Mixing Methods to Explain Emigration Policies: A Post-Positivist Perspective

Abstract

Growing numbers of researchers are using mixed methods to study migration, often highlighting the practical reasons connected with policy engagement. However, in this article we emphasize epistemological and theoretical rather than purely practical reasons for using mixed methods in the study of migration. Specifically, we argue that mixed methods designs are well suited to research that attempts to explain sociopolitical action within a post-positivist epistemological framework. We provide an example of this approach in the Diaspora Engagement Policies Project, a 5-year project to explain the global proliferation of formal government institutions for emigrants and their descendants.

Keywords

diaspora engagement policies diaspora institutions post-positivism explanatory theory mixed methods research

As this special issue shows, growing numbers of researchers are using mixed methods to study migration. Researchers often highlight practical justifications connected with reaching and engaging diverse audiences of scholars, practitioners, and policymakers. However, researchers have written surprisingly little about the epistemological and theoretical reasons why mixed methods may be suitable for the study of migration. In this article, we address this gap in the literature. We outline our understanding of post-positivist epistemology and show how it opens the door to mixed methods approaches. As an example, we present the Diaspora Engagement Policies Project: research that used mixed methods to pursue a post-positivist–style explanation of sociopolitical action within the broader field of migration studies. Our central argument, threaded throughout the article, is that mixed methods research designs are well matched to migration studies projects informed by the post-positivist worldview. We contend that this is because post-positivist research emphasizes explanation and such explanation requires both description of general patterns of social action (which is best achieved through quantitative methods) and making sense of what such actions mean to the social actors involved (which is best achieved through qualitative methods). This account of the fit between mixed methods and post-positivist epistemology runs counter to the widespread notion that post-positivism is merely a “soft” version of positivism linked primarily to quantitative approaches.

The Match Between Post-Positivism and Mixed Methods Research

We used a mixed methods research design to study diaspora engagement policies because such a design fit with our underlying post-positivist worldview. Below, we explain what post-positivism is and why we think it is such a good fit with mixed methods research.

Post-positivism is a relatively recent arrival on the epistemological scene, but one with extensive philosophical roots (Phillips, 1983). Many central tenets of post-positivism, such as critical self-reflection and suspension of judgment about the existence and nature of true reality, are consistent with aspects of skeptical philosophy stretching back to the Ancient era (see below). However, as a distinctive philosophical perspective, post-positivism arose during the 20th century (Scotland, 2012), and it has since been used everywhere from geography (Sharp et al., 2011; Smith, 1979) to the health sciences (Henderson, 2011; Petty, Thomson, & Stew, 2012), to language education (Scotland, 2012), to legal theory (Bustamante, 2011; Petroski, 2011). Post-positivism is now one of the more widespread worldviews informing social science research, including research in the field of migration studies.

Philosophical Roots of Post-Positivism

Various disciplinary debates have led to quite different understandings of the term post-positivism, which therefore deserves more detailed discussion. Positivism itself is an empiricist–foundationalist epistemology. Foundationalist epistemology assumes that claims to knowledge have a “secure foundation,” but different types of foundationalism disagree on what this foundation is. Rationalist foundationalism assumes that secure knowledge claims are based on rational thought, a view encapsulated in René Descartes’ proposition, cogito, ergo sum (I think, therefore I am—or in a version closer to the original expression, “I doubt, therefore I think, therefore I am” (Descartes, 1644/1984, p. 6). By contrast, empiricist foundationalism assumes “the mind to be,” as John Locke put it, “white paper void of all characters . . . furnished . . . [with] all the materials of reason and knowledge . . . [by] EXPERIENCE . . . [upon which] all our knowledge is founded (Locke, 1690, as quoted in Phillips & Burbules, 2000, p. 6). As an empiricist–foundationalist position, positivism thus assumes that knowing something entails a sensory experience of it: Knowledge, in the positivist worldview, is fundamentally empirical.

Historical Roots of Post-Positivism

Positivism is often associated with the realist premise that one true reality exists and can be apprehended through empirical observation. However, positivism is not synonymous with realism. At the empiricist extreme, epitomized by the so-called logical positivists of interwar Vienna, positivism objects to a theory of ultimate reality as meaningless metaphysics: What cannot be observed cannot be said to exist, so scientific inquiry should be limited to the study of observable behavior (Crotty, 1998; Phillips, 1983). As the positivist founding figure Auguste Comte put it, scientific knowledge “gives up the search after the origin and hidden causes of the universe and a knowledge of the final causes of phenomena” (Comte, 1970, as cited in Phillips & Burbules, 2000, p. 8). Nor is positivism synonymous with quantitative methods, as is sometimes claimed. Positivists may rely on qualitative data, just as nonpositivists may rely on quantitative data (Phillips & Burbules, 2000). What defines positivism is not the form of observational data it depends on but, rather, its claim that the foundation of all knowledge is observation as opposed to inference.

In the broadest sense, post-positivism denotes everything that came after positivism (e.g., see Petty et al., 2012; Scotland, 2012). However, post-positivism is not a sympathetic, quantitatively focused successor to positivism; it is a series of critiques of positivism (Fox, 2008; Johnson & Gray, 2010; Phillips & Burbules, 2000). One strand of critique informing post-positivism derives from Max Weber’s assertion that social reality needs to be understood from the perspective of the subject rather than explained objectively—a perspective encapsulated in the concept of verstehen or “understanding” (Tucker, 1965). Another strand has roots in the work of Karl Popper (1934/2002), who argued that scientific discovery does not result in true statements but rather in theories: best guesses at the nature of reality, which may be rejected when previously unobserved information comes to light (c.f. Phillips & Burbules, 2000). Post-positivism also takes inspiration from Thomas Kuhn (1962/2012), whose notion of “paradigm shifts” in scientific thinking challenged the positivist premise that science is an objective pursuit of universal truth. This worldview allowed scientific activity to be treated as social activity embedded within social institutions and processes and therefore subject to the same sudden and irrational changes.

Just as positivism emerged during the European Enlightenment of the 17th and 18th centuries alongside the modernist rejection of faith and tradition in favor of rationality and order as sources of knowledge, post-positivism emerged alongside the profound disillusionment at the complicity of rationalism and empiricism with the irrational horrors of 19th-century colonialism and 20th-century totalitarianism. As Phillips and Burbules (2000) note,

“Postpositivism” is not a happy label (it is never a good idea to use a label that incorporates an older and defective viewpoint) but it does mark the fact that out of the ruins of the collapsed positivistic approach, a new (if diverse and less unified) approach has developed. (p. 4)

The Range of Post-Positivisms

The range of post-positivisms emerging from these critiques of positivism vary along a spectrum from “realist post-positivism” to “constructivist post-positivism” (Fox, 2008). Whereas realism (as mentioned above) is associated with the belief in an ultimate reality of which all sense impressions are an approximation, constructivism is associated with the view that reality is constructed in the mind of the observer (c.f. Schwandt, 1994). At one extreme, realist post-positivism accepts positivist ontology but rejects positivist epistemology: It asserts the existence of a single reality but rejects the notion that science can provide objectively true statements about this reality. Instead, this reality can only ever be experienced through imperfect filters of observer bias, including observers’ sociocultural frames of reference (Henderson, 2011; Maxwell & Mittapalli, 2010). The goal of realist post-positivism is therefore to minimize bias in the formation of theories that, although never 100% accurate, accord better with the facts (Petty et al., 2012). In these respects, realist post-positivism shares premises and objectives with so-called critical realism (c.f. Maxwell & Mittapalli, 2010). At the other extreme, constructivist post-positivism rejects both the ontology and the epistemology of positivism. From this perspective, there is no single objective reality but as many subjective realities as there are individuals and/or social groups (Blaikie, 1993; c.f. Phillips & Burbules, 2000). Truthful statements about reality are therefore relative, not absolute; and the focus of social science should be to interpret the meaning of social life to the social actors who create social reality through their interactions (Scotland, 2012).

