2.1 Approaches to Sociological Research

Introduction

When sociologists apply the sociological perspective and begin to ask questions, no topic is off limits. Every aspect of human behaviour is a source of possible investigation. Sociologists question the world that humans have created and live in. They notice patterns of behaviour as people move through the world. Using sociological methods and systematic research within the framework of the scientific method, sociologists have discovered workplace patterns that have transformed industries, family patterns that have enlightened parents, and education patterns that have aided structural changes in classrooms.

Sociologists often begin the research process by asking a question about how or why things happen. It might be a unique question about a new trend or an old question about a common aspect of life. Once a question is formed, a sociologist proceeds through an in-depth process to answer it. Depending on the nature of the topic and the goals of the research, sociologists have a variety of methodologies to choose from. The following sections describe these approaches to acquiring knowledge.

Science vs. Non-Science

Contemporary society is going through an interesting time in which the certitudes and authority of science are frequently challenged. In the context of the natural sciences, people doubt scientific claims about climate change and the safety of vaccines. In the context of the social sciences, people doubt scientific claims about the declining rate of violent crime or the effectiveness of needle exchange programs.  Sometimes there is a good reason to be skeptical about science, when scientific technologies prove to have adverse effects on the environment, for example. Sometimes skepticism has dangerous outcomes, when people act on conspiracy theories and misinformation or epidemics of diseases like measles suddenly break-out in schools due to low vaccination rates. In fact, skepticism is central to both natural and social sciences; but from a scientific point of view, the skeptical attitude needs to be combined with systematic research in order for knowledge to move forward.

In sociology, science provides the basis for being able to distinguish between everyday opinions or beliefs and propositions that can be sustained by evidence. In his paper, The Normative Structure of Science (1942/1973), the sociologist Robert Merton argued that science is a type of empirical knowledge organized around four key principles, often referred to by the acronym CUDOS.

For Merton, therefore, non-scientific knowledge is knowledge that fails in various respects to meet these criteria. Types of esoteric or mystical knowledge, for example, might be valid for someone on a spiritual path, but because this knowledge is passed from teacher to student through direct transmission and it is not available to the public for open debate, or because the validity of this knowledge might be specific to the individual’s unique spiritual configuration, esoteric or mystical knowledge is not scientific per se. Claims that are presented to persuade (rhetoric), to achieve political goals (propaganda, of various sorts), or to make profits (advertising) are not scientific because these claims are structured to satisfy private interests. Propositions which fail to stand up to rigorous and systematic standards of evaluation are not scientific because they can not withstand the criteria of organized skepticism and scientific method.

The basic distinction between scientific and common, non-scientific claims about the world is that in science “seeing is believing” whereas in everyday life “believing is seeing” (Brym, Roberts, Lie, & Rytina, 2013). Science is, in crucial respects, based on systematic observation following the principles of CUDOS. Only on the basis of observation (or “seeing”) can a scientist believe that a proposition about the nature of the world is correct. Research methodologies are designed to reduce the chance that conclusions will be based on error. In everyday life, the order is typically reversed. People “see” what they already expect to see or what they already believe to be true. They do not systematically test what they believe to be true. Prior intellectual commitments or biases predetermine what people observe and the conclusions they draw.

Many people know things about the social world without having a background in sociology. Sometimes their knowledge is valid; sometimes it is not. It is important, therefore, to think about how people know what they know, and compare it to the scientific way of knowing. Four types of non-scientific reasoning are common in everyday life: knowledge based on casual observation, knowledge based on selective evidence, knowledge based on overgeneralization, and knowledge based on authority or tradition.

Many people know things simply because they have experienced them directly. Someone who has grown up in Manitoba has probably observed what plenty of kids learn each winter, that it really is true that one’s tongue will stick to metal when it is very cold outside. Direct experience may provide accurate information, but only if the observer is lucky. Unlike the fictional detective Sherlock Holmes, in general, people are not very careful observers. In this example of the “winged ship” in Figure 2.4, the observation process is not deliberate or systematic. Instead, the observers come to know what they believe to be true through casual observation. The problem with casual observation is that sometimes it is right, and sometimes it is wrong. Without any systematic process for observing or assessing the accuracy of observations, a person can never really be sure if their informal observations are accurate.

