It is quite obvious that the reliability of the facts and conclusions obtained by the researcher depends on how the latter came to these facts and conclusions, that is, on the method he or she used. Until recently, the philosophy and methodology of science were dominated by a simplified view of the logic and procedure of scientific research, which can be designated as the “traditional image of science.” This view ignored the complex relationships between the most general philosophical and theoretical concepts included in a specific research program, and more specific assumptions, from which the scientist planning a specific research explicitly or implicitly proceeds. The “traditional image of science” presented the research process as a simple linear sequence (Foster, Ghani, Jarmin, Kreuter, & Lane, 2016):
- a testable general theory from which the main theoretical hypothesis is derived,
- the definition of basic theoretical concepts in terms of specific operations on measure, that is, their operationalization,
- a decisive experiment, leading to the unambiguous acceptance or rejection of the hypothesis, and at the same time the general theory from which the hypothesis was derived. It was assumed that the negative result of the “decisive experiment” ‑ the empirical demonstration of the falsity of the predictions derived by logical deduction from theoretical premises, testifies to the falsity of these premises.
In practice, however, it is always possible to attribute an “unsuccessful” observation either to unaccounted for peculiarities of the initial conditions of a particular experiment, or to the falsity of many auxiliary hypotheses and assumptions used to test the main theoretical hypothesis. In particular, a negative result can always be attributed not to a testable hypothesis but to artifacts of the method used or errors in operationalization and measurement of individual indicators. Thus, there is a significant discrepancy between the above-described “traditional way of science” and the real logic of research. This discrepancy becomes especially evident when looking at the social sciences for the following reasons (Zhao, Ross, Li, & Dennis, 2021):
- There are relatively little developed formal theories from which it would be possible to deduce testable hypotheses in a rigorous way, and for each of such hypotheses, already at the moment of its presentation, many counterexamples can be found;
- The possibilities of the experimental method are obviously limited, and the available data on naturally occurring events or on the results of special surveys rarely allow separating the main and side effects;
- There are simultaneously several highly influential research programs (for example, behavioral, interpretive, and structuralist), each of which has its own set of methodological norms, favorite research techniques, and exemplary theoretical interpretations.
In this context, it should be noted that sociology is a science that, relying on empirically confirmed data, studies the activities of people in a specific social and cultural context of the functioning of society, its institutions and organizations. It implies investigation of the contradictory development of social consciousness, conscious and unconscious motives of behavior (Connelly Playford, Gayle, & Dibben, 2016). Empiricism in sociology is a principle of sociological research, which was formed simultaneously with the justification of sociology as an independent science based on a specific analysis of social facts (Baral, 2017; Daniel, 2016). Currently, the principle of empiricism is increasingly linked to the theoretical analysis of social problems.
Social processes, from the very first steps in the formation of science, trying to understand the deep connections of objective reality, were considered as the subject field of sociology. In the modern world, the accumulation of quantitative and qualitative changes, their irreversibility and ambiguity, the tightness of time and social space represent such an interweaving of interrelated social processes, the study of which can find its positive conclusion only if there is a clear methodological base (Chew, 2019). Namely in this fundamental issue, researchers find themselves in a situation where there are no clear reference points of cognition adequate to complex social reality (Ramlagan et al., 2021). Today, the methodology of sociological science is represented by three levels: general scientific, which determines approaches to research; general sociological, represented by theoretical sociology and subject methodology, based on the applied logic of private sociological theories. Over time, fundamental sociology has transformed from knowledge that performs descriptive, analytical, and instrumental functions of social life cognition into a self-sufficient theoretical system that formulates the principles and laws of cognition of the objective domain of the world and a method of this cognition.
Research in the field of sociology is a system of consistent and logical methodological and organizational-technical procedures, interconnected by the common goal of obtaining objective, reliable data about the process or phenomenon under study for their subsequent study in practice. Sociological research aimed at solving any practical problems, with the aim of developing scientifically grounded forecasts about the development trends of certain social phenomena or processes, is called applied. Research aimed primarily at the development of scientific theories is fundamental (Du Plessis, 2019; O’Brien et al., 2016). Study focused on collecting and analyzing data using methods and techniques of sociological research is called empirical and can be carried out within the framework of both fundamental and applied sociology, depending on what goal it sets for itself. Accordingly, it defines the choice of study design, based, first of all, on the overall research philosophy.
