北京师范大学
教育研究方法讲座系列(2):
教育政策研究
第二讲
教育政策研究的知识论基础：理论视域的探讨
A. Perspectives in Policy Studies in Education: An Overview
1.Analytical-technical perspective
a.Epistemological premise: Public policiesaresocial facts. They are social actions, programs and projects undertaken by the modern stateto intervene the state of affairs of particular public domains in a modern society.
b.Aims of enquiry: Accordingly, policy studies is scientific enquiry aims to provide causal explanation for the question why the state undertaking particular policy actions and not the otherwise. More specifically, it aims to analytically identify and verify the antecedent conditions that caused the policy action to take place.
c.Practical premise: Based on the causal relation verified by policy studies, policy makers can then make prediction, means-ends calculation, and technical engineering about the policy situation concerned. It aims to impose technical control over the situation.
2.Interpretive-political perspective:
a.Epistemological premise: Public policies are social construction of realities. They are meanings, values, preferences and desires attributed by the modern state and others interest groups to the state of affairs of particular public domains in a modern society.
b.Aim of enquiry: Accordingly, policy studies is social enquiry aims to interpret and explain why particular meanings and values are signified in a policy “text” in a policy context, and not the otherwise.
c.Practical premise: Based on the interpretations and understandings revealed by policy studies, policy participants can then engage in communication and dialogue which aim to facilitate mutual understanding,to nurture consensus, and plausibly to work out politically reciprocal solution to the policy issue in point.
3.Discursive-critical perspective:
a.Epistemological premise: Public policies are authoritative valuesand even “effective discursive totality”legitimized and imposed by the modern state on the state of affairs in a particular public domain in a modern society.
b.Aim of enquiry: Accordingly, policy studies is critical enquiry aims to reveal how and why particular policy discourses are legitimized in a policy arena.
c.Practical premise: Based on the critical studies on policy discourse, policy critics can then reveal and assess the possible systemic biases and distortionshypostatized and legitimatized in particular policy domainand to strive to liberate human and social potentials from these biases and distortions.
(I)
The Analytic-Technical Perspectivein Policy Studies
A.The Epistemological Basis: Analytical Positivism
1.Stuart Nagel’s conception of policy analysis model
a.“Public Policy analysis can be defined as determining which of various alternative public or governmental policy will most achieve a given set of goals in light of the relations between the policies and the goals. That definition brings out four key elements of policy evaluation which are:
i.Goals, including normative constraints and relative weights for the goals.
ii.Policies, programs, projects, decisions, options, means, or other alternatives that are available for achieving the goals.
iii.Relations between the policies and the goals, including relations that are established by intuition, authority, statistics, observation, deduction, guesses, or other means
iv.Draw a conclusion as to which policy or combination of policies is best to adopt in light of the goals, policies, and relations.” (1986, p. 247)
b.Stokey and Zeckhauser’s Framework for policy analysis
i.Establishing the Context. What is the underlying problem that must be dealt with? What specific objectives are to be pursued in confronting this problem?
ii.Laying out the alternatives. What are the alternative courses of action? What are the possibilities for gathering further information?
iii.Predicting the consequences. What are the consequences of each of the alternative actions? What techniques are relevant for predicting these consequences? Of outcomes are uncertain, what is the estimated likelihood of each?
iv.Valuing the outcomes. By whatcriteria should we measure success in pursuing each objective? Recognizing that inevitably some alternatives will be superior with respect to certain objectives and inferior with respect to others, how should different combinations of valued objectives be compared with one another?
v.Making a choice. Drawing all aspects of the analysis together, what is the preferred course of action?
c.In searching of causality, prediction, and prescription for policy action and logical positivism emerged from natural science seemingly pointing the way. And three basic premises of logical logical positivism
i.Methodological monism
ii.Logical empiricism as the ideal-typical method of verification
iii.Deductive-Nomological model as the idea-typical model of explanation
2.Deductive-Nomological (D-N) explanation:The ideal-typical model of causal explanation in logical positivism:
The D-N explanation is the type of explanation commonly used in researches in natural sciences. It makes up of three parts:
a.The explanatory premises or the casual law (covering law), which is a universal statement of the sufficient and necessary conditions (explanans/cause) for the truth of the explanandum (effect).
