Qualitative vs. Quantitative Research Methods in Social Science

Introduction

Two of the most prominent approaches in social science research are qualitative and quantitative research methods. Qualitative research is believed to operate from a subjective, constructionist view of reality, whereas quantitative research operates from an objective, positivist viewpoint of the world. There has been quite a bit of debate over the merits of each approach, often with one paradigm belittling the assumptions of the other. This literature review explores the philosophical foundations of each paradigm, compares their practical differences, and discusses the strengths and weaknesses of both approaches as they relate to research in the social sciences and to human resources (HR) research. The rationale for mixed-methods research, where the two paradigms are combined, is also discussed.

In recent years there has been substantial interest concerning the role of specific paradigms and philosophical assumptions with regard to conducting research. There has been a growing concern regarding the adequacy of research methods in social sciences and HR studies (Anderson, 2004). This interest developed alongside increasing attention to cross-cultural influences in the social sciences and as they related to understandings of employees and management. Ambiguities have emerged in interpreting and utilizing findings from previous areas of research. Many researchers and their findings have been accused of producing scientific studies that do not fully explain phenomena as they occur in the real world.

A good deal of HR research has been largely empirical and predicated on a positivist research philosophy. This differs from the phenomenological research viewpoint that has inspired research in traditional HR management disciplines. Much of the empirical research in HR has evolved largely from doctoral dissertations, underscoring the importance of sound research grounding — predominantly where quantitative approaches, primarily hypothesis testing, are concerned (Anderson, 2004). In its traditional context, human resources management has not been a quantitative discipline, preferring the qualitative approach (Anderson, 2004; Huselid, 1995). The difference between qualitative and quantitative research reflects two distinct philosophical viewpoints.

The predominant school of empirical thought has traditionally used a quantitative approach based on research in "hard" sciences such as biology and medicine (Blalock, 1984). Beginning in the 1970s, serious concerns were raised about the limits of quantitative research methods associated with the prevalent paradigm of the time — positivism (Blalock, 1984). Positivism presupposes that the world is objective and that scientific methods can inherently measure and represent aspects of the world, explaining and predicting causal relations among central variables. Critics countered, however, that positivistic methods strip context from meanings during the process of developing quantified measures (Lincoln & Guba, 2000). These critics charge that quantitative measures omit subjective interpretations from data, impose the researcher's interpretations on the data, and require statistical samples that may not be representative of specific groups, preventing generalization to individual cases.

Positivism was the dominant influence on social research for many years; however, it has been challenged by critics from two different traditions: interpretive constructionism and critical postmodernism (Blalock, 1984). Constructionism and postmodernism have offered alternative theoretical and methodological approaches regarding management and organizations by generating interest in political and social issues that positivist-oriented research did not address. Positivistic research relies on experimental or survey methodologies that interpretivists have criticized for imposing a worldview on subjects rather than describing and understanding those views. Postmodernists have argued that positivist methods imply forms of knowledge that support capitalist structures and inequality. As a result of these criticisms there has been growing interest in qualitative research methods that do not follow the quantitative interpretation of variables and statistical methods, but instead seek to explore the subjective experiences of research participants (Blalock, 1984). Both quantitative and qualitative research methods should be viewed as complementary rather than as opposites.

To understand the paradigms of quantitative and qualitative research one should briefly trace their origins. These origins date back at least to the seventeenth-century philosophical debates regarding the nature of knowledge and its relationship to reality. Burrell and Morgan (1979) suggest that assumptions about the nature of reality can be conceptualized in terms of a subjective (qualitative)–objective (quantitative) dimension.

The subjective (qualitative) view is centered on the supposition that the social world outside of individual cognition is composed of nothing more than labels, names, and concepts that serve as artificial creations. Their usefulness is based on their convenience as devices that allow for describing, interpreting, and negotiating the world (Anderson, 2004). This view holds that a social action takes place when an actor assigns a meaning to their conduct or environment and, by means of that meaning, relates their act to the acts of others. Human actions are reciprocally oriented to one another not in a mechanistic stimulus-and-response manner, but through an interpretive process (Blalock, 1984). In order to understand the significance and meaning of social phenomena it is therefore first necessary to appreciate this interpretive process — to discover and understand the motives and goals that guide people to act. For the subjectivist researcher, understanding and interpretation take precedence over simple causal explanations or group-level differences (Firestone, 1987).