It would be wrong to reify the various labels listed above. Rather than monolithic, unitary doctrines shared by homogeneous research communities, positivism and the various forms of post-positivism are, like all philosophical stances, constellations of ideas used in different combinations by different researchers as heuristic toolkits (Maxwell & Mittapalli, 2010). For example, post-positivists may be open but not committed to both positivist and constructivist ontology, seeing the notion of a single true reality as yet another best guess, open to falsification (Phillips & Burbules, 2000). Such a stance draws on Popper (1934/2002) and Kuhn (1962/2012), as well as the wider skeptical principles of critical self-reflection and suspension of judgment about the existence and nature of reality. This stance has roots in a philosophical tradition stretching back through the work of Descartes to that of the Platonic Academy under Arcesilaus (from ca. 264 BCE) and his successors up to Antiochus of Ascalon (from 90 BCE) (Adamson, 2014). Moreover, despite their ontological differences, the realist and constructivist versions of post-positivism have much epistemological overlap. For example, both see the attainment of social knowledge as a process of interpreting and understanding the subjective frames of reference of social actors, and in this sense, both bear the influence of interpretivism—the view that social research is about interpreting how social actors go about socially constructing their own realities (c.f. Schwandt, 1994).

Post-Positivism’s Emphasis on Explanation

The aspect of post-positivism we wish to foreground in this article is its emphasis on explanation. For some positivists (especially those associated, sometimes inaccurately, with David Hume and therefore often called “Humean”), the goal of science is to observe reliable patterns of physical and social behavior, not to engage in metaphysical speculation about why such patterns occur (Lorkowski, 2015). For post-positivists, much as for realists, this behaviorist approach is inadequate: A previously reliable pattern may be coincidental and therefore prone to change, unless one can identify a good reason why it should not—that is, a causal explanation (Maxwell & Mittapalli, 2010; Phillips, 1983). Conversely, some forms of interpretivism are more concerned with describing, understanding, and interpreting a complex social world than in identifying why it apparently exists (Blaikie, 1993; Petty et al., 2012). For post-positivists, this too is inadequate because it gives up entirely on the idea of a single reality to settle for the assertion of subjective beliefs. Thus, although they fall at different ends of one epistemological spectrum, positivism and interpretivism share an opposition to explanation. Positivism rejects the inference on which explanation depends, instead favoring “pure” observation with a view to reliable prediction. Meanwhile, interpretivism asserts that explanation is not possible because neither the object nor the subject of observation is real. In contrast to both worldviews, post-positivism emphasizes the role of theory in explaining, rather than merely predicting or interpreting, social action (Phillips & Burbules, 2000). Put differently, a central goal of post-positivist social science is to build theories that explain rather than just describe social reality.

Importantly, however, explanation does not mean the same thing to all post-positivists. In the natural sciences, explanation may involve uncovering the regular physical, chemical, and biological laws—“causal mechanisms”—that govern the natural and physical world (Cruickshank, 2011). However, social science deals not with physical laws but, rather, “social facts,” to use Émile Durkheim’s phrase: realities that are constructed through cognition and social interaction; as realist mixed methods writers have put it, the mind is a part of reality (Maxwell & Mittapalli, 2010). The causes of social phenomena therefore lie within the psychological states and sociocultural frames of reference that motivate and shape human behavior (Phillips, 1983). Moreover, because humans are capable of reflecting on and modifying their own behavior in accordance with what they know, theories of social science become part of these sociocultural frames of reference and themselves shape action, rather than remaining objective—in what Giddens (1987) called a “double hermeneutic” (p. 30). Therefore, from a post-positivist perspective, explaining why social and political actors establish particular institutional arrangements—such as the diaspora institutions we are about to introduce below—is a matter of observation and interpretation, and calls for a mixture of methods to address both of these objectives.

Why Post-Positivist Explanation Invites Mixed Methods Research Strategies

This methodological need for both observation and interpretation can be a call for mixing research methods. Quantitative methods suit the observation of large-scale patterns in data: By arranging observations into discrete categories and assigning a numerical value to each, researchers can compare the relative size and importance of social phenomena. However, quantitative methods rely heavily on probabilistic inference when attempting to derive causal explanation from correlative studies, assuming causal relevance if actions coincide reliably in the same sequence. This is especially problematic in the social sciences, where the actions of interest are those of thinking, feeling human beings, who are not locked into predictable behavior and may have reasons for behaving unpredictably in any given circumstances. These reasons may be quite unreasonable and need not be instrumentally rational or even value rational, as Weber put it; they may be entirely irrational. But the essence of explaining social action lies in understanding these unreasonable reasons: It is about making sense of the way people explain what they are doing and why. This is not possible using quantitative methods; it requires qualitative methods such as interviews to elicit and interpret explanations for social action from the social actors themselves.

Post-positivism also invites mixed methods approaches insofar as post-positivists tend to reject what they see as a false dichotomy between the positivist-quantitative and interpretivist-qualitative research paradigms (c.f. Philip, 1998; Sheppard, 2014). Post-positivism holds that research strategies should be driven not by commitment to one epistemological doctrine against all the others but by pragmatic concerns about effectiveness in generating knowledge or solving problems by applying an appropriate range of available quantitative and qualitative research methods (Henderson, 2011; Maxwell & Mittapalli, 2010; Phillips & Burbules, 2000). This approach to the different emphases of qualitative and quantitative research methods chimes with McKendrick’s (1999) comment that one’s “epistemological position only determines how methods can be used: it does not preclude the use of particular methods” (p. 40). In this view, even random clinical trials, seen as a holy grail by many quantitative, empiricist researchers, may make good use of qualitative, interpretative approaches (Plano-Clark et al., 2013). Moreover, much like critical realism, the post-positivist worldview is compatible with both quantitative and qualitative methodologies.

In sum, this article argues that mixed methods research designs are well suited to social science projects informed by the post-positivist worldview, because the post-positivist emphasis on explanation calls for both description of general patterns of social action, made possible by quantitative methods, and making sense of what such actions mean to the social actors involved. This account of the fit between mixed methods and post-positivist epistemology runs counter to the widespread notion that post-positivism is merely a “soft” version of positivism, linked primarily to quantitative approaches. However, even though mixed methods strategies are well suited to the post-positivist worldview, relatively little has been written about how this connection can work in practice. For example, as one anonymous reviewer of this article pointed out, a search for “post-positivist” and “post-positivism” across three key journals on mixed methods research (this journal, Quantity and Quality, and International Journal on Multiple Research Approaches) yields no hits. (An exception is Pirret, 2013.)

