policy as an academic field.


Introduction
This paper is a theoretical exercise that aims to explore issues related to the levels of approach and abstraction used by researchers in the field of education policy when presenting their research reports 2 .The development of reflections on this issue emerged as a necessary task in the context of the research on education policy epistemologies, which we have been developing as part of the Red de Estudios Teóricos y Epistemológicos en Política Educativa (ReLePe) 3 .These studies involve the conduct of theoretical studies and empirical research that aim to understand how the researchers in this field have been working with the issues of an epistemological nature 4 .
In this paper, we initially present a basic theoretical framework on the focus of education policy, epistemological studies and meta-research, and then we present the three levels of approach and abstraction of research in this field: description, analysis and understanding.
We understand 'approach' as the researcher's decisions regarding the development of research and report elaboration.Abstraction levels are related to the results of the investigation as a whole.Epistemological positions assumed by the researchers are closely related to methodological approaches and influence throughout the research process, they assist in the development of research questions, the research design and the definition of data collection procedures.
The field of education policy is comprehensive and inclusive.Researchers use a diversity of theoretical perspectives in the development of different types of research: theoretical research, empirical research, comments or critiques.Research reports have different purposes and formats.Thus, identification of levels of approach and abstraction demand the analysis of research reports of education policy as a whole, through systematic reading, seeking to understand the context of the research, the theoretical framework, the objectives of the investigation, the purpose of the text, the researcher insertion level in the specific field of the education policy, among other aspects.

The Focus of the Education Policy Epistemologies and Meta-Research
Based on Bourdieu, Tello (2012) considers that the focus of the epistemologies of education policy approach (EEPA) is an analytical and conceptual scheme that can be used by the researcher him/herself to exercise reflexivity and epistemological vigilance (Bourdieu et al., 2008) as well as develop education policy meta-research studies.
The EEPA has three analytical components: the epistemological perspective, the epistemological position and epistemethodological focus (Tello, 2012).Assuming that EEPA is a focus with epistemological concern, that is, about the methodical and systematic knowledge produced in a scientific way, the epistemological perspective is a theoretical perspective that the researcher applies in his/her investigation process (e.g.Marxism, neo-Marxism, structuralism, poststructuralism, pluralism, etc.).The epistemological position arises from the very epistemological perspective or it should arise from it in a consistent and coherent investigation.It is the political position of the researcher.Some examples of epistemological position are: critical, critical-radical, critical and analytical, reproductivist, neo-institutionalist, legal and institutional, empiricist, neoliberal, etc.The epistemological position can also be understood as a derivation, a variable of the epistemological perspective.The history of knowledge and sociology of science has produced, from the complexity of social reality, multiple derivations and interpretative declinations of theoretical perspectives.The epistemological position can be understood as a pendulum that moves from one extreme to another within a particular theoretical perspective.Thus, we find, for instance, as epistemological position within the epistemological perspective of post-structuralism, positions like historical institutionalism, neo-institutionalism, Lacanian post-structuralism or Latin American poststructuralism 5 .
Finally, the epistemethodological focus is on how to methodically build research from a certain epistemological perspective and epistemological position.No methodology is neutral and, therefore, when making its epistemological foundations explicit, the researcher must be concerned with the epistemological vigilance in his/her research (methodology, data analysis, arguments, conclusions, etc.), whose construction comes from the epistemological perspective and epistemological position.To Rawolle and Lingard (2013), an important aspect of Bourdieu's theory is his rejection of the dichotomy between theory and data and between theory and methodology.Instead, he recognizes the necessary relationship between them and the impact upon one another.Lingard (2014) explains that Bourdieu rejects both 'theoreticism' and 'methodologism', i.e. the view that methodology refers only to data collection techniques and the view that theory is something distant from the data and the empirical reality.The term 'epistemethodology' aims to exactly express the existing articulateness between epistemological decisions and research methodology.Research characterized by consistency and internal coherence, the epistemological perspective, the epistemological position and epistemethodological focus show themselves articulated and integrated.A relevant issue in the context of education policy studies is that not always are such aspects explained by the researchers.In this case, they can be perceived through the systematic reading done in meta-research.In research reports and publications, it is noticeable, through meta-research, that there is not always coherence between epistemological perspective, epistemological position and epistemethodological focus.This occurs for several reasons, mainly due to the insufficient level of reflexivity and epistemological vigilance.