Accordingly, a causal law in natural science must comprises the following components
i.The factual truth of both the the explanandum (i.e. the phenonmenon to be explained) and the explanans.
ii.The conditionality between the explanandum and explanans
-Sufficient conditions: It refers to the kinds of conditionality between the explanandum and explanans, in which the explanans can exhaustively but not universally explain the truth of the explanandum.
-Necessary conditions: It refers to the kinds of conditionality between the explanandum and explanans, in which the explanans can universally but not exhaustively explain the truth of the explanandum.
-Sufficient and necessary conditions: It refers to the kinds of conditionality between the explanandum and explanans, in which the explanans can both exhaustively and universally explain the truth of the explanandum.
iii.The temporal order of the explanans must be in precedence to the explanandum
b.The initial condition, which defines the property of a specific case of the explanandum.
c.The conclusion, which state the exhaustive explanation of the specific explicandum by the explanans.
3.The compromised model: Statistical-Probabilistic (S-P) explanation:
The S-P model is the type of explanation commonly use in quantitative researches in social sciences. It is also made up of three parts similar to those in nomological-deductive explanation. There are two differences in probabilistic explanation. One is that the explanatory premises is not in the form of law-like / nomological statement of the sufficient and necessary conditions of the truth of the explanandum but only a probabilistic statement specifying the likelihood of the causal relationship between the explanans and explanandum. The second difference is that in the conclusion, the specific explanandum under study cannot be exhaustive explained by the explanans but can only be explained in probabilistic terms.
4.Logical-empiricism: The exemplary method of verification in logical-positivism
a.By empiricism, it refers to the method of verification based primary by sensory experiences of human being. More specifically, it is based on recorded experiences methodically collected by scientists. More importantly, these recorded experiences will then be set against their respective propositions to see whether they correspond each other. And it is through this operation of so call correspondence principle that scientific propositions will be verified against the external national world.
b.Apart from empirical verification that rely on human experiences, scientists can also rely on pure logical inference and mathematical calculations to verify their propositions. For example propositions in in geometry and mathematical physicalare usually not verified with empirical data but pure mathematical and logical inferences.
B.The Ontological Basis: Technical Rationalism
1.The concept of instrumental and technical rationality
a.The concept of rationality: Rationality can be defined as conscious and knowledgeable ways human beings approaches and even masters the world around them. Rationality therefore is a state of mind and a way of life, in which human beings strive to master their physical and even their social environments.
b.The concepts of instrumental and substantive rationality
i.Instrumentalrationality refers to conscious and knowledgeable process through which human beings calculate and choose the most expedient means to achieve preconceived and/or predetermined end.
ii.Substantive rationality refers to conscious and knowledgeable process through which human beings decide the ends most worthy of achieving.
c.The instrumental-technical turns in policy studies
i.Policy scientists who adhere to value-neutral or even value-free method of inquiry advocate that substantive choice of policy end are political decisions and should be left to politicians.
ii.Accordingly, they contend that policy scientists should confine themselves to the technical issues of choosing the best, or more specifically the most cost-effective policy instruments or means to attain the “politically” pre-determined ends.
2.Technical-rational perspective in policy studies
a.Following the conclusions drawn from analytic-positivist policy studies, the next taskto be performed by policy analysts is to work out, if possible to the last technical details, the action plan to carry out the policy measures. Hence, it is a task guarded by instrument and technical rationality.
b.Assumptions of comprehensive (technical) rational model in policy studies: (Forester, 1989, Pp. 49-54)
i.The agent/actor: A single decision-maker (or a group of fully consenting decision makers) who is a utility-maximizing, instrumentally rational actor
ii.The setting: Analogous to the decision-maker’s office, “by assumption a closed system”
iii.The problem: Well defined problem, “its scope, time horizon, value dimensions, and chains of consequences are clearly given” and close at hand.