To address this gap in the literature, we document an example research project that used a mixed methods strategy informed by a post-positivist worldview. Again, our overall argument is that mixed methods strategies are uniquely capable of supporting research oriented toward social-scientific explanation as conceived in post-positivist terms. Our hope is that strengthening this connection might contribute to discussions about future directions in the development of mixed methods research (Hesse-Biber & Johnson, 2013; Johnson & Onwuegbuzie, 2004; Mertens, 2013, 2014; Tashakkori & Creswell, 2007) and, given the nature of our case study, to ongoing development of methodologies in the field of migration studies (c.f. Findlay & Li, 1999; Gamlen, Cummings, Vaaler, & Rossouw, 2013; Massey, 1987) and other emerging subdisciplines (c.f. Van der Roest, Spaaij, & Van Bottenburg, 2015).

Case Study: The Diaspora Engagement Policies Project

Above, we suggested that mixed methods research designs could provide a particularly powerful route to explanation. Below, we provide an example to support this argument, outlining the research strategy of the Diaspora Engagement Policies Project. This 5-year project integrated quantitative and qualitative methods to explain why states are forming institutions dedicated to emigrants and their descendants—a type of migration policy that was almost unheard of in the 1980s but is now found in well over half of all governments around the world and is being zealously promoted by the United Nations. In this project, we first observed correlative patterns through quantitative analysis, to gain a theoretically informed descriptive overview of how certain social actors—states—behave. We then worked qualitatively with specific senior state policymakers as research participants to understand what motivated and shaped their behavior. We discuss this research strategy in terms of overall design, sampling, data collection, analysis, validation, and integration of methods (see O’Cathain, Murphy, & Nicholl, 2008). Detailed discussion of methodologies and results is outside the scope of this article but can be found in Gamlen (2014a, 2014b, 2015a, 2015b), Délano and Gamlen (2014), Gamlen et al. (2013), and Gamlen et al 2017.

Brief Overview of the Research Project

The Diaspora Engagement Policies Project was a research initiative that used a mixed methods research strategy to explain how and why states establish formal institutions devoted to emigrants and their descendants in the diaspora. The project was motivated by the observation that, even though migration policy is often thought of as immigration policy, every immigrant is also an emigrant with ties to a state of origin, and origin states are increasingly proactive in managing these “diasporic” ties (Gamlen, 2006, 2008). In particular, we were struck by the phenomenal global spread of formal executive and legislative government offices responsible for diaspora populations—we call these offices “diaspora institutions” (Figure 1). Such institutions have rapidly become a regular feature of political life in many parts of the world (Agunias, 2009; Agunias & Newland, 2012; Gamlen, 2006, 2008). Examples of diaspora institutions include India’s Ministry of Overseas Indian Affairs, Serbia’s Ministry of Religion and Diaspora, Ireland’s Irish Abroad Unit within the Department of Foreign Affairs; and special electorates for overseas voters that exist in Italy, France, and several other states.

Figure 1.

The rise of diaspora institutions: Percentage of United Nations member states with formal offices for emigrants and their descendants, by institution type, 1980-2014.

Diaspora institutions provide empirical and theoretical surprises worth studying. In 1980, only a handful existed; but since then, such institutions have proliferated very rapidly, to the point where well over half of all UN member states now have one. Prior to 1990, most diaspora institutions were departments, usually located in the foreign ministry. These are still most common and are found in at least 100 states. But from approximately 1990 onward, the number of full government ministries dedicated to the diaspora has grown rapidly: There are now some 30 of these around the world. Diaspora institutions merit empirical research because of their surprisingly fast growth, both in terms of the number of states and organizations involved and in terms of their importance to wider political and policy processes.

Diaspora institutions are also worth studying because they are theoretically surprising. They project domestic policies beyond state borders in ways that seem inconsistent with the territorial definitions of citizenship and sovereignty that underpin the international system. They potentially interfere in the domestic affairs of foreign states by asserting responsibility for a population living within another government’s jurisdiction. They potentially undermine the principle of freedom of exit: Some emigrants may want to sever ties to their former homeland but can find themselves subject to the attentions of their homeland government through the mechanisms of diaspora institutions. In addition, by representing diaspora interests in legislative and executive processes at “home,” diaspora institutions potentially give influence over lawmaking to people living abroad who are not subject to the resulting laws. Given these deviations from the basic territorial norms of the modern international system, it is particularly surprising that diaspora institutions are being praised and recommended by international organizations in and around the United Nations as a brave new approach to international cooperation over migration.

The Diaspora Engagement Policies Project aimed to explain the remarkable rise of these diaspora institutions from a post-positivist perspective. As we argue throughout this article, that approach explicitly called for a mixed methods research strategy, in which we could identify and describe a reliable pattern of state behavior through quantitative research while also making sense of this behavior through qualitative interviews, in which we asked senior policymakers to explain what they were doing and why, in their own terms. For quantitative evidence, we collected and statistically analyzed new cross-national data on the establishment and maintenance of diaspora institutions across UN member states between 1939 and 2014. This stage of the research allowed us to achieve, apparently for the first time, a relatively complete historical and geographical overview of this rapidly growing but underresearched phenomenon. From our perspective, this approach exemplifies how mixed methods strategies may be useful for researchers with a post-positivist orientation working not only on migration but across the social sciences and health sciences more broadly.

Design

Our methodological choices were guided by a hypothetico-deductive theoretical framework, consistent with the post-positivist worldview. To generate our hypotheses, we began by dividing previous literature on diaspora institutions into two camps: a rationalist camp, which explained the emergence of state diaspora institutions as the product of strategic efforts by states to capture the material resources of emigrants and their descendants, and a constructivist camp, which saw diaspora institutions as expressions of the idea that nations are ethnic constructs unconfined by territorial borders. The rationalist approach implied hypotheses that diaspora institutions would tend to emerge in states with low per capita income, high receipts of migrant remittances, and high rates of brain drain through skilled emigration (Gamlen, 2014b; Gamlen et al., 2013). By contrast, the constructivist approach implied greater likelihood of their emergence in states that are ethnically homogeneous, authoritarian, and/or led by right-wing governments. The project also aimed to provide an alternative theoretical approach, informed by institutional theories such as “world society” theory (Meyer, 2010), to include a focus on the role of international organizations and experts in promoting the dissemination of new migration policy norms. Our central hypothesis was that international organizations were seeking to promote more multilateral responsibility-sharing over migration by encouraging states to form diaspora institutions. Accordingly, we hypothesized that diaspora institutions would tend to emerge in states with high levels of membership in migration-related international organizations, larger numbers of professional research papers on the national diaspora, and higher numbers of peer states with diaspora institutions. Our approach to deriving these hypotheses was consistent with the post-positivist approach of attributing a “reality” to social facts and seeking causal explanations. However, as we outlined above and discuss further in the context of our empirical study below, the demands of explanation within a post-positivist social science framework are quite specific in needing a phenomenological understanding of social action from the perspectives of the social actors involved.