Bourdieu's concepts of rejection of epistemological innocence, reflexivity and the need for objectivity requires the researcher to define his/her position within the political sociology field6 and within the national education policy field (Hardy, 2009, as cited in Rawolle & Lingard, 2013).Positioning here refers to the researcher's position in relation to the object of study and the relevant academic field or related fields.To Rawolle and Lingard (2013), the position of the researcher can be defined as a position within several fields, including the field of the research object and the academic field/s in which the research is positioned.Rizvi and Lingard (2010, p. 47-48) suggest that such positions require reflexivity and consideration of the researcher's position in relation to the field and the research object, real location in terms of analysis, theoretical and methodological posture, spatial and temporal location.To Rawolle and Lingard (2013), in a certain sense, this is the reflexive application of Bourdieu's concept of 'socioanalysis' to the positioning of the researcher of policy sociology.Socioanalysis, for Bourdieu, is a way of understanding how individuals are social products and that provisions and commitments of people with practice are related to their social history, which is incorporated into their habitus.Socioanalysis involves providing a context to examine the relationship between the researcher's own arguments on social objects and his/her social history.This context involves a recounting of significant social events and social trajectory through different fields that are relevant to the research.Thus, socioanalysis represents a rethinking of a researcher's statement of interest and the impossibility of disinterested investigation.Bourdieu's argument here is that the recognition of that allows the realization of better research in Social Sciences (Rawolle & Lingard, 2013).
To Bourdieu, the concept of reflexivity is also central in the dissemination of the research.The rejection of epistemological innocence and recognition that every research is simultaneously empirical and theoretical as well as practical demands, according to Bourdieu, openness and vulnerability, true honesty in the presentation of research, whether in oral or written genre (Rawolle & Lingard, 2013).
Meta-research7 refers to the process of taking a set of texts as an object of reflection and analysis.In the case of meta-research based on education policy with an epistemological focus, we seek to identify how researchers work with epistemological issues, theories or concepts that underlie their research and how they are presented in their research reports.Thus, we seek to identify a number of elements and features such as: the epistemological perspective, the epistemological positioning, epistemethodological focus, the type of research (theoretical, empirical research, comments or critique), theoretical frameworks (concepts), levels of approach and abstraction and other aspects related to the use of theories and epistemological perspectives in research on education policy (Tello & Mainardes, 2015).
Meta-research contributes to the expansion of knowledge produced in this field, subsidizes the reflection on the possibilities of increasing scientificity and epistemological vigilance and helps to intensify the exchange of information and critiques in reference to the production of knowledge within this field which can gradually lead to steps forward in qualitative research.The results of these researches provide important elements for understanding the development of the education policy field and, at the same time, allow to outline some challenges and tensions that can inspire advancements in the production of knowledge in this field.
It is important to highlight that, based on Bourdieu's theory of social fields (Bourdieu, 2003;Wacquant, 1989Wacquant, , 2007)), education policy can be considered as a specific and autonomous academic field.However, in the case of Latin America, education policy is a new field and still in consolidation process.
Academic fields are constituted as such when they are appointed, i.e. when an institutional space is created, for example, in universities or research centers, with the creation of departments, disciplines, lines or research groups; and the creation of specialized periodics, specific associations, and so on.Based on this definition, we consider that in Latin America, education policy as an academic field emerged after the 1940s, such is the case with Mexico; after the 1950s, Argentina, and after the end of the 1960s, Brazil 8 .
