Studying Complex Interactions and Outcomes Through Qualitative Comparative Analysis

A Practical Guide to Comparative Case Studies and Ethnographic Data Analysis

Markus Kröger

• Publisher: Routledge
• Year: 2021
• eBook ISBN: 9781000412833
• Edition: 1
• Topic: Psychology
• Subtopic: Research & Methodology in Psychology
• Index: Psychology

Studying Complex Interactions and Outcomes Through Qualitative Comparative Analysis: A Practical Guide to Comparative Case Studies and Ethnographic Data Analysis offers practical, methodological, and theoretically robust guidelines to systematically study the causalities, dynamics, and outcomes of complex social interactions in multiple source data sets. It demonstrates how to convert data from multisited ethnography of investment politics, mobilizations, and citizen struggles into a Qualitative Comparative Analysis (QCA).

In this book, Markus Kröger focuses on how data collected primarily via multisited political ethnography, supplemented by other materials and verified by multiple forms of triangulation, can be systematically analyzed through QCA. The results of this QCA offer insight on how to study the political and economic outcomes in natural resource conflicts, across different contexts and political systems. This book applies the method in practice using examples from the author's own research. With a focus on social movement studies, it shows how QCA can be used to analyze a multiple data source database, that includes results from multiple case studies.

This book is a practical guide for researchers and students in social movement studies and other disciplines that produce ethnographic data from multiple sources on how to analyze complex databases through the QCA.

1
Systematic data collection and analysis

Multi-sited political ethnography, QCA, and other methods for assessing complex causalities

Chapter abstract: This chapter explains what Qualitative Comparative Analysis (QCA) is and how it can be used and combined with other methodological approaches. The chapter also explains how I conducted the field research that provides the illustrative data for this book. This transparent guide to the fieldwork approach aids in demonstrating how QCA can be utilized from the first moment of fieldwork through to the final data analysis. This will be particularly helpful for those struggling with the selection of proper methods, when to use QCA, and how to justify the use of QCA. Detailed tables, which uncover the protocols through which the causal conditions and factors are identified as being active or not, are offered as guidelines on how these tables should be crafted to make the analysis transparent and replicable. In this chapter, there is a discussion of when to use QCA in its different variations (e.g., crisp-set, fuzzy-logic, multi-value, temporal) and how to identify when it is not a suitable method. The limitations of this approach are explored, and insight is offered on how these limitations can and should be addressed.

QCA guidebooks, tools, and aims

Traditionally, QCA has been primarily seen as the best tool for case-oriented studies that have a moderate number of cases, what is called a medium-n study. However, the uses of QCA have expanded, through the introduction of new techniques, to include studies with more cases (larger n) used alongside studies with fewer, more detailed cases (small n) (Rihoux 2020). The use of QCA has even expanded beyond academia to policy-making circles, where it can be used to perform meta-analyses of projects, policies, and other cases observed in real-life settings (Rihoux 2016). This book will use a medium-n example to show how to use QCA across a database with 23 total cases in order to assess the possible causal linkages between 22 causal conditions and several political and economic outcomes. It should be noted that these causal conditions can also be called explanatory factors, and they can also include strategies.

I recommend reading this book in parallel with a more general QCA guidebook. The most recent (and excellent) guide on how to design and apply QCA studies is Mello’s (2021) Qualitative Comparative Analysis: Research Design and Application. Kahwati and Kane’s (2019) Qualitative Comparative Analysis in Mixed Methods Research and Evaluation focuses on helping students to apply QCA in mixed-methods research designs and contributes to mixed-methods textbooks. A useful complement to this mixed-methods textbook is the chapter by Rihoux, Concha, and Lobe (2021), which addresses more specifically how to utilize multiple case studies. These discussions are linked with the broader methodologies of process tracing, which aim to uncover causal mechanisms and the kind of empirical evidence that could be used to look for them, and also linked with how to sequence QCA with process tracing (Beach and Pedersen 2019; Schneider and Rohlfing 2013). The discussions here are part of this broader analysis of processes and causal mechanisms: In particular, in this project I was interested in identifying and following the key causal mechanisms and paths that might explain how resistance might impact global natural resource flows. My ethnographic approach was based on process tracing in the sense of not focusing on the broadest possible set of things to observe (as in traditional ethnography in a single place), but instead tracing the process of the global commodity boom of 2005–2015 and observing when and how this was possibly resisted. This constituted a multi-site, process-tracing exercise on two key processes: The pushing and resisting processes. Later, I divided this focus into more specific mechanisms, such as resistance strategies, observing the processes through which these, and especially their combinations, might have been causally linked to different kinds of outcomes. I found QCA to be an excellent tool to utilize for this fine-grained process tracing, which takes place later on in this book.

