Topic 4 John Gerring PDF

Preview

Summary

Summary of reading...

Description

What is a case study? 1. Qualitative (small-N) 2. Ethnographic, clinical, participant-observation (in the field) 3. Characterized by process tracing 4. Investigates the properties of a single case 5. Investigates a single phenomenon/ common usage 1-3: inappropriate as general definitions - Best understood as describing certain kinds of case studies 4: equates case study with the study of one case (N=1) research design Wrong Case studies always employ more than one case 5: Centering on phenomenon/instance/ example → correct but ambiguous Proposition of author: ● Define case study as an intensive study of a single unit for the purpose for understanding a larger class of similar units ● ● ● ● ●

Population comprises of sample (studied cases) + unstudied cases Sample comprises of several units Each unit is comprises cases Case comprises relevant dimensions (variables) Each variable built upon an observation/ observations

● ●

The Case Study Method Considered as an Empirical Endeavor Casual relationships: covariational in nature Cause + effect are covary 共變

The N Question ● A study provides a case (N) of France ○ Study French Revolution ○ Broaden analysis: include a second revolution (American Revolution) ○ Common to describe the study comprising 2 cases (but it’s a gross distortion) ○ Correct to describe such study as comprising 2 units rather than 2 cases ○ Case study of a single event examines the event over time 3 types of case studies ● Type I: observe a single unit over time ○ E.g. France is observed before, during and after the event to see what changed and what remained the same after the event ○ Patterns of covariation offer the empirical clues that are required to reach conclusions about causation ○ Create multiple cases out of individual units (i.e. before + after revolution, N=2) ● Type II: no temporal 時態的 variation ○ ○

E.g. French Revolutions is examined at a single point in time Covariational patterns within that unit

○

●

If primary unit analysis is nation-state, within-unit cases might be constructed from provinces, localities, groups or individuals Type III: Combines both temporal + within-unit variation Ambiguities - Necessary and Unnecessary

What is a case study? Intensive study of a single unit wherever the aim is to shed light on a question pertaining to a broader class of units This definition helps clarify 6 ambiguities 1. Problem of distinguishing different types of covariational evidence - Complexity of within-unit analysis - Often multiple and ambiguous 2. Blurry line between a unit that is intensively studied (the case study) and other adjacent units that may be brought into the analysis in a less structured manner - Case study refers to a set of units broader than the one immediately under study - Writer must have some knowledge of additional units - To choose a unit for special treatment - Identify plausible hypothesis - Case studies are not immaculately 一塵不染 conceived - Additional units always loom in the background - Distinction between formal and informal units - Formal unit: the unit chosen for intensive analysis - E.g. the person, group, organization, phenomenon which the writer has in-depth knowledge - Informal unit: other units that are brought into analysis in a peripheral way - Typically in an introductory or concluding chapter - Studied only through secondary literature - More superficially surveyed than the formal unit under study - Distinction between a formal and informal unit is a matter of degree - The more equality of treatment granted to peripheral units, the more the study leans toward a cross-unit style of analysis - The greater the predominance of a single unit, the more it merits the appellation case study 3. Single work combines single unit and across unit analysis in a formal manner - Sample composed of more than one units - Case studies - Have a population of cases - Series of case studies= sample - More case studies, less intensively each one is studies - More confident in representativeness, more likely to describe them as a sample rather than a series of case studies 4. Generally partake of 2 empirical worlds - Studies tout court + something more general - Population restricted to unit under investigation

5. Status of a work may change as it is digested and appropriated 挪用 by a community of scholars - Meta analysis: systematic attempts to integrate results of individual studies into a single quantitative analysis, pooling individual cases drawn from each study into a single dataset - Treating literature review as case studies in some larger project 6. Case study is intended to prove or demonstrate - Can be both suggestive and falsifiable Types of Inference: Descriptive vs Causal - Descriptive: Case A is like/ unlike other units (B and C) - Complicated descriptive case: classificatory relationship among A, B and C - Implicitly comparative (must have a cross-unit reference point) - E.g. green = not blue Scope of Proposition: Breadth and Boundedness vs Depth - Breadth + Boundedness: cross-unit cases - Study more than one country to make an argument about the universe of nation states - Case study of France offers better evidence for an argument about Europe than for an argument about the world - Boundedness: degree to which it exploits cross-unit variance - Case study often produce inferences with poorly defined boundaries - Because their focus is so tight - Depth: detail, richness, completeness, wholeness and degree of variance - Dilemma: knowing more about less/ knowing less about more Unit Homogeneity: Case Comparability vs Representativeness - Single unit studies: cases are likely to be comparable - Fall short in representativeness (causal relationships evidenced by that single unit may be assumed to be true for a larger set of (unstudied) units - Addition of units can hardly increase the case comparability of a sample - Unlikely that addition of units will decrease representativeness of a sample Causal insight: causal effect vs causal mechanisms - Quantitative researchers: estimation of causal effect - The effect of Y on a given change in X - Causal effect: cross-unit studies - Because single unit study cannot assume the behavior of one unit will be indicative of the behavior of other units - Causal mechanism - X Must be connected to Y in a plausible fashion - Unclear whether a pattern of covariation is truly causal in nature - Case studies enjoy a comparative advantage - Allow one to peer into the box of causality to the intermediate causes lying