Although it is a field with a significant accumulation of research on education policy, there are still many theoretical and epistemological aspects to be developed or looked into more thoroughly, such as: theoretical and methodological issues of research in education policies (including the elaboration of new theories, concepts and approaches to policy analysis); research into the history of the constitution of the field in different countries; reflection on their objects of study 9 ; analysis of epistemological perspectives and theoretical frameworks that have been used; and so on.We argue that the consolidation of the field of education policy demands, among other elements, the theoretical development of the aspects previously mentioned.

Approach and Abstraction Levels in Education Policy Research
In the context of research relating to the EEPA, the definition of some categories and classifications become necessary.One of them refers to the levels of approach and abstraction that can be identified in education policy research.
From the beginning, it is important to highlight that every typology or classification is arbitrary and it is related to specific purposes.Moreover, due to its arbitrary nature, the same objects can be classified in different ways 10 .Thus, the classification that we present below is a theoretical exercise on education policy research, with no pretension of judgements or creation of hierarchies.
The field of education policy in Latin America (and in other contexts) is characterized by being a comprehensive field, involving studies of a theoretical nature, policies and programs analysis, 8 Regarding the establishment of the academic field of education policy in Brazil, see Stremel (2016). 9Regarding the discussion on objects of studies of education policy, we point out the book organized by Tello (2015). 10Thiry-Cherques (2006, p. 29) explains that: 'Although heir to the philosophy of science, Bourdieu refuses to apply classification systems to objects that he investigates (Bourdieu, 1992a: 184).He understands that any typology crystallizes a situation, that is, it tends to be arbitrary, as it dismisses the typologies that do not conform and in cases that are equivocal, i.e. cases that are not clearly distinguishable.He owes to Bachelard (1984) the idea that thinking operates as a pincer movement, which discovers, integrates and overcomes the limitations of theories in a conceptual, increasingly comprehensive composition'.policy of education and school management, education financing, curriculum policies, education legislation, teaching work (valuation, remuneration, career, etc.).When developing their investigations, researchers use the theoretical frameworks in many different ways.In research reports, different levels of abstraction can be identified, which can range from predominantly descriptive studies to studies with a higher level of complexity and theory (understanding).The following are the main characteristics of these three levels of approach and abstraction.

Description
In the case of research in education policy, studies that are predominantly descriptive are those that present a set of ideas (papers of theoretical or bibliographical nature) or empirical data, with little analysis of the ideas or the data presented.
Although they may provide some theoretical foundation, they are studies with a low level of integration between theory and data.Among the descriptive studies, there are those that present a significant and relevant set of data (e.g.statistical information), which are poorly explored in the light of theoretical frameworks.There are also studies that present results of researches with few subjects or are very focused on a specific context (local).In this case, what it demonstrates is not the amount of subjects or scope of the research, but the approach style (purely descriptive).
In general, they do not have arguments or original analysis.In some cases, they are researches that are based on linear models of policy analysis (agenda formulation, implementation, evaluation), but there are also studies that are based on theories or approaches considered consistent, but are used in a precarious and incipient manner.
Descriptive studies focus on characterizing the selected object of study.In general, we observe that they meet the basic components of research in education policy; in other words, they comply with what Ball calls 'surface epistemology' (Ball, 2011), because there is an object of study, definition of objectives and methodology 11 .However, although they technically meet such components, they do not develop in the analysis of the political process, characterizing it superficially.
In many cases, in descriptive studies, the researcher has ideas and a priori answers, hindering a proper dialogue between the empirical data and theoretical perspectives employed in the research.In this context, the theory is 'applied' to the data with templates already pre-established to the reality and to what the researcher assumes that it 'should be'.
It is important to highlight that what we call here a descriptive level differs from Clifford Geertz's concept of 'thick description'.The thick description is not simply a matter of presenting relevant details, but instead, describing the social action in a dense form based on a complex and interpretive process (Schwandt, 2015).