Like other broad guidebooks that deal with QCA, the previously mentioned books primarily focus on introducing the methodology and explaining it step-by-step. Other useful books of this nature include Configurational Comparative Methods: Qualitative Comparative Analysis (QCA) and Related Techniques, edited by the pioneers of QCA methodology, Rihoux and Ragin (2009). These contributions all are substantially different from my book and the work on which it is based. To really master the QCA method, I recommend reading these more abstract, introductory books, which contain some examples of application, in addition to this book, which focuses on how the method is applied in practice. One should also always stay informed about the latest summarizing review articles and book chapters on the subject. For example, Rihoux (2020) provides an updated review of the state-of-the-art in QCA research, while Thomann and Maggetti (2020) offer valuable advice on how to design QCA-based studies.

The QCA method has also been criticized from a quantitative methods perspective, as well as from other viewpoints [see the Symposium in Sociological Methodology 44(1)), with contributions from Ragin, Collier, and many others]. Vaisey (2014) and Ragin (2014), among others, provide good answers to these critiques, showing their many weaknesses and misunderstandings. Most of these critiques are based on assumptions that approach QCA from the logic of conventional quantitative methods. A typical misunderstanding of QCA is related to it being a configurational method, not a method to study interactions between all factors. QCA solution terms refer not to interactions but to configurations (which is the concept that should be used): This refers to “the result of additive processes reaching a threshold” (Vaisey 2014). Most of the critiques against QCA have also been leveled against other methods that usually focus on finding parsimonious solutions or providing overtly complex algorithmic tools (Collier 2014).¹ Simplicity and intuitive reasoning are recommended across techniques, rather than using very complex and complicated methods or computational tools if there is no apparent or clear need. These methods and tools should only be introduced later after careful consideration of whether they are truly needed. This decision should also include a consideration of how much extra time and effort would be needed to master and utilize more complex methods and tools and what would be gained from their implementation. David Collier (2014) recommends that QCA users abandon complex algorithm tools and simplify the method. I follow his advice here; I do not make much use of the typical, complex QCA algorithm tools but instead base my analysis on very rich and detailed, intimate case knowledge. Ragin (2014) also recommends this approach as a better solution than the use of invented counterfactuals, if one has knowledge available on the studied cases. Besides my book, the importance of deep knowledge of cases, understood as complex and dynamic wholes, and their comparison via QCA are also presented by Gerrits and Verweij (2018). Their book, The Evaluation of Complex Infrastructure Projects: A Guide to Qualitative Comparative Analysis, contains an empirical and methodological approach that is akin and complementary to the focus on investment politics herein.

Normally QCA explores the role of approximately three to seven conditions, due to the issue of increasing limited diversity. The problem of limited diversity refers to a set-theoretic approach, wherein one does not want to have too many logical remainders, that is, causal condition complexes that are not covered by the QCA combinations, which are the cases with a specific set of explanatory and outcome factors (Rihoux and De Meur 2009). To accommodate this, I divided the 22 explanatory factors into different sets and observed the same outcomes for these different sets. These are both important steps, because in the typical progression of QCA one needs to first consider a larger array of explanatory factors and then focus on those factors that are found to be the most important. Marx and Dușa (2011) provide helpful advice and benchmarks for considering the number of case conditions to use in different QCA applications. The contribution here is to show how one set of conditions, focusing on movement strategies, can be juxtaposed with another set of conditions, which focuses on contextual and contingency factors, to see which set better explains the outcomes. I also mix between these sets of factors to look for combinations. I did this because political ethnography, and especially multi-sited ethnography, produces a myriad of different possible explanations and factors pertinent to those explanations that could matter in different cases. When one adopts QCA based on this type of field research, guidance is needed on how to apply the method when one needs to take complexity seriously to be able to draw meaningful conclusions. One cannot just see the world and attempt to explain it accurately via limited and very broad factors.

Normally QCA is not applied to ethnographically oriented data. Ethno-graphically oriented research tries to produce thick descriptions and understanding of complexities instead of parsimonious explanations on more generalizable mechanisms, which is more akin to typical QCA. However, I argue that QCA-type approaches can be used for making sense of complex, multi-sited, political ethnographic data. However, the key focus is not on trying to produce just the most parsimonious solution, but instead offering a tool for systematic data handling, comparisons, and drawing probabilistic findings from complex data sets. Thus, this book partakes in broadening the uses of QCA, which follows the trend in the field, specifically in the expansion of the use of more probabilistic procedures (Rihoux 2016).

Parsimonious explanations can also be pursued, but I will show how this is not the only way QCA should be applied. I provide a contribution to the bulk of QCA usages, especially by showing how medium-n databases across an abnormally large number of causal conditions (over 20) and outcomes (over 5) can be made sense of via an approach that is primarily and initially probabilistic, yet is in sequence parsimony-seeking.