Many people know things because they overlook disconfirming evidence. Suppose a friend declared that all men are liars shortly after she had learned that her boyfriend had deceived her. The fact that one man happened to lie to her in one instance came to represent a quality inherent in all men. But do all men really lie all the time? Probably not. If the friend is prompted to think more broadly about her experiences with men, she would probably acknowledge that she knew many men who, to her knowledge, had never lied to her and that maybe even her boyfriend did not generally make a habit of lying. This friend committed what social scientists refer to as selective observation by noticing only the pattern that she wanted to find at the time. She ignored disconfirming evidence. If, on the other hand, the friend’s experience with her boyfriend had been her only experience with a man, then she would have been committing what social scientists refer to as overgeneralization, assuming that broad patterns exist based on very limited observations.

Another way that people claim to know what they know is by looking to what they have always known to be true. There is an urban legend about a woman who for years used to cut both ends off of a ham before putting it in the oven (Mikkelson, 2005). She baked ham that way because that is the way her mother did it, so clearly that was the way it was supposed to be done. Her knowledge was based on a family tradition or traditional knowledge. After years of tossing cuts of perfectly good ham into the trash, however, she finally asked her mother why she did it and learned that the only reason her mother cut the ends off ham before cooking it was that she did not have a pan large enough to accommodate the ham without trimming it.

Without questioning what one thinks one knows is true, one may wind up believing things that are actually false. This is most likely to occur when an authority tells us that something is true, a case of authoritative knowledge. Mothers are not the only possible authorities people might rely on as sources of knowledge. Other common authorities people might rely on are political figures, churches and ministers, media influencers and social media networks. Although it is understandable that someone might believe something to be true if they look up to, or respect the person who has said it is so, this way of knowing differs from the sociological way of knowing. Whether quantitative, qualitative, or critical in orientation, sociological research is based on the scientific method.

The Scientific Method

Sociologists make use of tried-and-true methods of research, such as experiments, surveys, field research, and textual analysis. But humans and their social interactions are so diverse that they can seem impossible to chart or explain. It might seem that science is about discoveries and chemical reactions, or about proving hypotheses about elementary particles right or wrong, rather than about exploring the nuances of human behaviour. However, this is exactly why scientific models work for studying human behaviour. A scientific process of research establishes parameters that help make sure results are objective and accurate. Scientific methods provide limitations and boundaries that focus a study and organize its results. This is the case for both positivist quantitative methodologies, which seek to translate observable phenomena into unambiguous numerical data, and interpretive qualitative methodologies, which seek to translate observable phenomena into definable units of meaning. The social scientific method in both cases involves developing and testing theories about the world based on empirical (i.e., observable) evidence. The social scientific method is defined by its commitment to systematic observation of the social world, and it strives to be objective, critical, skeptical, and logical. It involves a series of established steps known as the research cycle.

No matter what research approach is used, researchers want to maximize the study’s reliability and validity. Reliability refers to how likely research results are to be replicated if the study is reproduced. Reliability increases the likelihood that what is true of one person will be true of all people in a group. Validity refers to how well the study measures what it was designed to measure. If the researcher wishes to examine people’s depth of religiosity — i.e., how strong is someone’s religious belief? How central is religious belief to their life? — does a measure like “frequency of church attendance” accurately measure that? Maybe not. People attend church for a variety of reasons and some religions are not organized on the basis of churches and congregations.

A subtopic in the field of political violence would be to examine the sources of homegrown radicalization: What are the conditions under which individuals in Canada move from a state of indifference or moderate concern with political issues to a state in which they are prepared to use violence to pursue political goals? The reliability of a study of radicalization reflects how well the social factors unearthed by the research apply to similar individuals who were not directly part of the research. Would another sociologist come up with the same results if they replicated the study? How well can the researcher extrapolate from the research subjects studied to individuals in the broader society? Does research on violent jihadi radicalization apply to violent neo-Nazi radicalization?