iv.Information: Assumed to be “perfect, complete, accessible, and comprehensible.”
v.Outcome: A single best solution or the most optimum resolution
C.Criticismsand Revisions
1.Herbert Simon’s concept of bounded rationality
a.Simon’s defines that “rationality denotes a style of behavior (A) that is appropriate to the achievement of a given goals, (B) within the limits imposed by given conditions and constraints.” Simon, 1982, p.405)
b.The concept of satisfice: Simon differentiates two stances in regard to (A), i.e. the degree of “appropriateness to goal achievement.
i.Maximizing or optimizing stance of the “economic man”: “While economic man maximizes - selects the best alternative from among all those available to him”
ii.Satisificing stance of the “administrative man”: “Administrative man satifices - look for a course of action that is satisfactory or ‘good enough’. (Simon, 1957, p. xxv)
c.The concept of bounded rationality: In regard to (B), Simon indicates that “It is impossible for the behaviour of a single, isolated individual to rearch any high degree of rationality. The number of alternatives he must explore is so great, the information he would need to evaluate them so vast that even an approximation to objective rationality is hard to conceive. Individual choice takes place in an environment of ‘givens’– premises that are accepted by the subject as base for his choice; and behaviour is adaptive only within the limits set by these ‘givens’.” (Simon, 1957, p. 79; my emphasis)
Simon specifies limitations imposed by the environment of givens are
i.Limitation of the knowledge
- Incomplete and fragmented nature of knowledge,
- Limits of knowledge about the consequences, i.e. predictability of knowledge
ii.Limitations of the cognitive ability of the decider makers
- Limits of attention
- Limits on the storage capacity of human mind
- Limits of the learning ability of human beings, i.e. observation, communication, comprehension, ….
- Limits on changes of status quo, i.e. human habits, routine, mind set, …
- limits on organizational environments.
2.Dahl and Lindblom’s conception of rational calculation
a.Limitations and difficulties in means-end rational calculation
i.Information deficiency: Relevant or even essential information to the means-end rational calculation may be incomplete, unavailable, difficult to obtain, …
ii.Communication problem: Available information may not be able to be dissimulated to all decision-making parties or the information may appear to be difficult to comprehend.
iii.The number of variables involved is too many to be exhausted.
vi.The complexity of the relations among variables is too complicated to be comprehended not to mention exhausted.
b.Scientists’ solutions to cognitive deficiency in means-end rational calculation
“Scientists deal with the problem of information by systematic observation, with the problem of communication by developing a precise and logical language usually including the language of mathematics; with the problems of an excessive number of and complex relations among variables by specialization, controlled by experiment, quantification, rigorous and system analysis, and exclusion of phenomena not amenable to these methods.” (Dahl & Lindblom, 1992, p. 78) In summary, these methods include
i.Codification: Method of reducing and unifying numerous, complicated and disorderly information into comprehensible units
ii.Quantification: Method of quantifying information and units into comparable values.
iii.Sampling: Selectively analyzing a fragment, a specimen of the phenomenon under observation.
vi.Observations in control situations or by randomization.
v.Modeling: Model “is a purposeful reduction of a mass of information to a manageable size and shape, and hence is a principal tool in the analyst’s work-tool. Indeed, we will be employing models throughout this book.” (Stokey & Zeckhauser, 1978, p.9)
3.Choices under calculated risk
a.Risk can be construed as “the residual variance in a theory of rational choice” (March, 1994, p. 35) or more specifically, the unexplained variance in a causal modeling equation. It is basically grown out of the epistemological constraints of the scientific means-end rational model.
b.Therefore, “calculated risks are often necessary because scientific methods have not yet produced tested knowledge about the probable consequences of large incremental changes…and existing reality is highly undesirable.” (Dahl & Lindblom, 1992, p. 85)
c.Growing industry for risk estimation and risk management in public policy
4.Charles Lindblom’s science of muddling through
Charles Lindblom agrees with Simon on the limitations of human rationality, yet Lindblom diagnoses that the sources of these limitations are more than the cognitive capacity of human mind. He suggests that limitations are integral parts of the very process of policy making. Lindblom characterizes this process as “successive limited comparison” and “muddling through”.