11 Ball (2011) also mentions the deep epistemology.To Ball, a deep epistemology involves itself with wider issues and deep assumptions of power, truth, subjectivity (Mainardes, 2015a).The explicitness of deep epistemological research involves reflections and problematizations about the theoretical assumptions employed, as well as structural discursive or economic bases that are being used to understand the object under investigation.To Ball, dealing with both epistemologies, in research, is not a re-articulation of macro and micro, but '[...] an erasure of that binary to see policy as a set of techniques, categories, objects, and subjectivities' (Mainardes, 2015a, p. 167).
The result of meta-research has allowed the verification that the epistemological positioning resulting from this level of abstraction, in many cases, can be designated as 'empiricist', as there is a concern with the presentation of data, while the analysis is undeveloped.

Analysis
In predominantly analytical studies, data and ideas are worked, categorized, compared.One of the important characteristics of analytical studies is more integration between theory and data.The theories are not merely applied, because the effort of analysis results in the generation of concepts, categories, typologies, empirical generalizations.Due to a more systematic use of a theoretical process and a more comprehensive and systematic analysis, discoveries and research findings become more universal, with a higher level of generality, being able to be extended and applied to other contexts.
In general, they are studies in which researchers explicit their arguments, constructing a more unique and consistent study.In many studies of this level, the epistemological position of the researcher becomes evident in the analysis, due to being a central element during argumentation, and from the establishment of relations between theory and data and generation of theoretical elements from the data.
During the analytical education policy research process, the initial phase becomes more complex because there are no previous or fixed responses and, therefore, the object of study becomes more dynamic and somehow, in Bourdieu's terms (Bourdieu, Chamboredon, & Passeron, 2007), the object is constructed as one advances in research.The construction of the object refers to the research process as a whole, including the reflection on the theoretical perspective itself, which can be taken as hypothesis (Brandão, 2002).
In meta-research, we can observe that there are different levels of analysis (more developed, less developed, etc.) and the theoretical framework is an essential element for the construction of the analytical process.

Understanding
The level of understanding is the highest and most advanced level of abstraction.This level may contain some degree of description and a significant number of analyzes, which are subsumed by understanding.They are studies that present a double dimension of the research process: explaining and understanding (the interpretive and explanatory character).They are studies that seek to address the thematics (theoretical or empirical) in a more totalizing way, exploring in depth the relationships and the determination involved in the policy under investigation or the issue that is being discussed.In general, they are studies that present greater richness and depth to the analysis, and may also serve as a basis for further research.In these studies, we can observe a strong and consistent relationship between the epistemological perspective, epistemological position and epistemethodological focus, even when the epistemological perspective is not explicitly shown.Applying the principles of grounded theory (Glaser & Strauss, 1967), the studies of understanding present an essential aspect in the knowledge production process: generation of theory.The generation of theory can be identified through the development of concepts, categories, typologies, explanation, or through sensitizing concepts12 , which, due to their level of generality and coherence, represent advances in the knowledge production of the field.Susen (2011) explains that, in the philosophy of Social Sciences, the tension between these two epistemological positions (explanatory design and interpretationist design) is generally addressed in terms of the difference between the method called Erklären (explain) and the method called Verstehen (understanding).In essence, the tension between these epistemological perspectives illustrates that Bourdieu's concept of Social Science is based on two concerns: on one hand, Bourdieu refers to Social Science as an instrument through which one explains the nature of the underlying structures that determine the involvement of human actors with the world; on the other hand, Bourdieu considers Social Science as a tool to understand the nature of the specific discourses of the field through which human actors interpret the world.In the first sense (explain), the task of Social Science is to uncover the interest-laden lawfulness os the human world.In the latter sense (understanding), the task of Social Science is to examine the meaning-laden discursiveness of the human world.Thus, from a Bourdieusian point of view, Social Sciences are confronted with the task of producing ambivalent types of knowledge both explanatory and interpretive, that is, both representational and expressional, factual and constructional, objective and subjective, descriptive and normative.The social and scientific engagement with the world requires both an explanatory and interpretive study of human reality (Susen, 2011).