We aimed to test these hypotheses through a mixed methods approach, in keeping with our understanding of the methodological implications of a post-positivist worldview, with a focus on quantitative methods for description and qualitative methods for interpretation (see the previous section). To this end, the project adopted an approach akin to what Creswell (2009) calls a “sequential explanatory” mixed methods research design.¹ This involved providing an overview of global patterns in diaspora policy making, informed by the collection and analysis of new quantitative data on state diaspora institutions. Recognizing that these quantitative methods are better at describing correlations than at explaining the behavior of the underlying social actors involved, we then examined these global quantitative patterns through in-depth qualitative interviews, fieldwork, and document analysis. Our aim was to explain—rather than merely describe—the establishment of specific institutions. The quantitative data allowed us to reveal large-scale patterns across all UN member states for a significant period—a task not attempted by previous studies. However, to confidently answer the central question of why states engage their diasporas, we could not simply look at what kinds of states did and did not implement specific types of programs; we also needed to speak with the people involved in the establishment and operation of diaspora institutions, asking them to clarify what they were doing and why in their own terms. For this reason, the project also conducted elite interviews (Dexter, 2006; Goldstein, 2002) with high-level policymakers (e.g., senior civil servants and government ministers) in as many states that currently maintain diaspora institutions as possible. Taking this approach allowed us to recognize correlative patterns in global migration policies as “social facts” in the Durkheimian sense, while also explaining such patterns in phenomenological terms through the perspectives of the social actors involved. In this sense, taking a mixed methods approach allowed us to pursue distinctly post-positivist social science explanations for the phenomena of interest.

Sampling and Data Collection

The project drew on a variety of specific qualitative and quantitative techniques, in keeping with the pragmatic approach emphasized by mixed methods researchers and post-positivist social scientists (outlined above). Our overall approach was cross-national comparison, with state-years as the basic unit of analysis (c.f. Hoffmeyer-Zlotnik, Harkness, & Zentrum für Umfragen, 2005). The qualitative component of the research defined, collected, coded, and analyzed comparative cross-country data on the types of diaspora institutions existing (or not) across all UN member states over the period 1990 to 2014. Geographically, the project conducted a census of states, but temporally, the database represents a sample restricted to the period after the disintegration of the USSR, when many new units of analysis (states) first came into existence. For the qualitative component, our sampling frame came from the attendance list of a specific international event: the inaugural Diaspora Ministerial Conference (DMC) convened by the International Organization for Migration (IOM) in 2013, involving high-level participation from most of the world’s diaspora institutions. The organizer of the DMC conference, Peter Schatzer, had heard us describing the Diaspora Engagement Policies Project at another conference and invited our lead researcher to participate in the DMC. After the conference, Schatzer and his assistants also generously provided us with full recordings of the proceedings and a full list of participants. We then sent invitations in six languages for these participants to meet us for telephone and/or in-person interviews where possible. Schatzer also emailed the participants drawing their attention to our research project and providing our contact details.

We compiled quantitative data by integrating a range of quantitative and qualitative sources. As well as speaking directly to policymakers, we consulted primary archives and conducted Internet searches of government websites based on standardized keywords to determine the existence and type of diaspora institutions for each UN member state in each year. We catalogued the sources from which we learned of each institution (there is neither space nor need to list all the sources here). Rather than a simple process of data extraction and compilation, this involved researchers carrying out case studies (Yin, 2003) of each individual UN member state, using multiple sources, to make judgments about the existence and type of diaspora institutions in each state over time. Multiple researchers were given the same coding scheme to test the validity and reliability of one another’s findings.

We also collected data in accordance with our hypotheses about the determinants of diaspora institution emergence, covering (for example) national-level data on external voting provisions; bilateral social security agreements; attitudes to UN policies; GDP and debt levels; insurance availability; the occurrence of natural disasters; access to finance; financial crises; diplomatic exchange; skilled emigrant stock; polity, freedom, and democracy measures; political orientation; citizenship data; membership in organizations; human rights norms; incentives for investment and remittances, and others. These data were compiled mainly from secondary sources such as the World Bank and the United Nations, though a few were derivative measures.

The project gathered qualitative information through in-depth interviews and formal statements. Our 45 in-depth interviews covered 20 states and three international organizations. The interviews were carried out by the project lead, over the telephone in most cases (40) but also face-to-face where the opportunity arose (in the case of 5 interviews, including with senior civil servants in India and Mexico). Wherever possible, multiple interviews were conducted or statements recorded at each institution to ensure reliability. The interviewees were not compensated in any way but were given the option to request a copy of their interview transcript or a copy of the published findings. The interviews ranged in length from 40 to 90 minutes. All were recorded, including a statement of informed consent, and the interviewees were given the opportunity to nominate part, or all, of the interview as “off the record.” In addition, at the DMC and other events, we took 51 formal public statements covering 42 states and 9 key international organizations and donor states. Both the interviews and the formal statements were then transcribed by hired assistants, and transcripts were sent to those who had requested them. This process yielded a database of qualitative information on a sample of 60 states with diaspora institutions, as well as several key international organizations.

This pragmatic approach to combining a variety of quantitative and qualitative techniques directly reflected our general agreement with the widespread view among both mixed methods scholars and post-positivist social scientists that researchers’ epistemological stance shapes how they use different methods rather than which specific methods they use (c.f. McKendrick, 1999).

Analysis

We analyzed our quantitative data statistically, aiming to identify which kinds of states tended to establish and maintain diaspora institutions and which did not. We used a range of statistical techniques, again reflecting the preferences of both mixed methodologists and post-positivist epistemologists. Geospatial techniques (Bailey & Gatrell, 1995) including geographical information systems and cartograms were used to explore and map key variables and examine simple bivariate correlations (e.g., between institution emergence and rates of remittances; see Figure 2). This allowed the detection of spatial clusters of states that are difficult to identify using other descriptive techniques. Multiple correspondence analysis (a version of correspondence analysis; see Abdi & Valentin, 2007) was used to observe the underlying structure of the categorical data and compare it with the theoretical expectations. Multiple regression analysis was then employed to analyze the extent to which diaspora institution emergence is a function of the numerical variables in the data set (Cohen, 2003), and event history techniques were used to examine the extent to which categorical variables might be associated with the emergence of diaspora institutions over time. In this sense, the project employed mixed methods not just in terms of integrating qualitative and quantitative approaches but also by combining a range of quantitative techniques, such as those described above.

Figure 2.

Cartogram analysis of the relationship between diaspora institution existence and migrant remittance receipts as a percentage of gross domestic product.

We analyzed the qualitative data thematically. All the transcripts were imported into specialized software (NVivo) and coded according to themes linked directly to the hypotheses of the research project (outlined above). The wider aim of this analytical strategy was to interpret whether the hypotheses formed based on theory and quantitative analysis were consistent with policymakers’ own understandings of their activity in relation to diaspora institutions, when we asked them what they were doing and why they were doing it. Through this analysis, we hoped to address the task of explanation in ways that could not be addressed adequately through identifying statistical correlations alone. In this way, the mixed methods analytical strategy again directly supported our goals of post-positivist social science explanation.