Validity ensures that the study’s design accurately examined what it was designed to study. Do the concepts and measures of radicalization accurately represent the actual experience of political radicals? An exploration of an individual’s propensity to plan or engage in violent acts or to go abroad to join a terrorist organization should address those specific issues and not confuse them with other topics such as why an individual adopts a particular faith or espouses radical political views. There is a key difference between religiosity, radicalization, and violent radicalization. As research from the UK and United States on jihadism has in fact shown, while jihadi terrorists typically identify with an Islamic world view, a well-developed Islamic identity counteracts jihadism. Similarly, research has shown that while it intuitively makes sense that people with radical views would adopt radical means like violence to achieve them, there is in fact no consistent homegrown terrorist profile, meaning that it is not possible to predict whether someone who espouses radical views will move on to committing violent acts without taking into account the specific stages in the process (Patel, 2011).

To ensure validity, research on political violence should focus on individuals who engage in political violence and be able to distinguish between simply holding radical political beliefs and acting violently on radical political beliefs. Dalgaard-Nielsen (2010) distinguishes between radical, radicalization, and violent radicalization as follows:

a radical is understood as a person harboring a deep-felt desire for fundamental sociopolitical changes and radicalization is understood as a growing readiness to pursue and support far reaching changes in society that conflict with, or pose a direct threat to, the existing order. Violent radicalization is a process in which radical ideas are accompanied by the development of a willingness to directly support or engage in violent acts.

The scientific method provides a systematic, organized series of steps that help ensure objectivity and consistency in exploring a social problem. These steps provide the means for accuracy, reliability, and validity. Typically, the scientific method starts with these steps, which are described below: 1) ask a question; 2) research existing sources; and 3) formulate a hypothesis.

Step 1: Ask a Question or Find a Research Topic

The first step of the scientific method is to ask a question, describe a problem, and identify the specific area of interest. The topic should be narrow enough to study within a specific geographical location and time frame. “Are societies capable of sustained happiness?” would be too vague. The question should also be broad enough to have universal merit. “What do personal hygiene habits reveal about the values of students at XYZ High School?” would be too narrow. That said, happiness and hygiene are worthy topics to study.

Sociologists do not rule out any topic but would strive to frame these questions in better research terms. That is why sociologists are careful to define their terms. Karl Popper (1902–1994) described the formulation of scientific propositions in terms of the concept of falsifiability (1963). He argued that the key demarcation between scientific and non-scientific propositions was not ultimately their factual truth, nor their verification, but simply whether or not they were stated in such a way as to be falsifiable; that is, whether a possible empirical observation could prove them wrong. If one claimed that evil spirits were the source of criminal behaviour, this would not be a scientific proposition because there is no possible way to definitively disprove it. Evil spirits cannot be observed. However, if one claimed that higher unemployment rates are the source of higher crime rates, this would be a scientific proposition because it is theoretically possible to find an instance where unemployment rates were not correlated to higher crime rates. As Popper said, “statements or systems of statements, in order to be ranked as scientific, must be capable of conflicting with possible, or conceivable, observations” (Popper, 1963).

Once a proposition is formulated in a way that would permit it to be falsified, the variables to be observed need to be operationalized. In a hygiene study, for instance, hygiene could be defined as “personal habits to maintain physical appearance (as opposed to health),” and a researcher might ask, “How do differing personal hygiene habits reflect the cultural value placed on appearance?” When forming these basic research questions, sociologists develop an operational definition; that is, they define the concept in terms of the physical or concrete steps it takes to objectively measure it. The concept is translated into an observable variable, a measure that has different values. The operational definition identifies an observable condition of the concept.

By operationalizing a variable of the concept, all researchers can collect data in a systematic or replicable manner. The operational definition must be valid in the sense that it is an appropriate and meaningful measure of the concept being studied. It must also be reliable, meaning that results will be close to uniform when tested on more than one person. For example, good drivers might be defined in many ways: Those who use their turn signals; those who do not speed; or those who courteously allow others to merge. But these driving behaviours could be interpreted differently by different researchers, so they could be difficult to measure. Alternatively, “a driver who has never received a traffic violation” is a specific description that will lead researchers to obtain the same information, so it is an effective operational definition. Asking the question, “how many traffic violations has a driver received?” turns the concepts of “good drivers” and “bad drivers” into variables which might be measured by the number of traffic violations a driver has received. Of course, the sociologist has to be wary of the way the variables are operationalized. In this example we know that black drivers are subject to much higher levels of police scrutiny than white drivers, so the number of traffic violations a driver has received might reflect less on their driving ability and more on the crime of “driving while black.”