In contrast, quantitative research methods originated from hard sciences such as chemistry, biology, and physics and follow an objectivist view centered on the ontological assumption that the social world exists independently of one's subjective experience of it (Burrell & Morgan, 1979). This objective view of reality as a tangible, physical structure supports an epistemological position that highlights the importance of studying the characteristics of relationships among the elements that make up that structure. Knowledge of the organization of reality from this viewpoint involves understanding and mapping the causal relationships among the components, ingredients, or factors of the structure (Firestone, 1987). This viewpoint encourages an objective mode of knowing that specifies the particular nature of the laws and relationships among phenomena, measured in terms of facts (Anderson, 2003; Blalock, 1984).

The phenomenological orientation (qualitative) disputes the notion that there can be any genuinely objective knowledge that can be spelled out and communicated in a tangible form, since what we know is simply an expression of how we have arbitrarily imposed our frame of reference on reality (Firestone, 1987). We mistakenly perceive reality as consisting of an external and separate sphere, but our interpretations of reality are inherently subjective.

The objectivist attitude (quantitative) supports an epistemology in the positivist tradition that attempts to describe and predict events in the world by identifying causal relationships and regularities among its components (Firestone, 1987). The research methodology here is influenced by experimental designs originating largely from the biological sciences. Research methods such as questionnaires can be used to investigate a range of topics, leading to the designation of research into experimental and correlational designs: true experiments can explain causal relations, while correlational designs explain relationships that cannot be assumed to be causal. Quasi-experimentation (Campbell & Stanley, 1963) is a research approach that can be used to examine causal relationships in situations where true experiments requiring random assignment to treatment conditions are difficult to achieve or altogether inappropriate. The quantitative approach highlights the use of statistical procedures, measures of association, and the expansion of measurement models. This approach has been developed on the basis of an ontology appropriate for the physical sciences and assumes that objects in the world and the relationships among people are interrelated by a lawful force called causation. Researchers can use experimental methods to ascertain these lawful relationships. This paradigm requires that the language of social behavior become the language of variables. The heavy reliance on experimental models of behavior derived from chemistry and physics can be criticized for its "closed" nature; such models may not be the only way to study the more "open systems" found in human organizations. Yet ignoring these principles can also lead to the investigation of ineffective methods of change (Blalock, 1984).

The Philosophy of Qualitative and Quantitative Research Methods

Wilden (1972) proposed that if one accepts a purely reductionist approach, there are always going to be causes that inevitably cause further causes, and so on. An argument can be made as to whether human action and behavior constitute a fundamentally distinct type of phenomenon — one that perhaps cannot be completely studied by a scientific approach that breaks its totality into separate components and variables for individual analysis (Firestone, 1987). Phenomenology (qualitative) supports the argument that the whole is always greater than the mere sum of its parts. From the phenomenologist's perspective, human behavior should be viewed in its totality; it must be experienced to be understood. If researchers perform an analysis that requires decomposing a phenomenon, they miss out on many of the true meanings of the action. The subjectivist adheres to the notion that human behavior can only be understood in terms of meaning and within the purview of the natural sciences. Phenomenologists therefore believe that the measurement-oriented, mechanistic models of explanation supported by the positivist approach are not appropriate for understanding human behavior (Blalock, 1984). This approach strongly supports the notion that researchers should seek to understand the ways in which members of a group or society create and shape their social roles through subjective interpretations of events. Objectivists rightly point out, however, that they do not believe all people are the same, but that there are similar trends that can be observed — and this is precisely what quantitative research seeks (Firestone, 1987).

The subjectivist (qualitative) viewpoint challenges the objectivist view on the basic issue of whether people can achieve knowledge that is independent of their subjective construction of reality. Subjectivists question whether true objectivity in the interpretation of behavior is even possible, since people are active agents through whom all forms of knowledge are experienced and perceived (Blalock, 1984). Because of this, it seems improbable that a researcher's personal assumptions, values, and purposes do not somehow shape the particular methods or techniques they choose to implement. Subjectivists therefore claim that knowledge of research methods should be balanced by an understanding of research as a characteristically human process through which researchers actually create knowledge. Qualitative research attempts to focus on the particular here-and-now situation; its focus is often situational relevance. Qualitative investigations seek to describe the meaning of social life and the unfolding of social processes, as opposed to describing the social structures that are the focal point of quantitative researchers.