Validation

A range of frameworks are available for the validation of mixed methods research (c.f. Daigneault & Jacob, 2014; Heyvaert, Hannes, Maes, & Onghena, 2013). Many mixed methods scholars have taken care to point out that, as Shadish et al. (2002, p. 34, as cited in Maxwell & Mittapalli, 2010, p. 158) put it, “validity is a property of inferences . . . not a property of designs or methods. . . . No method guarantees the validity of an inference.” As Maxwell and Mittapalli (2010, p. 159) put it, “The types of evidence that are relevant to a claim depend on the nature of the claim,” and “claims about meanings and perspectives . . . [that is] interpretive claims, require quite different sorts of evidence from claims about behavior, let alone claims about the relationship between variables.” In keeping with these insights, we did not adhere to any single set of prescriptive criteria that purports to guarantee the ‘quality’ of our mixed methods research design.

Instead, we undertook a range of specific measures to validate our quantitative data and analysis consistently with our post-positivist focus on explanation, using quantitative methods to describe and qualitative methods to interpret. To avoid false positives during quantitative data collection (i.e., mistakenly registering nonexistent diaspora institutions), we did not code an institution as existent unless we found convincing evidence from an authoritative official source or confirmatory accounts from several secondary sources, such as peer-reviewed academic studies or reputable newspaper reports. Meanwhile, to avoid false negatives (i.e., failing to register actually existing diaspora institutions), we conducted a census of UN member states, rather than collecting a sample: We searched systematically for diaspora institutions in each state, even when such institutions were not immediately evident. We documented sources of information on each diaspora institution, which were cross-checked by several different research assistants to avoid observer bias and error, and monitored these sources over time to keep the data current. We recoded the data several times until we were certain that our definitions of diaspora institutions were discrete and exhaustive—that is, that they included all relevant and comparable institutions and no other entities. These validation measures allowed us to be confident in indicating which states had diaspora institutions in specific years and which states did not, and to make meaningful comparisons between the two groups. To validate our analytical results at the exploratory stage, we used a range of techniques, from geospatial analysis to cross-tabulation, to get a descriptive overview of the data (Bailey & Gatrell, 1995). When moving to multivariate analysis, we used a range of standard regression diagnostic techniques to check the robustness of our results (Belsley, Kuh, & Welsh, 1980; Fox, 1997).

We also took appropriate measures to validate our qualitative data and analysis. To validate the results of our qualitative thematic analysis, all the material was coded twice, first by a research assistant and then by the lead researcher (Baxter & Eyles, 1997). Elite interviews were employed, wherein “the investigator is willing, and often eager to let the interviewee teach him what the problem, the question, the situation is” (Dexter, 2006, p. 19; c.f. Goldstein, 2002). Although this method can be used with interviewees holding any status, it is particularly useful for interviews with “people in important or exposed positions [who] may require VIP interviewing treatment on the topics which relate to their importance or exposure” (Dexter, 2006, p. 18). A common limitation of elite interviews is that their validity and utility may be constrained by the interviewer’s understanding of the context (Dexter, 2006). In our case, contextual understanding was provided by extensive experience with the research topic, including attendance, over the course of a decade, at numerous meetings, workshops, and conferences on the subject of “engaging diasporas,” often involving interviews with policymakers and their staff members. The purpose of this section is not to provide an exhaustive list of the validation measures used in this research, nor is there the space to do so. Rather, our aim is to indicate how our validation measures were consistent with best practice in the literatures on specific methodologies, and how we combined these methods and validation techniques in ways that directly supported our wider goal of post-positivistic social science explanation, using quantitative methods for description of overall patterns of behavior and qualitative methods for interpretation of the motives and intentions of the key actors themselves.

Integration of Methods

Mixing and integration of methods took place during several stages of the research. As the above section on data collection has already discussed, we integrated information from a range of qualitative sources into a large data set for quantitative analysis. As mentioned in our section on analysis, we also used an integrated mix of quantitative and qualitative techniques to test hypotheses, triangulate and validate findings, and provide a more reliable and complete assessment of our findings (c.f. Mertens & Hesse-Biber, 2012). Moreover, during the inferential phase of the research, we integrated the qualitative and quantitative results in our deductive and inductive reasoning. Below, we provide detailed examples to illustrate our approach to integration at the analysis and inferential stages—a central topic in the literature on mixed methods research (e.g., see Fetters, Curry, & Creswell, 2013).

Integration During the Analysis Phase

We integrated qualitative and quantitative methods of analysis to test and refine our hypotheses. For example, we had formed a two-tailed hypothesis regarding the relationship between diaspora institution establishment and origin-state regime type based on the theoretical and comparative literature. On one hand, some existing studies portrayed active state interest in emigrants and their descendants as an example of authoritarianism (Brand, 2006). On the other, other existing literature interpreted the same sorts of programs as expressions of democratic inclusivity stretching beyond borders (Rhodes & Harutyunyan, 2010). Accordingly, we offered two alternative hypotheses for the quantitative research phase, predicting that diaspora institution formation would be either positively or negatively associated with regime type measured on a conventional Polity IV scale from −10 (fully autocratic) to +10 (fully democratic) (Marshall, Gurr, & Jaggers, 2016). The results were intriguing but inconclusive: Initially, we could not find a strong relationship in either direction. But then, by incorporating a quadratic term in our regression equation (Belsley, Kuh, & Welsh, 1980; Fox, 1997), we detected a signal suggesting a second-degree functional relationship between regime type and diaspora institution formation: that is, we found evidence that the likelihood of diaspora institution emergence initially rose as states became less autocratic but fell once states became more fully democratic (around +4 on the Polity IV scale—see Figure 3³). In other words, the quantitative results in some models suggested that diaspora institutions might be characteristic of regimes in transition between autocracy and democracy (see Gamlen, Cummings, et al., 2013). In keeping with our broader argument, we recognized the power of quantitative analysis to identify such large-scale patterns, while remaining mindful of the need for detailed qualitative work to provide a “causal” explanation consistent with a post-positivist social-scientific worldview.

Figure 3.

Marginal effects of regime type (Polity IV scores) on the probability of diaspora institution establishment and maintenance, 1990-2005.

Keeping this tentative statistical finding in mind, we examined the qualitative database for statements related to regime type. We corroborated this finding in the interviews: Many policymakers noted what they understood as processes of democratic transition, or “democratization.” For example, an advisor to Burundi’s Directorate of the Diaspora told us that

Burundi is known as one of the conflict affected countries. . . . This was followed by massive movement of migrants—refugees—to neighboring countries. . . . To the government these were, let’s say, viewed as enemies and they also viewed the government like they were forced. . . . Now the ruling party is made of those who were refugees. . . . They don’t want to leave anybody behind this process. . . . Once you have said that you are born here, . . . they said, OK, you are welcome, you can come. . . .That is the reason why this Directorate was created, that you have to take those who want to come back. It’s part of this process of democratization.