Sociological research differs from social examinations primarily in terms of purpose. In the first case, the goal of a specialist’s work is knowledge as a self-sufficient value; in the second case, the value of the results achieved is determined by their information content and usefulness to society. Accordingly, in the first case, reliability is of the biggest importance, in the second – informational content. Sociological doctrines, sociological examinations and observations, and sociological research – the three main methodological styles in sociology – characterize the preferential orientations of researchers and rarely exist in pure form (Druckman & Donohue, 2019). Even the most abstract sociological doctrines contain references to facts and circumstances that illustrate the author’s thought, and this determines the expedience of combining methods. The purpose of sociological methodology is to substantiate generalizations – general judgments about a certain area of reality.
The concept of “experiment” involves the study of some phenomenon in a controlled environment. The main task of the experiment is to study one specific phenomenon in order to isolate factors that, in the general case, also affect the object, but are not of interest to the researcher within the framework of the scientific task set by him or her. Thus, Galileo’s experiments on measuring the speed of falling bodies showed that heavy and light objects fall at the same speed, thereby paving the way for the theory of gravity, but abstracting from other phenomena, such as air resistance or friction forces, which also affected objects experiment (Flynn, Kramer, & Laher, 2019). In this sense, experiments in the social sciences are akin to physical ones. Both, controlling the possible impact on the object of research of transient factors that are not of interest from the point of view of the theory being tested, were originally intended to test the theoretical hypotheses put forward by it in theoretically unambiguous circumstances.
A laboratory, or true, experiment is aimed at testing a theoretical hypothesis and is carried out under conditions of maximum control over the level of the independent variable influence and purification (isolation) of this effect from extraneous influences exerted by external variables. These variables are irrelevant from the position of the hypothesis under testing. Experimental control and isolation allow one to reject other possible explanations for the observed effect – competitive hypotheses (Das, Tarafder, & Nahar, 2016). “An important condition for the validity and reliability of the results obtained in a laboratory experiment is the possibility of a sufficiently reliable measurement of the dependent variable” (Baldassarri & Abascal, 2017, p. 56). In this case, with an infinite number of tests, the results of unavoidable random perturbations in the dependent variable will “cancel out” each other and the researcher will receive an accurate assessment of the effect of interest.
In practice, the described requirements for a true experiment can be fully implemented only in an infinite ideal experiment, during which the external, so-called exogenous variables remain unchanged, and only the independent variable changes. This fact ensures the complete validity of the conclusions about the studied “relationship between the independent and dependent variables” (Devlin, 2017, p. 42). An ideal experiment is a standard against which real experiments can be evaluated and compared, but literal fulfillment of all its requirements is usually impossible or even meaningless from the position of a particular scientific task facing the researcher (Bruggemann & Bizer, 2016). An ideal, that is, a perfectly valid experiment, fixes only the relationship between the variables that the experimenter plans to study, and “cuts off” any other sources of systematic variation in the results. The validity of the experiment, therefore, determines the conclusions reliability concerning the existence or lack of the alleged causal relationship and about the confirmation or non-confirmation of the theoretical hypothesis being tested in frames of the experiment.
The experimental approach allows expanding the empirical base of research. Indeed, today, social science can no longer afford to be limited only by data from outside observations ‑ observational studies; however, they certainly were and remain necessary. Thus, in economics, the “canonical” source is the data of statistical reporting, and in sociology ‑ the results of representative opinion polls. At the same time, both sciences are certainly not limited to these sources. Sociologists actively use focus groups, various types of interviews, participant observation, and in questionnaires, they have not limited themselves to clarifying the current state of affairs for a long time, but use, for example, projective methods and experimental questions to obtain information on topics, the answers to which may be sensitive for respondents.
