An illustrated graduate-level introduction to case study research, including qualitative comparative analysis (QCA), also known as configurational analysis, and covering information metrics methodology.
NEW IN THE 2016 EDITION
LIST OF ADVANTAGES OF QCA FOR CASE STUDY RESEARCH1. It is useful for causal explanation, leading to theory development. 2. QCA supports theory testing by determining if expected patterns conform to generalizations based on observed cases. QCA also supports policy decision-making by uncovering both necessary and sufficient conditions needed for a desired outcome. 3. It handles various types of causality, including asymmetric causality and multi- conjunctural causality (where the causal effect reflects a combination of multiple causes - a "causal package" - rather than a single cause). 4. Where correlational approaches tend to dismiss low-correlated causal variables, the set-theoretic asymmetrical approach of QCA can identify strong set theoretic relationships involving limited subsets of case. Where correlative studies may yield null findings, it is possible QCA may identify strong patterns involving the same variable and even situations where presence of the variable is important in one causal pattern while absence of the variable is important in a second causal pattern. 5. QCA incorporates counterfactual reasoning. The complex, parsimonious, and intermediate solutions of QCA in effect engage in "thought experiments" which treat "remainders" (logical causal patterns with no observed cases) differently, either excluding them (complex solution), including those which simplify the solution (parsimonious solution), or include those which simplify the solution and which are consistent with researcher-specified causal assumptions (intermediate solution). 6. QCA promotes recognition of equifinality (multiple causal paths leading to the same outcome). QCA is thus particularly salient when the researcher has reason to think that more than one causal path may lead to an outcome of interest and that the effect of any given condition may only come to light when that condition is associated with certain other conditions. 7. QCA involves causal asymmetry and often will show that the causal conditions associated with presence of the outcome are different from the causal conditions associated with the absence of the outcome. Multiple causal configurations may be associated with either high or with low membership in a given outcome state. Understanding multiple asymmetrical causal relationships often is more helpful to both theory building an policy analysis that are quantitative "percent of variance explained" solutions. 8. As a corollary, models in QCA need not assume unidirectional causality or be recursive as is required for most models using standard quantitative modeling techniques. 9. For ordinal and interval data in fuzzy set QCA, data are calibrated rather than using the usual raw data usual in quantitative social science. Calibration incorporates external standards and external knowledge about cutting points for inclusion and non-inclusion of cases in membership in an attribute, and about cross-over (tipping) points for likelihood of membership or non-membership. Unlike uniform methods of measurement in correlational analysis, calibration in QCA may be different for different cases as appropriate based on external knowledge). 10. Where correlational methods may dismiss a causal variable if its additive net effect is small when previously-entered variables are controlled, QCA takes a more holistic approach which shows how combinations of causal variables work in sets. 11. Generalization is possible whether the number of cases is small, medium, or large. 12. QCA lends itself to set-theoretic graphical analysis, in which Venn-type diagrams are extended to include causal arrows representing "necessary" and "sufficient" causal conditions. 13. QCA affords one useful basis of analysis in triangulation and multi-method approaches, which are often thought to be more valid than monomethod research designs. 14. QCA can be used to supplement quantitative research, as by identifying variables to include in a quantitative model or by spotlighting mediating variables.
The full content is now available from Statistical Associates Publishers. Click here.
Below is the unformatted table of contents.
CASE STUDY ANALYSIS & QCA Example datasets 6 Overview 6 Key Concepts and Terms 7 Types of case studies 7 Research designs for case study research 9 The grounded exploratory design 9 Counterfactual comparison design 10 Experimental single case research design (SCRD) 11 Cross-sectional comparison design 11 Longitudinal comparison design 12 Dynamic comparison design 12 Case selection strategies 13 Data collection 16 Data collection methods 16 Linear or iterative? 18 "Thick description" 18 Constant comparison 22 Case quality control 23 Pattern matching 24 Overview 24 Congruence testing 25 Explanation-building 25 Terminating data collection 25 Process tracing 27 Overview 27 Controlled observation 27 Time series analysis 28 Critical incident technique (CIT) 28 Meta-analysis 29 Overview 29 Implementation considerations 31 Quantitative meta-analysis of cases 33 Example: DOE (2010) 35 Configurational (QCA) analysis 37 Overview 37 Types of QCA 42 Coverage vs. consistency 42 Dealing with inconsistency and contradictions 46 QCA: Limits, criticisms, and assumptions 47 fsQCA software 49 Crisp set QCA 50 Crisp set QCA example 1 50 Crisp set QCA example 2 67 Crisp set QCA example 3 88 Longitudinal pattern matching 91 Research design considerations 91 Fuzzy-set QCA 96 Overview 96 Fuzzy set QCA example 1 97 Fuzzy set QCA example 2 109 Internal structural heterogeneity index (ISHI 110 Qualitative and Multi-Method Research periodical 112 Information metrics / structured-focused comparison 113 Data 113 Overview 113 Entropy, conditional entropy, and uncertainty 114 Information metrics example 115 Statistical packages for the uncertainty coefficient 122 Information metrics complements QCA 128 Assumptions of case studies methods 131 Frequently Asked Questions 131 What are common standards for case studies based dissertations? 131 Is Institutional Review Board/Use of Human Subjects approval necessary for case study research? 133 Is case study research a social science substitute for scientific experimentation? 133 Aren't case studies unscientific because of researcher bias? 133 Aren't case studies unscientific because findings cannot be generalized? 135 Is a "case history" the same as a "case study"? 135 Are all case studies forms of case study research? 135 Is a joint or team approach to case study research preferred over single-investigator research? 136 What is a "consortium benchmarking" case study? 136 How is case study research related to complexity science? 137 What is nVivo? 137 What is xSight? 138 What is Leximancer? 138 What type of sampling is required in QCA? 138 How many predictor variables can QCA handle? 139 How does calibration work in fsQCA? 140 What are "remainders" in QCA? 143 How does the QCA algorithm simplify the truth table? 144 How is QCA related to cluster analysis? 146 How are hierarchical data handled in QCA? 147 What support exists for QCA? 148 What QCA software is available? 149 Bibliography 151 Acknowledgments 162 Page count: 164