Similarly, a senior representative of the Department for Romanians Abroad within the Romanian Ministry of Foreign Affairs explained to us,

In 1989 was the revolution, which brought a democratic regime. . . . Before 1989 Romanians abroad were only exiled people. . . . They were basically the opponents of the regime, and the Romanian state considered them as enemies, so there was no reason to interfere with them in a positive way. . . . So I thought to myself, . . . our job is to stress the best possible way of viewing this phenomenon, and I started to count the major benefits that we could have from this phenomenon.

Meanwhile, Ethiopia’s ambassador to the United Nations related a similar story:

Ethiopia has started a journey of democratization in 1991. . . . This policy seeks to lay the ground, to consolidate the different aspects of this participation and assist our diaspora, facilitate our conditions for our diaspora to contribute constructively to building institutions in Ethiopia for democracy.

The executive director of the Institute for Mexicans Abroad within the Mexican Exterior Ministry put it even more explicitly:

What you see is possible because of the democratization of power in Mexico, and how power—political power—was decentralized. The independent variable is the democratization of Mexico; the dependent variable is [Mexico’s diaspora institution] the Program to Mexican Communities.

These qualitative findings supported our refined hypothesis that diaspora institutions are characteristic of transitional regimes (“anocracies”) rather than full democracies or autocracies. Moreover, they gave us an explanation for this connection grounded in the participants’ own understandings of their actions: The new leaders of previously autocratic states often saw their efforts to establish diaspora institutions as a way of welcoming back exiles into a newly democratic fold. As this example demonstrates, our integration of quantitative and qualitative methods of data analysis allowed us to enhance our interpretation of significant findings (Onwuegbuzie & Leech, 2004), testing and refining our hypotheses and building a more complete explanatory picture of why states engage their diasporas through dedicated institutions. The example also demonstrates how, by integrating the pattern recognition strengths of quantitative analysis with the insights about motivation and decision making from qualitative research, we achieved a type of social-scientific explanation that is consistent with the expectations of post-positivism, where explanation involves not only quantitative description of correlation but also qualitative interpretation of the motives and intentions of specific social action.

Integration During the Inferential Phase

We also integrated methods during the inferential stage of the research, such as when qualitative analysis yielded more conclusive results than quantitative techniques (or vice versa). For example, our quantitative analyses yielded results that contradicted hypotheses from the existing rationalist and constructivist theoretical literature: We found that remittance receipts (money transfers from migrants to their families in the homeland), high-skilled emigration rates, right-wing government orientation, and regime type did not matter in the ways predicted by the rationalist and constructivist theories identified at the outset (see above). However, the quantitative results also did not conclusively support our own alternative approach: There was no initially strong statistical support for the theoretical expectation that international organization membership, geographical proximity, or shared colonial heritage would predict diaspora institution emergence. However, our qualitative results were more conclusive on this topic. We found many important quotes from leaders of international organizations recommending that states form diaspora policies. For instance, the European Union representative at the IOM’s inaugural 2013 DMC made the following statement:

The EU and its member states firmly believe that the creation of effective governance frameworks in both countries of origin and destination plays a key role in creating the conditions for effective realization of the development potential of diasporas. . . . We expect this ministerial meeting to provide useful insights and orientation on key challenges such as how governments should engage diasporas and support diaspora groups. . . . It should also examine how consideration of migration and development issues, such as the role of the diaspora, can be reflected in the post-2015 development framework.

We found related statements from officials at organizations such as the World Bank, the UN, and, most important, the IOM.

Meanwhile, our qualitative data also contained quotes from high-level national policymakers across many states reporting that they were acting on such explicit or implicit recommendations from officials in international organizations when establishing their diaspora institutions. For example, the State Minister of Georgia for Diaspora Issues told us that the “IOM has served as one of the best protectors and advisers to Georgian emigrants—both irregular and legal. It has greatly contributed to designing diaspora policy” (Surguladze, personal communication, January 20, 2014).

We integrated these findings to make explanatory inferences. From the quantitative results, we concluded that the existing rationalist and constructivist theories could not provide a convincing general picture of what kinds of states form diaspora institutions. Meanwhile, from the qualitative results we concluded that states were often acting in ways that directly supported our approach. Overall, we concluded that our factors of interest played an overlooked role in explanations for why states form diaspora institutions. In this way, we integrated insights from quantitative and qualitative research to make inferences consistent with post-positivist social-scientific thinking.

Conclusions on the Use of Post-Positivistic Mixed Methods Designs

There are often practical reasons for the adoption of mixed methods—a choice made by a growing number of researchers in the study of migration, such as those included in this special issue. Disparate fields within migration studies often overlap: For example, migration scholarship and policy research may encompass fields as diverse as refugee law, labor economics, and the anthropology and psychology of immigrant adaptation. Increasingly, researchers across these fields are favoring mixed methods strategies for pragmatic and normative reasons, as part of wider efforts to enact social change.

By contrast, in keeping with our post-positivist approach, in this article we show how mixed methods research strategies are useful not just for changing social and political action surrounding migration but also for explaining such action. In our view, post-positivist explanation in social science must be descriptive of general patterns of social behavior and also make sense of this behavior from the perspective of the actors involved. This, we believe, is consistent with post-positivism’s combination of an ontological openness to a reliable “true” reality, an epistemological commitment to the fallibility of observation, and compatibility with the constructivist belief that in the social world all “facts” are socially constructed.

In this article, we focused on the methodological challenges encountered during the Diaspora Engagement Policies Project, a 5-year research initiative that examined the policies of migrants’ origin states toward emigrants and their descendants in the diaspora. With this project as an example, we argued that mixed methods are well suited to the post-positive approach to social science theory building because quantitative methods may help describe a reliable social reality and qualitative methods may allow researchers to understand how and why this social reality has been constructed and what it means to those involved.

Using mixed methods in this migration research project allowed us to minimize bias by combining research techniques with complementary strengths, and to triangulate our findings using more than one approach. It also allowed us to understand both how states are behaving and what incentives might drive them, as well as how this behavior is shaped by the social and political understandings of the policymakers themselves. On one hand, it allowed us to identify the social facts in our research field—the patterns of policy and institutional arrangements in place around the world—through large-scale cross-national comparison using quantitative methods. These social facts belong not to an ultimate objective reality of the kind identified in positivist studies but to an intersubjective reality that different social actors can agree on. On the other hand, this approach allowed us to move beyond the identification of correlations and get more directly at causation, as understood from a post-positivist social science perspective. This involved drilling down into the patterns of social behavior by speaking to the people concerned about what they were doing and why.

The Diaspora Engagement Policies Project provided a range of opportunities to use mixed methods in this distinctly post-positivist way. It allowed us to take a hypothetico-deductive approach and assume the social reality of our subject matter (consistent with post-positivism’s openness to the notion of reality), without assuming that we could provide unbiased or objective descriptions or inferences about that subject matter (consistent with post-positivism’s commitment to observer fallibility). It allowed us to combine a variety of methods, putting aside the common misconception that an epistemological perspective dictates what research methods can be used, rather than merely shaping how different methods are used. It allowed us to combine methods, minimize bias, and validate findings in a way consistent with post-positivist premises about social reality. And it allowed us to integrate insights from diverse research methods during the analysis and inference stages of the research and to arrive at well-supported post-positivist social science explanations.