between some cause and its purported effect Causal Relationship: Invariant vs Probabilistic - Invariant causal relationships: asserted to be always true, given a set of background circumstances - Take form of necessary, sufficient arguments - Probabilistic arguments: true in probabilistic fashion - A cause increases the likelihood of an outcome or the magnitude of a scalar outcome - Case study research: easier to address invariant causes Strategy of Research: Exploratory vs confirmatory - Case study: advantage in exploratory nature - Serve as impediments in work of a confirmatory nature - Can test a multitude of hypotheses in a rough-and-ready way - Theory generating - Multiple-unit study: allow testing of only few hypotheses but a greater degree of confidence - Confirmatory nature - Less room for authorial intervention because evidence gathered can only be interpreted in a limited number of ways - Prosaic 平淡 - Case studies are often useful in conjunction with a cross-unit study for the purpose of elucidating causal mechanisms - General theories rarely offer the detailed and determinate predictions on withinunit variation that would allow one to reject hypothesis through pattern matching (Without additional cross-unit evidence) - Theory confirmation is not the case study’s strong suit Q1. According to Gerring (2004), there are two main types of research: case study and multipleunit study, the former one is conducive to theory generating whereas the latter is more useful in theory confirmation. Is it necessary for a researcher to propose and test one’s theory through engaging in both case study and cross-unit study so that his/ her theory can be a confirmatory one? Which kind of study is more important or popular in modern political science scholastic world, theory generating or theory testing? Would theory generating become rare as time passes since more well-known theories have been established to explain social phenomena, and researchers would rather focus on testing the falsifiability of theories by applying it to more units? Useful variance: Single vs multiple unit Preferred research designs 1. Laboratory/ field experiments 2. Natural experiments (single unit undergoes unmanipulated change through time that approximates a true experiment) 3. Thought experiments (counterfactuals)

4. Statistical controls (a quasi-experimental method of neutralizing irrelevant variables so as to isolate the true causal effects of one/few factors of theoretical interest) - Experiment involving single unit may be more useful than multiple units that attempt to mimic the virtues of the experimental method with purely statistical evidence - E.g. Investigating the relationship between campaign efforts and voter turnout - One be more convinced by a field experiment conducted in a single community than by a multiple cross-community studies or individual poll data that rely on a host of unsatisfactory quasi-experimental controls - E.g. investigating the role of electoral systems in conditioning public policy outcomes - One might be more convinced by a single natural experiment (a change in country’s electoral system) than a cross country study (employing statistical controls to examine complex causal relationships - E.g. Whether early democratization leads to a quiescent working class and lower levels of social welfare development is difficult to investigate cross-nationally for the simple reason that only one country granted suffrage to the male working class prior to industrialization - Historical study focused on the US may provide the most compelling evidence of a general proposition

-

Q2. According to Gerring (2004), when there are not any other useful variance across units, the case study method is considered as the only way to provide evidence of a general proposition. Yet, case study has the drawback of lacking representation. In this case, should the researchers investigate a topic with minimal examples (i.e. maybe only one) available for study? Or should they change to another topic which may provide a profusion of units for them to compare and study to enhance the study’s representativeness and credibility? If researchers insist on conducting case studies with few examples, how can they ensure that their studies can become authoritative in the future? Case studies often focus on rare historical events

Ontological Considerations Ontology: vision of the world as it really is - A more or less inherent set of assumptions about how the world works - Cannot be proved or disproved - Q3. According to Gerring (2004), when adjacent units are thought to be entirely non comparable, the case study method is impossible. Whereas if the units are highly similar to each other, it is said that nothing can be learned about one unit by studying another. However, if analogous units possess identical characteristics, they may also be useful for strengthening the study’s representativeness. So, to what extent are highly similar units useful in the arena of political science? Do political scientists put more emphasis on the study’s comparability or representativeness?

Purported 聲稱 Profusion of 豐富 Purview 範圍 Ontological presupposition 本體論預設 Affinities 親和力 Pejorative 貶義 Breadth 寬度 Commensurate 相稱 Gauge 測量 Intuit 意會 Revelatory 啟示的 Superfluous 多餘 Scalar 純量 Conjectures 推測 Prima facie causal connection 表面看起來不經意的連接 Discern 辨別 Heuristic 啟發式 Prosaic 平淡 A corpus of 一個語料庫 XXX is a bane to sth XXX 是禍根 Mimic 摹擬 Quiescent 靜 Time-honoured 老字號 question Presuppositions 前提 Vetted 審核 Tangentially 切線 Cognizant 認識 Disjuncture 脫節 Penumbra 半影 Conflated with… 與...混淆

Questions

Q1. According to Gerring (2004), there are two main types of research: case study and multipleunit study, the former one is conducive to theory generating whereas the latter is more useful in theory confirmation. Is it necessary for a researcher to propose and test one’s theory through engaging in both case study and cross-unit study so that his/ her theory can be a confirmatory one? Which kind of study is more important or popular in modern political science scholastic world, theory generating or theory testing? Would theory generating become relatively scarce as time passes since more prominent theories have been established to explain social phenomena, and researchers would rather focus on testing the falsifiability of theories by applying it to more units? Q2. According to Gerring (2004), when there are not any other useful variance across units, the case study method is considered as the only way to provide evidence of a general proposition. Yet, case study has the drawback of lacking representativeness. In this case, should the researchers investigate a topic with minimal examples (i.e. maybe only one) available for study? Or should they change to another topic which may provide a profusion of units for them to compare and study to enhance the study’s generalizability and credibility? If researchers insist on conducting case studies with few examples, how can they ensure that their studies can become authoritative in the future? Q3. According to Gerring (2004), when adjacent units are thought to be entirely non comparable, the case study method is impossible, whereas if the units are highly similar to each other, it is said that nothing can be learned about one unit by studying another. However, if analogous units possess identical characteristics, they may also be useful for strengthening the study’s representativeness. To what extent are highly similar units useful in the arena of political science? Do political scientists put more emphasis on comparability or representativeness when conducting a research?...