When carrying out meta-research, studies classified as 'understanding' stand out among the others because they present greater analytical and theoretical density, as well as more assertive and thoughtful interpretations and explanations.In general, in these texts, the epistemological perspective and epistemological position are made explicit by the author or can be easily identified from the analyses, arguments and conclusions.

Meta-Research and Levels of Abstraction
As already mentioned, the meta-research contributes to the comprehension of the development of the education policy field and, at the same time, helps outline some challenges and tensions that can inspire breakthroughs in the production of knowledge in this field.
It is important to note that meta-research is distinct from literature and systematic review.While, in the literature review and systematic review, the aim is to synthesize the results of research on a particular subject or theme, meta-research aims to analyze research reports and education policy publications, aiming to explore the theoretical and epistemological foundations or other aspects (objectives, methodology, research questions, etc.).
Meta-research has a dual dimension: 'reflective' and 'theoretical and analytical'.In its reflective dimension, it allows a mapping of what has been researched and on which theoretical concepts' (culture, institutions, social structure, personality, etc.).Sensitizing concepts do not involve fixed and specific procedures to identify a set of phenomena, but instead they offer a sense of reference and guidance in the approach of empirical instances.So, while the definitive concepts provide prescriptions of what to see, the sensitizing concepts only suggest directions where to look (Blumer, 1954).It is a relevant concept in the grounded theory, as 'the sociologist should be theoretically sensitive enough to be able to conceptualize and formulate a theory from the data' (Glaser & Strauss, 1967, p. 46).Some examples of sensitizing concepts in the field of education policy include: education policy of containment and education policy of release (Cunha, 1975);bureaucratic-authoritarian State (O'Donnell, 1982); Globally structured educational agenda (Dale, 2004); hegemonic globalization and counter hegemonic globalization (Santos, 2004); the concept of symbolic analysts (Brunner & Sunkel, 1993); the theory of policy enactment (Ball, Maguire, & Braun, 2012); concept of performative implementation (Ball, Maguire, & Braun, 2012), among others.perspective the research is based.In its theoretical and analytical dimension, the analysis and the conclusions of meta-research can serve as a reference for more thorough researches in the field or to build new perspectives and research approaches.To Bourdieu (2011, p. 38), when the research has as object of study the universe where the research takes place, the acquisitions that it ensures 'can be immediately reinvested in scientific work (...)'.Thus, the meta-research contributes to the expansion of knowledge produced in the field, subsidizes the reflection on the possibilities of increasing scientificity and epistemological vigilance and helps to intensify the exchange of information and critiques about the production of knowledge in the field that can gradually lead to qualitative leaps in research in this field.
A basic scheme of meta-research in education policy may include, among others, the following aspects: a) thematics investigated in a given period, place or form of disclosure; b) type of research: theoretical, empirical, comments or critique; c) cited authors; d) elements of the education policy epistemology focus: epistemological perspective, epistemological position and epistemethodological focus; e) levels of approach and abstraction; f) theoretical frameworks (concepts used throughout the text); g) scope (global, national, local, etc.).The apprehension of levels of approach and abstraction demands reading and analysis of text, seeking to understand the author's reasoning, argumentation, the use of theoretical tools, analyzes, conclusions, the dialogue with the research area, etc.The level of approach and abstraction is not something that is located in a part of the text.It is something that can only be noticed as a whole and is strongly related to the use of the theoretical framework, data analysis (in the case of empirical research), arguments and conclusions.It is also important to highlight that the levels of approach and abstraction are not homogeneous within the same level, as there are different levels of description, analysis and understanding, some being more original, coherent and reflective than others.

Final Considerations
In this paper, we presented the main features of the three levels of approach and abstraction that can be identified in studies of education policy.Sayer (1984), based on the critical realism, considers that the complex systems can be understood in terms of events, mechanisms and structures.Researches of description level can be understood as those that exploit events (for example, the formulation and implementation of policies or programs), with unique focuses, in particular contexts (Ball, 2011), whereas the level of analysis relates to the identification of more general 'mechanisms', regularity identification, logic of intervention identification, etc. (Dale, 2004(Dale, , 2007(Dale, , 2010)).The level of understanding involves the identification of structures, that is, of the general conditionings of the policies (processes and structures that fit the agendas of education policies and run them) 13 .