a.“Incrementalism is a method of social action that takes existing reality as one alternative and compares the probable gains and loses of closely related alternatives by making relatively small adjustments in existing reality, or making larger adjustments about whose consequences approximately as much is known as about the consequences of existing reality, or both.” (Dahl & Lindblom, p. 82)
b.Lindblom’s two models of decision-making
Rational comprehensive / Successive limited comparison1a / Clarification of values or objectives distinct from and usually prerequisite to empirical analysis of alternative policies / 1b / Selection of values, goals and empirical analysis of the needed action are not distinct from one another but are closely intertwined
2a / Policy formulation is therefore approached through means-ends analysis: first the ends are isolated; then the means to achieve them are sought / 2b / Since means and ends are not distinct, means-ends analysis is often inappropriate or limited
3a / The test of a ‘good’ policy is nthat it can be shown to be the most appropriate means to desired ends / 3c / The test of a ‘good’ policy is typically that various analysts find themselves directly agreeing on a policy (without their agreeing that it is the most appropriate means to an agreed objective)
4a / Analysis is comprehensive; every important relevant factors is taken into account / 4b / Analysis is drastically limited:
a. important possible outcome are neglected;
b. important alternative potential policy are neglected;
c. important affected values are neglected
5a / Theory is often heavily relied upon / 5b / A succession of comparisons greatly reduces or eliminates reliance on theory
5.John Forester’s typology of bounded rationality (see Table 4)
a.Bounded rationality I: Cognitive limits
b.Bounded rationality II: Social differentiation
c.Bounded rationality III: Pluralist conflict
d.Bounded rationality IV: Structural distortions
1
W.K. Tsang
Policy Studies in Education
1
W.K. Tsang
Policy Studies in Education
D.A Case Study HKSAR
1.Analytic-positivist perspective in education policy
a.Codification
i.The case of MOI: MIG, EMI-capables, EMI-schools,
ii.The case of HKSAR education reform: Quality schools, quality teachers, quality school management, quality indicators
b.Quantification:
i.The case of MOI:
- MIGAI, II, and III,
- EMI-capables: the top 40% in pre-S1 HKAT,
- EMI-schools = on a 3-year average, 85% of S1 intakes being EMI-capables
ii.The case of HKSAR education reform
- Performance Indicators for HK Schools, which are analytically divided into 4 domains, 14 areas, 28 components, 186 performance evidences each of which is in turn measures a 4-point scale
- 23 Key Performance Measures,
- Value-added index of SVAIS,
- Basic Competence Assessment
- The teacher Competencies Framework , which are analytically divided into 4 domains, 16 areas, and 46 indicators; which are in turn are measured by 5-point scales and ranked into 3 levels.
- 5-level Language Proficiency (Benchmark) Assessment for teachers
c.Sampling and randomization: Sampled schools and their attributes found in policy studies can be applied to other non-sampled schools in the assumption that factors other than those controlled by policy measures can be randomized.
i.The case of MOI policy:
- Classification of MIG-I, -II & -III or EMI-capable students
- Classification of EMI and CMI schools
ii.The case of HKSAR education reform
- Classification of quality schools
- Classification of professionally competent and/or linguistic proficient teachers
- Classification of students and schools passed the Basic Competence Assessment
d.Modeling:
i.Monolingual model of mother-tongue instruction vs. triglossiic model
ii.Input-output model: Value-added model, linear regression model, ordinary-least-square model, multi-level regression model, …
iii.Educational process model: Quality school model, Quality-Assurance Inspection, School-Self Inspection model, External School Review…
iv.Model of professional teachers: Analytically fragmented and technocratic professional vs. heuristic learned-professional
2.Technical-rational perspective in education policy
a.Technocratic procedures of assessments and evaluation
i.Procedures and criteria of the application for EMI schools
ii.Procedures and criteria of TSA, pre-S1 HKAT, value-added index
ii.Procedures of QAI, SSE, and ESR in Quality Assurance Mechanism