Step 2: Research Existing Sources

The next step researchers undertake is to conduct background research through a literature review, which is a review of any existing similar or related studies. A visit to a university library and a thorough online search will uncover existing research about the topic of study. This step helps researchers gain a broad understanding of work previously conducted on the topic at hand and enables them to position their own research to build on prior knowledge. It allows them to sharpen the focus of their research question and avoid duplicating previous research. Researchers — including student researchers — are responsible for correctly citing existing sources they use in a study or sources that inform their work. While it is fine to build on previously published material (as long as it enhances a unique viewpoint), it must be referenced properly and never plagiarized. To study hygiene and its value in a particular society, a researcher might sort through existing research and unearth studies about childrearing, vanity, obsessive-compulsive behaviours, and cultural attitudes toward beauty. At the literature review stage it is important to sift through this information and determine what is relevant. Using existing sources educates a researcher by showing what others have found relevant about a topic, and helps refine and improve a study’s design.

Step 3: Formulate a Hypothesis

A hypothesis is an assumption about how two or more variables are related; it makes a conjectural statement about the relationship between those variables. It is an educated guess because it is not random but based on theory, observations, patterns of experience, or the existing literature. The hypothesis formulates this guess in the form of a testable or falsifiable proposition. However, how the hypothesis is handled differs between the positivist and interpretive approaches in sociology.

Hypothesis Formation in Quantitative Research

In scientific research, quantitative or positivist approaches are often referred to as hypothetico-deductive methodologies. A hypothesis is derived from a theoretical proposition. On the basis of the hypothesis, a prediction or generalization is logically deduced. If the prediction is confirmed by observation, the theoretical proposition is regarded as valid (at least provisionally). In quantitative studies, the hypothesis predicts how one form of human behaviour influences another. How does being a black driver affect the number of times the police will pull the driver over?

Successful prediction will determine the adequacy of the hypothesis and thereby test the theoretical proposition. Typically, quantitative or positivist approaches operationalize variables as quantitative data; that is, by translating a social phenomenon like health into a quantifiable or numerically measurable variable like “number of visits to the hospital.” This permits sociologists to formulate their predictions using mathematical language, like regression formulas, to present research findings in graphs and tables, and to perform mathematical or statistical techniques to demonstrate the validity of relationships.

Variables are examined to see if there is a correlation between them. When a change in one variable coincides with a change in another variable there is a correlation. This does not necessarily indicate that changes in one variable causes a change in another variable, however — just that they are associated. A key distinction here is between independent and dependent variables. In research, independent variables are the cause of the change. The dependent variable is the effect, or thing that is changed. For example, in a basic study, the researcher would establish one form of human behaviour as the independent variable and observe the influence it has on a dependent variable. How does gender (the independent variable) affect rate of income (the dependent variable)? How does one’s religion (the independent variable) affect family size (the dependent variable)? How is social class (the dependent variable) affected by level of education (the independent variable)?

Note. Typically, the independent variable causes the dependent variable to change in some way.

It is important to note that while there has to be a correlation between variables for there to be a causal relationship, correlation does not necessarily imply causation. The relationship between variables can be the product of a third intervening variable that is independently related to both. For example, there might be a positive relationship between wearing bikinis and eating ice cream, but wearing bikinis does not cause eating ice cream. It is more likely that the heat of summertime causes both an increase in bikini wearing and an increase in the consumption of ice cream.

The distinction between causation and correlation can have significant consequences. For example, Indigenous Canadians are overrepresented in prisons and arrest statistics. As noted in Chapter 1. An Introduction to Sociology, in 2020 Indigenous people accounted for about 5% of the Canadian population, but they made up 30% of the federal penitentiary population (Correctional Investigator Canada, 2020). There is a positive correlation between being an Indigenous person in Canada and being in jail. Is this because Indigenous people are racially or biologically predisposed to crime? No. In fact there are at least four intervening variables that explain the higher incarceration of Indigenous people (Hartnagel, 2004): Indigenous people are disproportionately poor and poverty is associated with higher arrest and incarceration rates; Indigenous lawbreakers tend to commit more detectable “street” crimes than the less detectable and actionable “white collar” crimes of other segments of the population; the criminal justice system disproportionately profiles and discriminates against Indigenous people; and the legacy of colonization has disrupted and weakened traditional sources of social control in Indigenous communities.