Economists, for their part, have long mastered poll data to determine agent preferences. These methods are used, for example, in such cases as expert assessments of the degree of efficiency of state institutions in different countries, identification of consumer preferences when choosing a type of urban transport, expectations of business leaders regarding the upcoming economic situation, and many others. Moreover, both in terms of research methods and methods of processing their results (regression and factor analysis, multilevel models, nonparametric and Bayesian methods, and so on), the approaches of modern economists, sociologists, and political scientists do not fundamentally differ (Salganik, 2019). Representatives of the social sciences who study the same problem see it from different angles set by the specifics of their “picture of the world” but basically have the same understanding of what a scientific hypothesis is and how it should be related to theory, how to test it, as well as what statistical models and methods should be used for this verification (Creswell & Creswell, 2018). Thus, in modern social sciences, there is a convergence both in subjects and in research methods.
An important methodological problem of social psychology is to consider the structure of an experiment as a way to test a hypothesis. In an experimental study, the subject of which is social and socio-psychological phenomena, at certain stages such methods as interviews, conversation, observation, questionnaires, tests can be used (Altmejd et al., 2019). The problems of using laboratory and natural experiments in socio-psychological research, their correlation and validity are differently solved by different social psychologists. Various authors, stating the widespread use of laboratory experiments in socio-psychological studies of small groups, indicate that the facts obtained in this way have a low “ecological” validity, weakly correlate with the social context of behavior and activity (Bruggemann & Bizer, 2016). Supporters of the opposite point of view believe that a laboratory experiment provides additional opportunities for more complete control over variables and isolation of the studied psychological phenomenon “in its pure form” (Goodman, 2017). American specialists of this profile also emphasize that a laboratory experiment is capable of providing unambiguous evidence of causality, more fully controlling external variables, and also measuring the values and parameters of complex experimental variables, that is, a laboratory experiment has certain advantages that cannot be ignored (Altmejd et al., 2019). The answer, apparently, is not whether or not to use a laboratory experiment, but what goals should be achieved and how to apply its results, what circumstances and techniques allow validating the “data obtained in a laboratory experiment” (Altmejd et al., 2019, p. 10). The study of joint activities and group behavior in stressful and extreme conditions requires a combination of laboratory and natural experiments, since full-scale reproduction of extreme and stressful conditions is not acceptable to researcher for ethical reasons.
A number of practical experimental techniques exist and are applied in the field of social sciences. First, there are natural experiments, when a certain process occurs by itself, and the researcher only observes what people do in control and experimental conditions. An example is experiments from the field of the “theory of broken windows”: researchers measured, depending on the degree of litter in a given area of the city, the willingness of ordinary residents on the street to throw garbage on the ground, and not in a trash can (Devlin, 2017). This approach is supported by field experiments, which have become widespread in the last decade, particularly in political sciences.
The data that are collected during the experiments are usually not representative, and the issue of representativeness is usually not even considered a condition for collecting data of acceptable quality. Accordingly, from the standpoint of generally accepted sociological approaches, the findings cannot be extended to broader populations than the participants in the experiment, whose number, usually, is not more than several hundred people. Finally, the results of most experiments are not robust tested and do not exhibit external validity (Pajo, 2017). For this reason, it is often difficult to say whether the results of an experiment conducted a month or a year later with another similar group of participants will differ significantly from those obtained in the current experiment.
The solution to this problem and, accordingly, obtaining better information should be sought along the path of “theorizing” experiments. For example, in the formation of hypotheses and ideas about the motives of human behavior, one should use knowledge about social and other determinants of behavior. Different methods of collecting and analyzing data on current social phenomena and processes should not compete but complement each other. Thus, it becomes necessary to use also non-experimental research methods.