From a phenomenologist viewpoint, human behavior — what people do or say — is a result of how those people subjectively define and construct their reality. For qualitative research, the mission of the researcher is to grasp this process of interpretation. This can be viewed as a form of "inquiry from the inside" (Evered & Louis, 1981; Mahoney & Goertz, 2006). This type of investigation often requires the direct or indirect involvement of the researcher, an absence of a priori analytical categories, and a sincere intent to understand the individual situation. This frame of view attempts to understand the world from the respondent's frame of reference. It operates on the assumption that the researcher can best come to know the respondent's subjective reality by being there, by becoming part of the flow of activities and events, and by becoming immersed in the phenomena under study. Active participation in the situation is essential, since knowledge can best be validated experientially.

The quantitative (objectivist) school would counter by stating that subjectivity certainly exists, but that there are indeed certain principles that apply to most people. If this were not the case and all subjective views were equally valid, there would be no laws governing human conduct (Firestone, 1987). There are indeed generalities that apply to large groups. The positivist model (quantitative) operates under the assumption that research is a neutral and mechanical process by which researchers discover or uncover knowledge. Quantitative researchers have techniques that allow them to answer their questions scientifically, precisely, theoretically, and systematically (Blalock, 1984). A necessary prerequisite of any scientific answer to a research question is an acceptable empirically based database comprising representativeness, random selection, and other statistically based requirements. Theoretical knowledge results from the methodical application of the scientific method to phenomena in the world, distinguishing the researcher's scientific account from ordinary lay views. Through scientific methods — especially random sampling and replication — research findings can be generalized from particular experiments to a body of theoretical statements that are universally applicable, in the same way that gravity, chemical reactions, and the contraction of diseases are universally applicable (Firestone, 1987). Even findings not subjected to strict methodologies, such as Maslow's hierarchy of needs (Maslow, 1954), are perceived to be applicable on an overall basis.

Qualitative researchers claim to know relatively little about a given sample of observed behavior without first knowing and understanding the context in which that behavior takes place, and without viewing that behavior from the stance of the originator (Firestone, 1987). Subjectivists believe that such contextual understanding and empathetic objectives are not likely to be achieved without a firsthand and intimate knowledge of the particular research setting. As a result, qualitative researchers believe that the research process is a form of social interaction during which the researcher interacts with, converses with, and learns from the phenomena being studied. This is particularly true when the object of inquiry is a human being or a group of people (Blalock, 1984). Moreover, subjectivists believe that this interactivity with research subjects makes it possible for the researcher to serve as a knowledgeable instrument in the research process by using sensitivity, adaptability, and responsiveness to acquire information and ideas (Blalock, 1984).

Positivists (quantitative) believe that knowledge receives validation through methodological procedures and logic (Blalock, 1984). Researchers are guided by the belief that reality is external and composed of facts arranged in a law-like formation. Researchers therefore adopt the "inquiry from the outside" attitude (Evered & Louis, 1981) that requires detachment on the part of the researcher. Data are gathered according to a priori analytical categories, with the goal of discovering knowledge that can be generalized across many environments or situations. Researchers most often preselect their variables based on their particular paradigm or frame of reference and relevant to their hypotheses. Hypotheses are phrased in the language of variables and only data relevant to those variables are collected. In the quantitative approach it is crucial that the researcher maintain a careful distance from the participants. There are particular methodological safeguards that must be put in place to protect the research design against reactivity on the part of subjects, as such reactivity can influence the outcome of the research (Firestone, 1987).

Researchers from both camps have come to understand that the people and groups under study have a lifestyle and a culture that belongs to them. If researchers truly wish to understand the behavior of people and the interactions in groups or organizations, they need to be able both to describe and to appreciate the cultural dimensions of those behaviors. Qualitative research proponents such as Morgan and Smircich (1980) have argued that researchers at the organizational level should no longer be content to remain as external observers merely measuring their observations. Instead, researchers are encouraged to move into the subjective world of their subjects in order to investigate the object of study from within and to utilize techniques consistent with that approach. To operate from a qualitative mode requires that the researcher work in linguistic symbols and attempt to reduce the distance between theory and data, between indicator and indicated, and between context and action. Qualitative research employs interpretive techniques that attempt to translate, decode, describe, and come to terms with the meaning — as opposed to just the frequency — of phenomena under study in their social context.