At this point, a researcher’s operational definitions help measure the variables. In a study asking how tutoring improves grades, for instance, one researcher might define “good” grades as a C or better, while another uses a B+ as a starting point for good. Another operational definition might describe “tutoring” as “one-on-one assistance by an expert in the field, hired by an educational institution.” Those definitions set limits and establish cut-off points, ensuring consistency and replicability in a study. As the chart shows, an independent variable is the one that causes a dependent variable to change. For example, a researcher might hypothesize that teaching children proper hygiene (the independent variable) will boost their sense of self-esteem (the dependent variable). Or rephrased, a child’s sense of self-esteem depends, in part, on the quality and availability of hygienic resources.

Of course, this hypothesis can also work the other way around. Perhaps a sociologist believes that increasing a child’s sense of self-esteem (the independent variable) will automatically increase or improve habits of hygiene (now the dependent variable). Identifying the independent and dependent variables is very important. As the hygiene example shows, simply identifying two topics, or variables, is not enough: Their prospective relationship must be part of the hypothesis. Just because a sociologist forms an educated prediction of a study’s outcome does not mean data contradicting the hypothesis are not welcome. Sociologists analyze general patterns in response to a study, but they are equally interested in exceptions to patterns.

In a study of education, a researcher might predict that high school dropouts have a hard time finding a rewarding career. While it has become at least a cultural assumption that the higher the education, the higher the salary and degree of career happiness, there are certainly exceptions. People with little education have had stunning careers, and people with advanced degrees have had trouble finding work. A sociologist prepares a hypothesis knowing that results will vary.

Hypothesis Formation in Qualitative Research

While many quantitative sociologists rely on the positivist hypothetico-deductive method in their research, qualitative researchers operate from an interpretive approach. While still systematic, this approach typically does not follow the hypothesis-testing model that seeks to make generalizable predictions from quantitative variables. Instead, an interpretive framework seeks to understand social worlds from the point of view of participants, leading to in-depth knowledge. It focuses on qualitative data, or the meanings that guide people’s behaviour. Rather than relying on quantitative instruments, like fixed questionnaires or experiments, which can be artificial, the interpretive approach attempts to find ways to get closer to the informants’ lived experience and perceptions.

Interpretive research is generally more descriptive or narrative in its findings than positivist research. It can begin from a deductive approach by deriving a hypothesis from theory and then seeking to confirm it through methodologies like in-depth interviews. However, it is ideally suited to an inductive approach in which the hypothesis emerges only after a substantial period of direct observation or interaction with subjects. This type of approach is exploratory in that the researcher also learns as they proceed, sometimes adjusting the research methods or processes midway to respond to new insights and findings as they evolve.

For example, Glaser and Strauss’ (1967) classic elaboration of grounded theory proposed that qualitative researchers working with rich sources of qualitative data from interviews or ethnographic observations need to go through several stages of coding the data before a strong theory of the social phenomenon under investigation can emerge. In the initial stage, the researcher is simply trying to listen carefully and to tentatively categorize and sort the data. The researchers do not predetermine what the relevant categories of the social experience are, but analyze carefully what their subjects actually say. For example, what are the working definitions of health and illness that hospital patients use to describe their situation? In the first stage, the researcher tries to distinguish and succinctly code or label the numerous themes emerging from the data: different ways of describing the experience of health and illness. In the second stage, the researcher takes a more analytical approach by organizing the initial interview data into a few key reoccurring themes: Perhaps these are key assumptions that lay people make about the physiological mechanisms of the body, or the metaphors they use to describe their relationship to illness (e.g., a random occurrence, a battle, a punishment, a message, etc.). In the third stage, the researcher returns to the interview subjects with a new set of questions that would seek to either affirm, modify, or discard the analytical themes derived from the initial categorization of the interview material. This process can then be repeated back and forth until a thoroughly grounded theory is ready to be proposed. At every stage of the research, the researchers are obliged to follow the emerging data by revising their conceptions as new material is gathered, contradictions accounted for, commonalities categorized, and themes re-examined with further interviews.

Once the preliminary work is done and the hypothesis is defined, it is time for the next research steps: choosing a research methodology, conducting a study, and drawing conclusions. These research steps are discussed in the next section (Chapter 2.2: Research Methods).