One of the popular methods of non-experimental research is a longitudinal study. This study involves “sequential multiple registration of certain indicators at strictly set intervals in order to determine the dynamics of their change and mutual influence” (Creswell & Clark, 2017, p. 40). Longitude research involves the long-term study of one set of persons. Initially, longitudinal research or the method of “longitudinal sections” was formed in child and developmental psychology, where it was used as a kind of alternative to the method of “cross sections.” They differ in the general concept, the number and composition of participants, the scale of the event, but all of them are aimed at understanding how the life or actions trajectories of the participants change, depending on various factors.
Later, longitudinal research penetrated into other studies, in particular in sociology, where it is understood as a non-experimental f research in frames of which data is collected from the same sample at more than one time point. Since the same sample population is observed for a long time, the data obtained in different observations are rightfully interpreted as changes in time, characterizing the same person or object, and not as differences characterizing different samples, albeit obtained on the same object (Reio, 2016; Swart, Kramer, Ratele, & Seedat, 2019). Moreover, monitoring is used, which is a way of observing, evaluating and predicting the state of an object, the development of a phenomenon for a sufficiently long time according to the same system of indicators and methods. Monitoring is currently used in the technical, natural, and social sciences.
Monitoring of a process or phenomenon is possible if two conditions are met: the monitored phenomenon or process changes over time; the phenomenon has acquired a massive or threatening significance. Since the phenomenon or event under study is dynamic, two or more current measurements can be applied for forecasting its development in the near or distant future (Ross, 2019). In addition, monitoring the dynamics of an event allows to compare several different points with each other, trace the trend or pattern of change, and apply sophisticated analytical and graphical tools to depict the phenomenon over time.
One of the classic non-experimental methods widely used in modern social science is the case study method. In general, “case analysis” can be defined as an in-depth sample study of a problem at one single, but representative object (Cooper & Meltzoff, 2017). At the same time, the subject of research and its relationship with the object are studied with particular care. Case study is a form of qualitative descriptive research, the object of which is an individual or a small group. Its subject can only be the real interaction of a foreseeable number of people and only in a very specific context. Sociologists who practice this method do not strive for global generalizations, the discovery of causal laws, or statistically representative information (Flynn et al., 2019; Pearce, 2016). Here one event or one community is studied in all details. The essence of such a study is the situation that, having studied in detail one or several cases, to reveal the content of the deep processes going on in society.
The difficulties and limitations of using the case study method are as follows: 1) Representativeness – as well as the sample population, a single case must be comparable with the general population; 2) Limited use of accurate quantitative methods. The advantages of the case study method lie in the variety of information content, as well as the accessibility of the method. The case study approach was used in their research by such scientists as Hans Jurgen Eysenck, Charles Ragin and Howard Becker, John Walton and others, who were convinced that this particular research method is the best source of the theory (Creswell & Clark, 2017). The main advantages of case study include the following features (Ross, 2019):
- The studied phenomenon, as a rule, is an urgent scientific problem that is not supported by sufficient knowledge at the time of research;
- The integrity of scientific research, which has a theoretical justification for the formulation of a problematic problem, the formation of a hypothesis, its verification, modification and generation of new knowledge;
- The cognitive integrity of cognition, expressed in the empirical (collection and primary analysis of data) and theoretical (associated with the interpretation of results) nature of the research;
- Flexible and adaptive methodological structure that allows making the necessary adjustments to the research process;
- A holistic approach aimed at studying a holistic phenomenon that is inseparable from the context allows applying case studies to the study of complex systems;
- Taking into account contextual factors makes it possible to study the dynamics of phenomena, taking into account changes in the environment and the consequences of the influence of various factors;
- Applicability of research results for further research by other scientists;
- Providing unique knowledge about a person, organizations, social and political phenomena;
- Obtaining detailed information about latent processes and mechanisms of social relations.
The limitations of case studies can be overcome by using quantitative comparative approach. A quantitative non-experimental causal comparative research design can be taken to answer the research questions proposed and to compare independent variable influencing security behaviors of IT professionals. The research question for this study is “To what extent do internal factors, external factors, or inherent factors influence the security awareness of the end user?” There are three hypotheses to be tested during this study:
- H01: Internal factors do not have more influence on end user security awareness than external or inherent factors.
