Critical Realism

This chapter evaluates critical realism, a term which refers to a philosophy of science connected to the broader approach of scientific realism. In contrast to other philosophies of science, such as positivism and post-positivism, critical realism presents an alternative view on the questions of what is ‘real’ and how one can generate scientific knowledge of the ‘real’. How one answers these questions has implications for how one studies science and society. The critical realist answer starts by prioritizing the ontological question over the epistemological one, by asking: What must the world be like for science to be possible? Critical realism holds the key ontological belief of scientific realism that there is a reality which exists independent of our knowledge and experience of it. Critical realists posit that reality is more complex, and made up of more than the directly observable. More specifically, critical realism understands reality as ‘stratified’ and composed of three ontological domains: the empirical, the actual, and the real. Here lies the basis for causation.

Mixed Methods

This chapter examines mixed-methods research, which is an approach that involves the integration of quantitative and qualitative methods at one or more stages of a research study. The central idea behind mixed-methods research is that the intentional combination of numeric-based methods with narrative-based methods can best provide answers to some research questions. The ongoing attempts to construct a simple and common conceptualization of mixed-methods provide a good indicator of the status of mixed-methods itself. mixed-methods research has emerged as a formalized methodology well suited to addressing complex problems, and is currently applied throughout the social sciences and beyond. Nowadays, researchers interested in combining quantitative and qualitative methods can benefit from the growing knowledge about the epistemological foundations, essential considerations, and rigorous designs that have been advanced for mixed-methods research.

D. Deductive, Inductive, and Retroductive Reasoning

This chapter looks at deduction, induction, and retroduction, which are three forms of reasoning that explain observations or develop new explanations from observations, by connecting sentences to a logical structure. Deduction explains individual occurrences of a phenomenon based on general sentences (laws) and respective circumstances. Induction derives general sentences (laws) from repeated observations of similar events. Retroduction, also often referred to as ‘abduction’, is an educated guess about the likely explanation for an observation, which can then be tested. The purpose of applying these forms of reasoning to observational studies is to make logic an explicit tool that applies extant knowledge, or develops new knowledge. While deduction applies extant knowledge, induction and retroduction develop new knowledge. The basic structure of all three forms of reasoning is derived from classical syllogisms (arguments), i.e. a structure in language that combines sentences (premises) to a conclusion. The chapter then considers examples of scientific work that applies the three forms of reasoning.

Concept Construction

This chapter highlights concept construction. All social sciences research projects, be they qualitative or quantitative, are dependent on concepts. The chapter first explains what concepts are and why social scientists should be self-conscious in the way they use them. It then describes the methodology of concept construction and presents three different ways to structure a concept. Finally, the chapter provides criteria to evaluate the quality of the concepts we have built ourselves or borrowed from others. Concept construction involves two basic operations beyond choosing a term to designate the concept: identifying the fundamental characteristics of the phenomenon of interest, and logically connecting these characteristics.

M. Meta-Analysis

This chapter illustrates meta-analysis, which is a specific type of literature review, and more precisely a type of research synthesis, alongside traditional narrative reviews. Unlike in primary research, the unit of analysis of a meta-analysis is the results of individual studies. And unlike traditional reviews, meta-analysis only applies to: empirical research studies with quantitative findings hat are conceptually comparable and configured in similar statistical forms. What further distinguishes meta-analysis from other research syntheses is the method of synthesizing the results of studies — i.e. the use of statistics and, in particular, of effect sizes. An effect size represents the degree to which the phenomenon under study exists.

Typology

This chapter assesses typology. The importance of typologies is contested. Some scholars view them as fundamental to concept construction, while others consider that they are temporary devices at best and actually discourage their use. The chapter focuses on the less problematic, heuristic roles of typologies. In this respect, typologies are a proven and widely used instrument for organizing knowledge and ideas at various stages of research, and especially for jump starting the process of generating hypotheses. Moreover, theories are often presented, e.g. for didactic purposes, by means of typologies. Before turning to the uses of typologies, the chapter differentiates between typology, classification, and taxonomy.

Multicausality and Equifinality

This chapter evaluates multicausality and equifinality, which refer to a research situation whereby an outcome is explained by more than one causal factor. The term ‘equifinality’ stems from systems analysis, and refers to a situation in which ‘the same final state may be reached from different initial conditions and in different ways’. ‘Equifinality’ also appears in related disciplines such as psychology, archaeology, or environmental studies, while ‘multicausality’ is often used in literature on social science methodology. Ultimately, multicausality and equifinality are important reasons why social phenomena are particularly challenging to study. Multicausality and equifinality are often explored by research traditions and methods of social inquiry that approach causal processes through a ‘causes-of-effects’ approach, rather than an ‘effect-of-causes’ approach.

Literature Review

This chapter highlights literature review. Reviewing the published literature is one of the key activities of social science research, as a way to position one’s academic contribution, but also to get a bird’s eye view of what the relevant literature says on a given topic or research question. Many guides have been created to assist academic researchers and students in conducting a literature review, but there is no consensus on the most appropriate method to do so. One of the reasons for this lack of consensus is the plurality of epistemological attitudes that coexist in the social sciences. Before initiating a literature review, the researcher should start by clarifying the need for and the purpose of the review. Once this has been clarified, the actual review protocol, tools, and databases to be used will need to be determined to strike a balance between the scope of the study and the depth of the review.

Discourse Analysis

This chapter describes discourse analysis. In linguistics, discourse is generally defined as a continuous expression of connected written or spoken language that is larger than a sentence. However, as a method in the social sciences, discourse analysis (DA) gave rise to diatribes about where to set the borders of discourse. As language constitutes the very entry point to the world, some discourse analysts argue that all that exists acquires meaning through language. Does this mean that discourse constitutes reality? Is there anything outside text and discourse? Or is discourse one among many means of social construction? The evolution of DA in social science unearths an ontological debate between ‘realists’ and ‘nominalists’, which eventually reverberates in epistemological strategies.

Time Series

This chapter focuses on time series analysis, a statistical method of longitudinal analysis which is suitable if researchers are interested in the temporality of social phenomena and want to analyse social change and patterns of recurrence over time. In contrast to other statistical methods of longitudinal analysis, time series analysis can be applied even if researchers have only a few cases (maybe even only one) and only a few (maybe even only one) variables. Time series can be built for any level of analysis, as cases can be persons, but are usually organizations or countries. In order to build a time series, the variables need to have been measured several times over a given period, and for each measurement one needs to know the measurement date. There are different goals when doing time series analysis, which can be used in descriptive, explanatory, and interpretive approaches.