Step 4: Design and Conduct a Study

Researchers design studies to maximize reliability, which refers to how likely research results are to be replicated if the study is reproduced. Reliability increases the likelihood that what happens to one person will happen to all people in a group or what will happen in one situation will happen in another. Cooking is a science. When you follow a recipe and measure ingredients with a cooking tool, such as a measuring cup, the same results is obtained as long as the cook follows the same recipe and uses the same type of tool. The measuring cup introduces accuracy into the process. If a person uses a less accurate tool, such as their hand, to measure ingredients rather than a cup, the same result may not be replicated. Accurate tools and methods increase reliability.

Researchers also strive for validity, which refers to how well the study measures what it was designed to measure. To produce reliable and valid results, sociologists develop an operational definition, that is, they define each concept, or variable, in terms of the physical or concrete steps it takes to objectively measure it. The operational definition identifies an observable condition of the concept. By operationalizing the concept, all researchers can collect data in a systematic or replicable manner. Moreover, researchers can determine whether the experiment or method validly represent the phenomenon they intended to study.

A study asking how tutoring improves grades, for instance, might define “tutoring” as “one-on-one assistance by an expert in the field, hired by an educational institution.” However, one researcher might define a “good” grade as a C or better, while another uses a B+ as a starting point for “good.” For the results to be replicated and gain acceptance within the broader scientific community, researchers would have to use a standard operational definition. These definitions set limits and establish cut-off points that ensure consistency and replicability in a study.

Step 5: Draw Conclusions

After constructing the research design, sociologists collect, tabulate or categorize, and analyze data to formulate conclusions. If the analysis supports the hypothesis, researchers can discuss the implications of the results for the theory or policy solution that they were addressing. If the analysis does not support the hypothesis, researchers may consider repeating the experiment or think of ways to improve their procedure.

However, even when results contradict a sociologist’s prediction of a study’s outcome, these results still contribute to sociological understanding. Sociologists analyze general patterns in response to a study, but they are equally interested in exceptions to patterns. In a study of education, a researcher might predict that high school dropouts have a hard time finding rewarding careers. While many assume that the higher the education, the higher the salary and degree of career happiness, there are certainly exceptions. People with little education have had stunning careers, and people with advanced degrees have had trouble finding work. A sociologist prepares a hypothesis knowing that results may substantiate or contradict it.

Sociologists carefully keep in mind how operational definitions and research designs impact the results as they draw conclusions. Consider the concept of “increase of crime,” which might be defined as the percent increase in crime from last week to this week, as in the study of Swedish crime discussed above. Yet the data used to evaluate “increase of crime” might be limited by many factors: who commits the crime, where the crimes are committed, or what type of crime is committed. If the data is gathered for “crimes committed in Houston, Texas in zip code 77021,” then it may not be generalizable to crimes committed in rural areas outside of major cities like Houston. If data is collected about vandalism, it may not be generalizable to assault.

Step 6: Report Results

Scientific results from studies are dispersed through a systematic and structured process aimed at ensuring accuracy, transparency, and accessibility of data within the scientific community. The two most common ways researchers typically disseminate their results are through academic peer-reviewed journals and conferences. For scholarly articles, this rigorous process typically begins with researchers compiling their findings into a manuscript format following the guidelines of a scientific journal. Once the manuscript is submitted, it undergoes a peer-review process which involves experts within the field assessing the manuscript’s overall quality, validity and/or reliability of the findings, and its contributions to the scientific field. They provide constructive feedback to the authors, recommending acceptance, revision, or rejection based on their evaluation. This critical examination process helps maintain scientific rigor and ensures that only high-quality research with meaningful contributions advances to publication.

Similarly, scientific results can be reported through conference presentations. Researchers submit abstracts of their study to conferences, which are reviewed by organizers of the conference. Accepted abstracts are presented orally at the conference, providing a platform for researchers to share their findings directly with colleagues, receive feedback, and engage in relevant discussions. Results from both journal articles and conferences are subjected to the scrutiny of other sociologists in the field. Before the conclusions of a study become widely accepted, the studies are often repeated in the same or different environments. In this way, sociological theories and scientific knowledge develop as the relationships between social phenomenon are established in broader contexts.