- Ha1: Internal factors have more influence on end user security awareness than external or inherent factors.
- H02: External factors do not have more influence on end user security awareness than internal or inherent factors.
- Ha2: External factors have more influence on end user security awareness than internal or inherent factors.
- H03: Inherent factors do not have more influence on end user security awareness than internal or external factors.
- Ha3: Inherent factors have more influence on end user security awareness than internal or external factors.
Comparison is a logical tool necessary in any cognitive activity: at its various stages and levels, regardless of its object; the comparative method is a narrower concept. Comparison can be used as a special research method only when the comparison procedure requires – for its effective implementation – special training and organization. Such a need usually arises when comparing complex objects and phenomena that are described by a large set of widely varying features.
The isolation of objects, processes, and phenomena is of relative nature. Whatever originality they may have, between them there is always a certain commonality, openness to each other, interdependence. Each object is an element of a large system, where everything is interconnected. Changes in one part of the system inevitably entail corresponding changes in other parts of it. Therefore, analysis’ objective is not so much to study the essence of isolated objects, but to find – as far as possible – connections between separate objects (Lahmami, 2020). Thus, when conducting a comparative analysis, any area is considered not in a narrowly subject matter, but systematically, that is, in the interaction of various objects or components of the system (Lahmami, 2020). The methodological foundations of comparative analysis are concentrated on identifying the nature of connections, patterns of interaction between objects and socio-economic, cultural, and other phenomena. At the same time, the analytical approach naturally develops into a synthetic, systemic one, which allows creating a picture of reality more adequately, tracing the cause-effect relationships that most often lie behind the narrow framework of a limited object (Gross, 2018). However, there are several limitations to this method that should be noted. The following main problems can be distinguished when evaluating objects according to many criteria (Hothersall, 2017):
- Inconsistency of criteria: improvement in one criterion usually leads to worsening in some other criteria;
- Impossibility of analytical – in the form of formulas – expression of links between assessments according to different criteria;
- Grades according to various criteria have a different form: numerical, meaningful (“excellent,” “good,” “yes-no,” and so on), point-based, in the form of rankings or other. In the general case, non-numerical data is understood as elements of spaces that are not linear or vector, in which there are no operations of addition of elements and their multiplication by a real number;
- Numerical estimates differ in dimension, they correspond to different physical quantities and are measured in different units. Moreover, there is difference in direction, as some criteria need to be minimized, others – to be maximized, in the range of values;
- The difference in the importance of criteria.
The main way to remove these problems in the process of evaluating objects is to identify and take into account the subjective judgments of an expert. Usually, the following information is required from a person (Kaushik & Walsh, 2019):
- A list of compared objects;
- A list of criteria by which the comparison will be carried out;
- Evaluation of objects by criteria;
- Judgments about the importance of the criteria – that is, information about which criteria are more important, which ones are less important;
- Restrictions on individual criteria;
- Judgments about the degree of admissibility of lagging by individual criteria, about compensation of some criteria by others.
For each object, a certain generalized score is calculated, which takes into account the scores for all criteria. The result of comparing objects should be some ordered sequence of them, placing the objects in the order of their preference. According to the principle of reducing object assessments to a single assessment, several classes of methods can be distinguished. However, for the above research question, methods based on the calculation of generalized assessments (generalized criterion) seem to be preferable, since it is necessary to assess the influence of three types of factors on overall security awareness of the end user. The principle of these methods is to calculate a generalized score for each of the objects based on their scores for individual criteria.
It should be noted, however, that there are ethical limitations of the above method. First of all, this is the potential subjectivity of expert assessments when ranking, as well as a high probability of distortion of the results due to changes in external factors that cannot be influenced during the study. It is not possible to completely avoid the influence of these restrictions, therefore, this should be taken into account and indicated both when describing the research results and when outlining directions for future research. In addition, triangulation is a useful tool for increasing the level of validity of a non-experimental study.