Based on this research experience, we argue that there are strong reasons for using mixed methods in the study of migration and in post-positivist explanatory research in the social sciences and health sciences more generally. This contrasts somewhat with the widespread focus on the uses of mixed methods research for practical, policy-facing research work and with the misconception of post-positivism as a soft, quantitatively focused version of positivism. Specifically, we argue that mixed methods are well suited to the twin tasks of describing and explaining the social world. In our example, quantitative methods played an important role in identifying and describing reliable social patterns, whereas qualitative methods allowed us to interpret and understand the psychological and sociocultural frames of reference of the social actors who created these patterns through their interactions. Mixed methods in this way can directly facilitate the kind of knowledge sought by the post-positivist worldview, which is open to the existence of a “true” social reality but recognizes that this reality is interpreted and re-created differently by each social actor and that knowledge consists of understanding these different interpretations from the perspectives of those involved. In this way, mixing methods has allowed us to help build explanatory theory in the increasingly important interdisciplinary field of migration studies.

Footnotes

Acknowledgements

We gratefully acknowledge support from Victoria University and the Paekakariki Institute in Wellington, and Stanford University’s Centre on Democracy, Development and the Rule of Law. We thank all the participants and researchers who have contributed, the editors and three anonymous reviewers at the Journal of Mixed Methods Research, and the guest editors of this special issue.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was funded by the Leverhulme Trust through the Oxford Diasporas Programme.

Notes

References

Abdi

Valentin

(2007). Multiple correspondence analysis. In Salkind

(Ed.), Encyclopedia of measurement and statistics (pp. 651-657). Thousand Oaks, CA: Sage.

Adamson

(2014). Classical philosophy: A history of philosophy without any gaps. Oxford, England: Oxford University Press.

Agunias

D. R.

(2009). Closing the distance: How governments strengthen ties with their diasporas. Washington, DC: Migration Policy Institute.

Agunias

D. R.

Newland

(2012). Developing a road map for engaging diasporas in development: A handbook for policymakers and practitioners in home and host countries. Washington, DC: Migration Policy Institute.

Bailey

T. C.

Gatrell

A. C.

(1995). Interactive spatial data analysis. Harlow, England: Pearson Education.

Baxter

Eyles

(1997). Evaluating qualitative research in social geography: Establishing “rigour” in interview analysis. Transactions of the Institute of British Geographers, 22, 505-525.

Belsley

D. A.

Kuh

Welsh

R. E.

(1980). Regression diagnostics: Identifying influential data and sources of collinearity. New York, NY: Wiley.

Blaikie

(1993). Approaches to social enquiry polity. Cambridge, England: Cambridge University Press.

Brand

L. A.

(2006). Citizens abroad: Emigration and the state in the Middle East and North Africa. Cambridge, England: Cambridge University Press.

10.

Bustamante

(2011). On MacCormick’s post-positivism: Comment on Karen Petroski’s “Is Post-Positivism Possible?” German Law Journal, 12, 693-727.

11.

Creswell

J. W.

Plano-Clark

V. L.

(2011). Designing and conducting mixed methods research. Thousand Oaks, CA: Sage.

12.

Cohen

(2003). Applied multiple regression/correlation analysis for the behavioural sciences. London, England: Lawrence Erlbaum.

13.

Creswell

J. W.

(2009). Research design: Qualitative, quantitative, and mixed methods approaches (3rd ed.). London, England: Sage.

14.

Crotty

(1998). The foundations of social research: Meaning and perspective in the research process. Thousand Oaks, CA: Sage.

15.

Cruickshank

(2011). The positive and the negative: Assessing critical realism and social constructionism as post-positivist approaches to empirical research in the social sciences (International Migration Institute Working Paper 42). Oxford, England: University of Oxford.

16.

Daigneault

P.-M.

Jacob

(2014). Unexpected but most welcome: Mixed methods for the validation and revision of the participatory evaluation measurement instrument. Journal of Mixed Methods Research, 8(1), 6-24.

17.

Délano

Gamlen

(2014). Comparing and theorizing state–diaspora relations. Political Geography, 41, 43-53.

18.

Descartes

(1984). Principia philosophiae [Principles of philosophy] ( Miller

R. P.

, Trans.). Berlin, Germany: Springer Science + Business Media. (Original work published 1644)

19.

Dexter

L. A.

(2006). Elite and specialized interviewing. Colchester, England: European Consortium for Political Research.

20.

Fetters

M. D.

Curry

L. A.

Creswell

J. W.

(2013). Achieving integration in mixed methods designs: Principles and practices. Health Services Research, 48, 2134-2156.

21.

Findlay

A. M.

F. L. N.

(1999). Methodological issues in researching migration. The Professional Geographer, 51(1), 50-59.

22.

Fox

(1997). Applied regression analysis, linear models, and related methods. Thousand Oaks, CA: Sage.

23.

Fox

N. J.

(2008). Postpositivism. In Given

L. M.

(Ed.), The SAGE encyclopedia of qualitative research methods (Vol. 2, pp. 659-664). Thousand Oaks: Sage.

24.

Gamlen

(2006). Diaspora engagement policies: What are they, and what kinds of states use them? (COMPAS Working Paper 06-32). Oxford, England: University of Oxford.

25.

Gamlen

(2008). The emigration state and the modern geopolitical imagination. Political Geography, 27, 840-856.

26.

Gamlen

(2014a). Diaspora governance. In Anderson

Keith

(Eds.), Migration: The COMPAS anthology (pp. 21-23). Oxford, England: Centre on Migration, Policy and Society.

27.

Gamlen

(2014b). Diaspora institutions and diaspora governance. International Migration Review, 48, S180-S217.

28.

Gamlen

(2015a). Explaining diaspora engagement policies. In Cohen

Story

(Eds.), The impact of diasporas (pp. 22-24). Oxford, England: Oxford Diasporas Programme.

29.

Gamlen

(2015b). The rise of diaspora institutions. In Sigona

Gamlen

Libertore

Neveu-Kringelbach

(Eds.), Diasporas re-imagined: Spaces, practices and belonging (pp. 166-171). Oxford, England: Oxford Diasporas Programme.

30.

Gamlen

Betts

Délano

Lacroix

Paoletti

Sigona

Vargas-Silva

(2013). Faultlines and contact zones: A new forum for migration studies. Migration Studies, 1(1), 1-3.

31.

Gamlen

Cummings

Vaaler

Rossouw

(2013). Explaining the rise of diaspora institutions (International Migration Institute Working Paper 78). Oxford, England: University of Oxford.

32.

Gamlen

Cummings

Vaaler

(2017) Explaining the rise of diaspora institutions. Journal of Ethnic and Migration Studies. https://doi.org/10.1080/1369183X.2017.1409163

33.

Giddens

(1987). Social theory and modern sociology. Stanford, CA: Stanford University Press.

34.

Goldstein

(2002). Getting in the door: Sampling and completing elite interviews. PS: Political Science & Politics, 35, 669-672.