Descriptive research is a principal method of data collection for a qualitative design. Among the qualitative methodologies currently employed, the ethnographic method has a distinguished history of usefulness. Ethnography involves a collection of methodological and interpretive procedures developed during the twentieth century (Lassiter, 2005). There are two primary approaches: symbolic interactionism, which concentrates on patterns of communication, interpretation, and adjustment between people (Griffin, 2012), and ethnomethodology, the study of the daily methods people use to construct social order (Garfinkel, 2002). These approaches involve techniques such as participant observation, in-depth interviews, content analysis, linguistic analysis, biography, and even forms of psychotherapy.

The level of analysis for the quantitative researcher is statistical (Mahoney & Goertz, 2006). The empirical surveys that dominate much of HR research — many of them large-scale — are good examples of methods that operate on the assumptions distinctive of an objectivist approach. They attempt to portray reality as a type of concrete structure. By analyzing data through sophisticated quantitative techniques such as multivariate statistical analysis, researchers can determine group differences that are statistically significant and not due to chance (Firestone, 1987).

In addition to philosophical differences, there are practical approaches to research that drive both qualitative and quantitative methods. Mahoney and Goertz (2006) discuss ten areas of discrepancy that differentiate qualitative and quantitative research studies:

1. Different approaches to explanation. Qualitative methods seek to explain individual cases and use a "causes of effects" approach. Quantitative methods estimate the mean or average effects of selected independent variables — an "effects of causes" approach.

2. Different conceptions of causation. Qualitative researchers conceptualize causes in terms of necessary and sufficient conditions. Quantitative research explains causes in terms of probability and statistical theories.

3. Multivariate explanations. Qualitative research operates on the assumption that individual events do not have a single cause and that understanding requires including an assortment of causally relevant factors. Quantitative research operates on assumptions of additive causes or statistical interactions — that is, differential effects at different levels of an independent variable (Holland, 1986).

4. Equifinality is the concept of multiple causation (George & Bennett, 2005). This refers to the notion that there are potentially multiple causal paths to the same outcome. It has been embraced by qualitative researchers but is less well accepted by quantitative researchers, who are more interested in the relative contributions to causal effects.

5. Scope and generalization. Qualitative research tends to adopt a narrow scope and avoids generalization. Quantitative research tends to adopt a broader scope to maximize statistical leverage and generalization.

6. Case selection practices. Qualitative research tends to seek out positive cases of interest (e.g., a researcher studying conflict management would seek out instances of conflict in organizations). Quantitative researchers look for random selection on independent variables and include all cases.

7. Weighing of cases. In qualitative research, the evaluation of a theory is sensitive to individual observations; a single non-conforming case can have important theoretical implications. In quantitative research, the overall pattern of fit is what matters, and all observations are given a priori equal consideration.

8. The treatment of substantially important cases. Some cases in qualitative research are more important than others — for instance, cultural factors may make certain cases more relevant. In quantitative research, unless the important aspects of a case are operationalized as an independent variable, all cases are treated equally.

9. The treatment of a lack of fit. In qualitative research, the tendency is to treat non-conforming cases with greater interest. In quantitative research, the trend is to treat them as error — though this does not mean such cases are unimportant to consider.

10. Concepts and measurement. In qualitative research, concepts are the center of attention and error can lead to a revision of concepts. In quantitative research, measurement tends to take center stage; however, error can be modeled and new indicators can be identified.

When one compares the differences between qualitative and quantitative research methods, one is inevitably navigating sensitive terrain. Researchers affiliated with either tradition will respond defensively to perceived criticisms of their disciplines or to mischaracterizations of their practices, goals, and assumptions. The potential for misconstructions of either viewpoint is considerable. This misunderstanding is compounded by the very labels "qualitative" and "quantitative," which do not adequately capture the differences between the two traditions. Quantitative analysis inherently involves numbers; however, all statistical analyses rely heavily on words and explanations for their interpretation. Conversely, qualitative studies frequently use numerical data, and a good number of qualitative techniques require quantitative information to be fully understood (Creswell, 2003). Thus, the differences are not always as clear-cut as they may seem. To compensate for the deficiencies of either model, some researchers rely on mixed-methods designs that approach analysis using both (Creswell, 2003).