With regard to the possibilities of applying experimental methods, it is preferable to use experimental research methods, for example, to study the results of programs to reduce the consumption of medicinal opioids in the United States in the framework of the fight against the current opioid crisis. The main hypothesis of such a study is the following: programs to reduce the use of opioid analgesics help to diminish the consumption of opioids on a national scale, both for medical and non-medical use, as well as facilitate improvement of the situation with an opioid crisis dissemination in the country. The experimental nature of the research in this case is determined by the purely practical orientation of the methods for solving the existing acute social problem, the causes of which have already been sufficiently studied and known. Among these reasons, the main place is occupied by the narrow economic interests of pharmaceutical companies that conduct active advertising campaigns for opioid painkillers. It is combined with the problem of insufficient competence and excessive workload of doctors (Liamputtong, 2019). The preferred method in this case is a laboratory-field experiment.
Field experiments, due to the greater vagueness and ambiguity of the results, explanations of the relationships are often less good than laboratory ones. However, they can serve to confirm the results of laboratory experiments and to establish how much the sign tested in the laboratory acts in a real situation (Gallifa, 2018). This allows experiment, intervention, and field research to be combined. In the above-described research case, the impact of the aforementioned programs in different population groups under real conditions is examined.
However, the ethical limitations of the method discussed above should also be noted. When pilot studies are conducted to examine the effectiveness and efficiency of opioid crisis programs, the ethical implications are even more obvious, especially as it concerns vulnerable populations. Researchers should ensure compliance with the requirements for research involving human participants, in ethical, legal, and regulatory areas, as well as applicable international requirements. Special attention should be paid to the various types of possible harm and benefits, as well as issues with the choice of the location of the study, which arise in conditions when many factors must be considered when assessing the benefit-harm ratio. These include poverty, lack of adequate access to health care due to lack of insurance, gender and ethnic inequality and other types of vulnerability.
In connection with the above, the systematic use of the advantages of qualitative and quantitative methodologies in experimental and non-experimental studies allows obtaining more reliable data. At the same time, the use of various methods of collecting and analyzing information makes it possible to outline the most significant points of the study. This is provided by triangulation – the use of versatile methods in the analysis of empirical events. Since each method involves a different line of action in relation to reality and each method allows discovering its different aspects, it is necessary to use different methods to allow these various aspects to manifest. Non-experimental and experimental methods can be applied in a longitudinal way, that is, for a long time, ad this is the most preferable for social science, psychology, and economics, due to high dynamics of influencing factors and environment. In conclusion, one can stress that in today development of science, there is changes of not only paradigms, concepts, and theories, but research methods themselves. They lose the speculative, asserting nature, become constitutive, shaping, and, often, transforming, shifting towards constructionism philosophy and paradigm. Thus, the development of the methodological arsenal of modern social science consists in a special consolidation of all research methods, when usually not one method is used in research, but a whole arsenal of various methods, which, intertwining and controlling each other, are intercomplementary.
References
Altmejd, A., Dreber, A., Forsell, E., Huber, J., Imai, T., Johannesson, M., Camerer, C. (2019). Predicting the replicability of social science lab experiments, PLoS One, 14(2), 1-18. Web.
Baldassarri, D., & Abascal, M. (2017). Field experiments across the social sciences. Annual Review of Sociology, 43, 41-73. Web.
Baral, U. (2017). ‘Research data’ in social science methods. Journal of Political Science 17, 82-104. Web.
Bruggemann, J., & Bizer, K. (2016). Laboratory experiments in innovation research: A methodological overview and a review of the current literature. Journal of Innovation and Entrepreneurship, 5(24), 1-13. Web.
Chew, B.-H. (2019). Planning and conducting clinical research: The whole process. Cureus, 11(2), e4112. Web.
Connelly, R., Playford, C., Gayle, V., & Dibben, C. (2016). The role of administrative data in the big data revolution in social science research. Social Science Research, 59, 1-12. Web.
Cooper, H. M., & Meltzoff, J. (2017). Critical thinking about research: Psychology and related fields (2nd ed.). Washington, D.C.: American Psychological Association.