It is important to highlight that education policy researchers develop researches with different purposes and each report or publication has a purpose, validity and importance.What we want to argue is that, in the current level of development of education policy research in Latin America, it is necessary to increase the number of understanding studies.Such studies are essential to strengthening the field, as they allow advances in knowledge about education policy and are studies that may provide the basis for further studies.Education policy research requires more studies of this level to establish itself as a field that allows the comprehension of the increasingly complex political, social and educational reality phenomena. 13An explanation of events, mechanisms and structures can be found in Mainardes (2015b).
Due to the socio-historical circumstances from decades past, education policies have focused on education and school management centered in education administration and organization of education systems, predominating the instrumental dimension of education policy.Somehow, the effect of 'know-how' prioritizes professionalism and interventionism instead of the dimension of theoretical and analytical reflection to understand reality.For this reason, the distinction of levels of abstraction and approach is a relevant issue in the present historical moment We must also point out that the historical matrices, along with epistemes of the period, have generated linear investigation forms that historically legitimize themselves and modes of analysis and investigation in education policy that are merely descriptive, through the analysis of events (policies, programs) and studies aiming at problem solving (reality overcoming) 14 .
Meta-research also allows to verify that the conscious and reflective use of the theory has an essential role in the setting of the level of understanding.Generally, they are studies with a high degree of integration between theory and data, with conceptual depth and generation of theory (categories, refined concepts, causal explanations, relationship establishing, etc.).In some cases, in texts of experienced researchers, the theory appears incorporated into the arguments, analysis and epistemological positioning, and it is not constituted in an isolated section.We also note that in studies of a theoretical or empirical nature that reach the level of understading, epistemological position of the researcher emerges as an essential and articulate element.It constitutes the element that seems to direct the analysis and approach of the theme.
Meta-research brings relevant contributions to the field of education policy, as it allows to identify trends and general standards in research and publication, as well as this, they may indicate some important referrals in researchers' education 15 .Regarding researchers' education, some referrals may be mentioned as relevant, such as: the need for theory enhancement in the education process; the study of different epistemological perspectives; discussions about the role of reflexivity, the epistemological vigilance and different possibilities of theory application.Although the field of education policy in Latin America is evolving by way of its continued strengthening and consolidation, it is essential to invest in innovative forms of research and theorizing, different ways of applying the theory in research, expansion of the conceptual discussions and break away from the repetition of orthodoxies.
14 According to Tello (2013), studies to overcome reality are those that aim to change the reality or improve it through concrete lines or intervention and overcoming proposals with the implementation of certain policies or programs.Generally, they are researches conducted by researchers linked to national or international organizations, foundations and research institutes, hired by official agencies (Ministry of Education, Departments of Education) or independent researchers.Tello (2013) argues that the technical reports and overcoming projects of the reality should not be confused with the academic knowledge production that has other characteristics.In general, academic research aims to comprehend the reality for a possible overcoming or transformation.The studies of reality overcoming can be compared to studies designated as 'problem solving approach' (Cox, 1996;Dale & Robertson, 2012).Problem-solving theories are oriented towards maintaining the status quo; they are ahistorical and aim to make the institutions a little better through small changes within the boundaries and parameters of the problems.Cox (1996) considers the critical theory opposed to the problem-solving approach.Dale & Robertson (2012) indicates the existence of three analytical models, with increasing levels of abstraction, namely: problem solving, critical perspective and explanatory perspective.This last perspective of analysis is based on the principles of critical realism. 15Meta-research in education policy can also be classified as description, analysis and understanding.The meta-research that reaches the level of understanding could bring important contributions to the advancement of research in education policy, especially concerning theoretical and epistemological issues and theorizing strategies that can be identified in the research of this field.