After examining the basic philosophical assumptions and practical differences between these two paradigms, it becomes clear that there are arguments for combining qualitative and quantitative methods. Indeed, several viewpoints have been advanced for why these methods should be combined. One argument states that the approaches should be merged because they both seek to reach the same goal: understanding phenomena in the world (Haase & Myers, 1988; Creswell, 2003). King, Keohane, and Verba (1994), in a discourse on the mixed-methods approach, assert that both qualitative and quantitative designs share a united logic, so the same principles of inference should apply to both.

Second, both schools are believed to be compatible because they share the notions of theory as important to understanding, the fallibility of human knowledge, the indetermination of theory by fact, and a set of standards regarding how the inquiry process should be managed. The two paradigms are also united by a joint commitment to comprehending and improving human lives through spreading practical knowledge, and by a shared commitment to rigor, meticulousness, and critique during the research process (Reichardt & Rallis, 1994). In essence, this viewpoint argues that qualitative and quantitative methods should be viewed as a continuum of research with specific goals based on the objective or question of the research itself.

Clarke and Yaros (1988) make the further point that combining research methods can be very useful in certain fields due to the complexity of the phenomena under study. HR, for instance, requires data from a large number of different perspectives. Thus, HR and many other fields in the social sciences may require a wide range of qualitative and quantitative techniques to fully answer their research questions.

Practical Differences between Qualitative and Quantitative Research Methods

Fourth, many philosophers of science claim that there should not be a preoccupation with the quantitative-qualitative debate because it is the type of debate that can never really be resolved and is unrealistic to expect resolution anytime in the future. Moreover, a preoccupation with epistemological purity does not solve research questions or get research done (Creswell, 2003).

In essence, none of these arguments actually addresses the underlying assumptions behind the differences between qualitative and quantitative research approaches. Reichardt and Rallis (1994) recognize the possibility of a severe clash between the two schools of thought, especially regarding their approach to the nature of reality. The two paradigms are indeed incompatible if the qualitative model presupposes that there are no external referents to explain reality — yet this philosophy is also incompatible with many of the qualitative methods themselves.

Howe (1992) made a rather convincing case for combining qualitative and quantitative methods by challenging the underlying assumptions of both paradigms. Howe states that both paradigms are actually grounded in the principles of positivism and that qualitative research is not solely based on constructivism or interpretivism. His argument denied any either-or selection between the two schools, proposing instead that both types of researchers should accept a positivism that includes a certain measure of interpretivism. This option is based on the critical social research model, which steers clear of the positivist/interpretivist division in support of the compatibility of both views.

Bryman (2006) examined 232 articles that combined both paradigms in social science research using mixed methods. It was found that when mixed-methods paradigms were followed, the most frequently used quantitative method was a self-administered questionnaire, followed by a structured interview. The most common qualitative method was a semi-structured interview, distantly followed by participant observation. This finding appears to support Howe's (1992) notions.

Moreover, a surprising finding was that the rationale for using mixed-methods designs was not well thought out in the vast majority of the studies reviewed. In only ten of the 232 reviewed studies was there a clear indication that qualitative and quantitative research had each been designed to answer specific research questions addressed by the study. This raises the concern that many researchers employing mixed-method designs do so in an attempt to be "trendy" or to preemptively address critics of either paradigm alone — which is not a productive basis for useful research.

There are several important take-home messages here for future HR research:

1. Neither paradigm is gospel in research circles. Neither of the philosophical assumptions on which these paradigms are based are invariant nor are they omnipotent.

2. Both paradigms make use of quantitative and qualitative data and methods to understand their findings.

3. Both paradigms actually seek to answer questions from different viewpoints that are equally important for understanding.

4. The choice of paradigm is dependent on the research question one wishes to address in a particular study.

5. Mixed-methods designs should be employed based on their appropriateness for the research questions the researcher wishes to address, not for methodological fashion.

Anderson, V. (2004). Research methods in human resource management. London, UK: Chartered Institute of Personnel and Development.