Creswell, J. W., & Clark, V. (2017). Designing and conducting mixed methods research. Newbury Park, CA: SAGE Publications.
Creswell, J. W., & Creswell, J. D. (2018). Research design: Qualitative, quantitative, and mixed methods approaches (5th ed.). Newbury Park, CA: SAGE Publications.
Daniel, E. (2016). The usefulness of qualitative and quantitative approaches and methods in researching problem-solving ability in science education curriculum. Journal of Education and Practice, 7(15), 91-100. Web.
Das, R., Tarafder, V., & Nahar, N. (2016). Research methods persistent in social science research. International Journal of Innovative Science, Engineering & Technology, 4(10), 5-15. Web.
Devlin, A. (2017). The research experience: Planning, conducting, and reporting research. Newbury Park, CA: SAGE Publications.
Druckman, D., & Donohue, W. (2019). Innovations in social science methodologies: An overview. American Behavioral Scientist, 64(1), 3-18. Web.
Du Plessis, H. (2019). Engaging with the challenges of social science research methods. South African Journal of Science, 115(11/12). Web.
Flynn, A., Kramer, S., & Laher, S. (2019). Transforming research methods in the social sciences. Johannesburg, South Africa: Wits University Press.
Foster, I., Ghani, R., Jarmin, R., Kreuter, F., & Lane, J. (2016). Big data and social science: A practical guide to methods and tools. London, UK: Chapman and Hall.
Gallifa, J. (2018). Research traditions in social sciences and their methodological rationales. Aloma, 36(2), 9-20. Web.
Goodman, S. (2017). How to conduct a psychological discourse analysis. Critical Approaches to Discourse Analysis Across Disciplines, 9(2), 142-153. Web.
Gross, N. (2018). Pragmatism and the study of large-scale phenomena. Theory and Society 47, 87-111. Web.
Hothersall, S. J. (2017). Everyday knowledge: A mixed-methods study using factor analysis and narrative approaches to explore social worker’s knowledge. Social Work and Social Sciences Review, 19(2), 33-64. Web.
Kaushik, W., & Walsh, C. (2019). Pragmatism as a research paradigm and its implications for social work research. Social Sciences, 8, 1-17. Web.
Lahmami, H. (2020). Methodology of social science research: The case of the sociology of action in relation to values. Revista Electrónica Interuniversitaria de Formación del Profesorado, 23(1), 59-73. Web.
Liamputtong, P. (2019). Handbook of research methods in health social sciences. New York, NY: Springer.
O’Brien, B. C., Ruddick, V. J., & Young, J. Q. (2016). Generating research questions appropriate for qualitative studies in health professions education. Academic Medicine, 91, e16. Web.
Pearce, N. (2016). Analysis of matched case-control studies. BMJ, 352. Web.
Pajo, B. (2017). Introduction to research methods: A hands-on approach. Newbury Park, CA: SAGE Publications.
Ramlagan, S., Shean, Y. L., Parker, S., Trollip, K., Davids, A., Reddy, S. P. (2021). Pushing the boundaries: Adapting research methodology to document the COVID-19 pandemic from a socio-behavioural perspective in a low/middle level income country: The case of South Africa. International Journal of Social Research Methodology. Web.
Reio, T. G. (2016). Nonexperimental research: Strengths, weaknesses and issues of precision. European Journal of Training and Development, 40(8/9), 676-690. Web.
Ross, A. S. (2019). Experimental and nonexperimental designs in social psychology. Milton Park, UK: Routledge.
Salganik, M. (2019). Bit by bit: Social research in the digital age. Princeton University Press.
Swart, L.-A., Kramer, S., Ratele, K., Seedat, M. (2019). Non-experimental research designs: Investigating the spatial distribution and social ecology of male homicide. In S. Laher, A. Fynn, & S. Kramer (Eds.). Transforming research methods in the social sciences (pp.19-35). Wits University Press. Web.
Zhao, P., Ross, K., Li, P., & Dennis, B. (2021). Making sense of social research methodology: A student and practitioner centered approach. Newbury Park, CA: SAGE Publications.