35.

Guest

(2013). Describing mixed method research: An alternative to typologies. Journal of Mixed Methods Research, 7(2), 141-151.

36.

Henderson

K. A.

(2011). Post-positivism and the pragmatics of leisure research. Leisure Sciences, 33, 341-346.

37.

Hesse-Biber

Johnson

R. B.

(2013). Coming at things differently: Future directions of possible engagement with mixed methods research. Journal of Mixed Methods Research, 7(2), 103-109.

38.

Heyvaert

Hannes

Maes

Onghena

(2013). Critical appraisal of mixed methods studies. Journal of Mixed Methods Research, 7(4), 302-327.

39.

Hoffmeyer-Zlotnik

J. H.

Harkness

J. A.

Zentrum für

Umfragen

. (2005). Methodological aspects in cross-national research. Mannheim, Germany: Zentrum für Umfragen, Methoden und Analysen.

40.

International Organization for Migration. (2013, June 18-19). Summary of discussions prepared by the Secretariat (IDM 2013: Diaspora Ministerial Conference). Retrieved from http://diaspora.iom.int/sites/default/files/pdf/Diaspora-Ministerial-Conference-Summary-19-June-2013.pdf

41.

Johnson

R. B.

Gray

(2010). A history of philosophical and theoretical issues for mixed methods research. In Tashakkori

Teddlie

(Eds.), SAGE handbook of mixed methods in social and behavioral research (2nd ed., pp. 69-94). Thousand Oaks, CA: Sage.

42.

Johnson

R. B.

Onwuegbuzie

A. J.

(2004). Mixed methods research: A research paradigm whose time has come. Educational Researcher, 33(7), 14-26.

43.

Johnson

R. B.

Onwuegbuzie

A. J.

Turner

L. A.

(2007). Toward a definition of mixed methods research. Journal of Mixed Methods Research, 1(2), 112-133.

44.

Kuhn

T. S.

(2012). The structure of scientific revolutions. Chicago, IL: University of Chicago Press. (Original work published 1962)

45.

Lorkowski

C. M.

(2015). David Hume: Causation. In Internet encyclopedia of philosophy. Retrieved from http://www.iep.utm.edu/hume-cau/

46.

Marshall

Gurr

T. R.

Jaggers

(2016). Polity IV project: Political regime characteristics and transitions, 1800-2015. Retrieved from http://www.nber.org/ens/feldstein/ENSA_Sources/CSP/Polity%20Score/p4manualv2015.pdf

47.

Massey

D. S.

(1987). Special issue: Measuring international migration: Theory and practice. International Migration Review, 21, 1498-1522.

48.

Maxwell

J. A.

Mittapalli

(2010). Realism as a stance for mixed methods research. In Tashakkori

Teddlie

(Eds.), SAGE handbook of mixed methods in social and behavioral research (2nd ed., pp. 145-168). Thousand Oaks, CA: Sage.

49.

McKendrick

J. H.

(1999). Multi-method research: An introduction to its application in population geography. The Professional Geographer, 51, 40-50.

50.

Mertens

D. M.

(2013). Emerging advances in mixed methods: Addressing social justice. Journal of Mixed Methods Research, 7(3), 215-218.

51.

Mertens

D. M.

(2014). A momentous development in mixed methods research. Journal of Mixed Methods Research, 8(1), 3-5.

52.

Mertens

D. M.

Hesse-Biber

(2012). Triangulation and mixed methods research: Provocative positions. Journal of Mixed Methods Research, 6(2), 75-79.

53.

Meyer

J. W.

(2010). World society, institutional theories, and the actor. Annual Review of Sociology, 36, 1-20.

54.

O’Cathain

Murphy

Nicholl

(2008). The quality of mixed methods studies in health services research. Journal of Health Services Research & Policy, 13(2), 92-98.

55.

Onwuegbuzie

A. J.

Leech

N. L.

(2004). Enhancing the interpretation of “significant” findings: The role of mixed methods research. Qualitative Report, 9, 770-792.

56.

Petroski

(2011). Is post-positivism possible? German Law Journal, 12, 663.

57.

Petty

N. J.

Thomson

O. P.

Stew

(2012). Ready for a paradigm shift? Part 1: Introducing the philosophy of qualitative research. Manual Therapy, 17, 267-274.

58.

Philip

(1998). Combining quantitative and qualitative approaches to social research in human geography: An impossible mixture? Environment and Planning A, 30, 261-276.

59.

Phillips

D. C.

(1983). After the wake: Postpositivistic educational thought. Educational Researcher, 12(5), 4-12.

60.

Phillips

D. C.

Burbules

N. C.

(2000). Postpositivism and educational research. Oxford, England: Rowman & Littlefield.

61.

Plano-Clark

V. L.

Schumacher

West

Edrington

Dunn

L. B.

Harzstark

. . . Miaskowski

(2013). Practices for embedding an interpretive qualitative approach within a randomized clinical trial. Journal of Mixed Methods Research, 7(3), 219-242.

62.

Pirret

A. M.

(2013). Nurse practitioner diagnostic reasoning (Doctoral dissertation, Massey University). Retrieved from https://mro.massey.ac.nz/bitstream/handle/10179/4929/02_whole.pdf?sequence=2&isAllowed=y

63.

Popper

(2002). The logic of scientific discovery. London, England: Routledge. (Original work published 1934)

64.

Rhodes

Harutyunyan

(2010). Extending citizenship to emigrants: Democratic contestation and a new global norm. International Political Science Review, 31, 470-493.

65.

Schwandt

T. A.

(1994). Constructivist, interpretivist approaches to human inquiry. In Denzin

N. K.

Lincoln

Y. S.

(Eds.), Handbook of qualitative research (pp. 118-137). Thousand Oaks, CA: Sage.

66.

Scotland

(2012). Exploring the philosophical underpinnings of research: Relating ontology and epistemology to the methodology and methods of the scientific, interpretive, and critical research paradigms. English Language Teaching, 5(9), 9-16.

67.

Sharp

McDonald

Sim

Knamiller

Sefton

Wong

(2011). Positivism, post-positivism and domestic water demand: Interrelating science across the paradigmatic divide. Transactions of the Institute of British Geographers, 36, 501-515.

68.

Sheppard

(2014). We have never been positivist. Urban Geography, 35, 636-644.

69.

Smith

(1979). Geography, science and post-positivist modes of explanation. Progress in Human Geography, 3, 356-383.

70.

Tashakkori

Creswell

J. W.

(2007). Editorial: The new era of mixed methods. Journal of Mixed Methods Research, 1(1), 3-7.

71.

Tashakkori

Teddlie

(2003). Handbook of mixed methods in social and behavioral research. London, England: Sage.

72.

Tucker

W. T.

(1965). Max Weber’s “Verstehen.” Sociological Quarterly, 6, 157-165.

73.

Van der Roest

J.-W.

Spaaij

Van Bottenburg

. (2015). Mixed methods in emerging subdisciplines: The case of sport management. Journal of Mixed Methods Research, 9(1), 70-90.

74.

Yin

(2003). Case study research: Design and methods. Thousand Oaks, CA: Sage.