The prior usages of QCA have based their logic mostly on set-theoretical and particular logical inferences, for example, through the usage of Boolean logic. This discussion opens up the prospect that one should take into consideration even logically possible combinations that are not present in the actual data sets, what are called logical remainders (Rihoux and De Meur 2009). These logical remainders are useful for making further analyses, but QCA users normally try to avoid having too many of them by limiting the number of conditions, as having too many conditions leads to many individualized case explanations. However, I will show, through multi-sited ethnography data that are reported via QCA tables, that one cannot and should not try to steer reality too much toward more parsimonious explanations that go beyond the nuance of the actual reality. Instead, one should use the tool in these kinds of data sets and situations for systematic comparison in order to identify the possible generalizable dynamics. There are necessary and sufficient causal condition complexes and factors that can be uncovered through a real-world database across 22 conditions and 23 cases. I would also like to note that the utilization of invented case scenarios is not recommended or necessary in analysis that is based on a limited data set of observed, real-world scenarios that are already sufficiently complex. However, I will show how the guidance in existing QCA technical guidebooks (e.g., in Rihoux and De Meur 2009, see box 3.6) on how to resolve contradictory configurations and rework the conditions and their sets is helpful and can be used in practice.

Specific technical tools can be used and are recommended when conducting QCA, especially across a database as vast as the one I used in my research. I used the Excel QCA add-in developed by Lasse Cronqvist (2019) to check my results and each subtable, which I obtained by first manually analyzing the tables and going through each combination. I highly recommend this Excel tool, which is free, open-access, and easy to use. It provides help with double-checking the truth table findings and determining the implicants (the cases and their factors that produced the outcome studied) and solutions (the different truth formulas). In addition, the tool is useful as it allows for checking what the analysis would be in cases where logical remainders are included. It also allows for quickly checking what would explain the 0 outcomes. When I refer to the “Excel tool” in this book, I refer to this QCA tool.

Thiem and Dușa (2013) discuss the issue of whether to use parsimonious or intermediate solution terms in QCA, which currently divides opinion. I show how one can pursue both and how even individualized accounts do serve a purpose. Most importantly, I put much more value on a probabilistic approach, wherein I observe, for example, that a set of conditions was present in most of the cases. There are so many contextual and contingency differences in the real world—especially when comparing across different cases in different parts of the Brazilian and Indian realities—that one cannot assume a parsimonious explanation can be found. Instead, I show how particular sets of conditions explain a particular set of cases, depending on the context, for example, civil war or peaceful setting. The Excel tool is helpful as it can be used to quickly go through different sets of case combinations and condition sets, selecting these based on contextual and other limitations that surface.

There are other tools that can be used in QCA and publications that offer guidance for their usage. For example, Thiem and Dușa (2013) focus on offering advice on how to use QCA with the R software package for statistical computing. They discuss how to use crisp-set, multi-value, and fuzzy-logic versions of QCA in R software, with the assumption that the reader already knows the fundamentals of QCA and R. In comparison, this book will offer guidance and demonstrate how one can perform good QCA without needing to learn software that is typically too complex or complicated for noninstructed users. This type of approach to QCA has been called for even by those who take a more critical stance toward the method: For example, Collier (2014), does not favor adopting the new algorithms and programs designed to study QCA tables, but instead prefers to refocus attention on case-based research, process tracing, and other conventional qualitative methods, which should be the primary method. This book answers the call for simpler approaches that show how QCA can be used. I demonstrate how necessary and sufficient causal condition complexes can be identified through Excel-based tables that are divided into subsets of truth tables that are easily read and interpreted. These tables are opened up and analyzed in this book, showing how they can be unpacked in practice.

On one hand, the added value of multi-sited ethnography here is to provide a broader comparison than is typical and not a detailed case study, which already abound in the resistance and mining literature. On the other hand, the usage of csQCA, and the several case studies that are compared, signifies that more abstraction and generalization are required to allow for conceptualization that captures the heterogeneity across the cases and also allows for a comparison of recurring dynamics across the contexts.

Writing a methodology section on multi-sited ethnography and QCA

Each research project should produce a description of the methodology used and justifications for the choices made. What follows is a concrete example of how that methodology section should be written, based partially on the methodological annex of Iron Will: Global Extractivism and Mining Resistance in Brazil and India (Kröger 2020). I wrote Iron Will using a subset of the empirical materials that I use in this book to support my examples. While this book focuses on how to the use the method, the focus in Iron Will is the analysis resulting from use of the method. The text that follows can be used as an example of how to justify the methodological choices of QCA and multi-sited political ethnography.

The data that make up the concrete examples used in this book were collected utilizing multi-sited eth...

Table of contents

• Cover
• Half Title
• Title
• Copyright
• Contents
• List of maps
• List of tables
• Acknowledgments
• List of abbreviations
• Introduction
• 1 Systematic data collection and analysis: multi-sited political ethnography, QCA, and other methods for assessing complex causalities
• 2 Studying complex interactions and outcomes: selecting key factors in social movement studies
• 3 How to apply QCA? An illustration through a QCA of investment politics outcomes in Brazil and India
• 4 Complementing QCA by a detailed opening of explanatory factors
• Conclusions
• References
• Index