Here we are again with our usual second installment of this series, where we explore the perspectives and approaches specific to the epistemology of the social sciences. Sociology falls within this field of study and analysis, which, through its methods, shares some of its reflections with other scientific disciplines interested in social dynamics, such as political science, anthropology, or history.
Therefore, this episode aims to illustrate the characteristics of quantitative methods, their strengths, and their critical points. In these, the research methodology will favor approaches that begin with questions about quantum and lead to the definition of numbers, percentages, and more generally quantifiable objects. In contrast, in qualitative methods—which we will explore shortly—the methodology will follow paths typical of in-depth investigations, aimed at analyzing the quality of the phenomenon in question. The culmination of these methods is descriptive narratives.
With this brief introduction in mind, we can delve a little further and try to understand the specific characteristics that distinguish quantitative approaches. For a comparison with qualitative methods, as I mentioned above, I refer you to the next article in this series.
Quantitative methods: the phases
Quantifying, indeed. This is the primary goal of any research that adopts quantitative methods and techniques. And by quantifying, it is possible to measure and count the social phenomenon under the researcher's magnifying glass. Let's quickly give an example. Suppose that the object of study chosen by our social researcher is the political participation of Italian citizens over a given period of time. The task and purpose of the research will then be to "measure" the political participation of the target population, namely Italians.
First—regardless of the type of research, whether quantitative or qualitative—we will proceed with a literature review to determine the state of the art of knowledge on the phenomenon, that is, what progress sociological/political science has made to date (the latter because it is relevant to the research in this example; otherwise, literature from other related disciplines can be drawn upon, in a multidisciplinary perspective). This step, therefore, is of fundamental importance as it allows us to identify (and justify) the variables we will use later. Reviewing and studying the scientific literature can be a time-consuming process: for this reason, it is crucial to choose the right reference theory, one that at least fits the paradigmatic framework the researcher has developed over the years.
Building hypotheses
Following this step, we move from theory to the construction of research hypotheses, which will give rise to concepts. These concepts must be rigorously defined, coherently, and exhaustively, so as not to give rise—as Sartori teaches—to ambiguity , on the one hand, and imprecision , on the other, in the conceptual dimension . But let's return to our example. Once we have found the appropriate literature for research on the political participation of Italians, the next step consists of identifying hypotheses consistent with the starting conjectures: for example, speculating on the fact that young individuals participate less in their country's politics.
This shift, due to the emergence of new hypotheses, can also provide a new impetus towards identifying more specific and sectoral research in the youth field (as well as empirical investigations). Indeed, the research process, at least initially, is still subject to ongoing modifications , in order to structure the very design of the ad hoc research , also allowing for a good dose of serendipity – as Merton himself taught.
Variables
We said variables. This is because quantitative research—to quantify, precisely, the magnitude of the phenomenon—is fueled by differences and heterogeneity among individuals, families, municipalities, and so on. Variables are the primary tool of the quantitative researcher, so they will be defined as "characteristics observed on any object of investigation to detect the variability of those characteristics within a defined population." Variables are a tool for reducing the complexity of the phenomenon, further simplifying a situation of great uncertainty, and breaking down the object of analysis into a series of characteristics, as we said, called "properties."
To return to the example: some variables that can be identified based on the theory are: gender, age, political orientation, occupation, income, etc. These are the most common. To try to understand if there is a relationship between these variables, we will first conduct a univariate analysis, then a bivariate analysis (trivariate, quadrivariate, etc.), and then a multivariate analysis. In short, this is to reach conclusions that can support the initial hypotheses.
The emergence of raw data
The next phase after data collection occurs with the emergence of the first results, even if they are still "raw." This is because they have not yet been substantively processed, that is, interpreted. The phase following the emergence of the results, in fact, is precisely their theoretical interpretation. That is, we must understand whether, based on the initial hypotheses, the results confirm or deny the hypotheses and to what extent they can be considered significant. Still referring to the example given above: the results may show that it is not true that young people participate little in their country's political life, or vice versa, confirm the initial hypotheses. This depends on whether the cross-referencing of the variables reveals a significant relationship or not. At this point, we simply return, through a process called "induction," to the theories, contributing to enriching the scientific landscape of existing theories, confirming some and falsifying others.
So, to summarize the process inherent in the logic of quantitative research, this process is divided into the following points:
- Literature review;
- Formulation of hypotheses;
- Data collection;
- Data Analysis;
- Data interpretation.
Quantitative methods: critical considerations
Although very brief, due to space constraints, the reasoning we've presented thus far requires some critical consideration. Although quantitative empirical research is often used to provide informed opinions, this "numbers-based" approach is not without its weaknesses. Let's examine some of them.
First of all , the reliability of a respondent's verbal behavior does not guarantee a reliable and certain measurement for statistical analysis. Ultimately, while large numbers provide a comfortable approximation of the "true" value of any measurement, they fail to account for the fundamental discrepancy between the interpretative categories of social scientists and those of common sense. Let me give just one example to illustrate. Suppose a questionnaire with structured questions asks about one's membership or closeness to a political party on a scale of 1 to 7: are we really sure that the respondent is able to retrace, in their past experiences, the actions, thoughts, and feelings that can be traced back to the concept of "membership" or "closeness" (or others like "sympathy," which is even more ambiguous) that the questionnaire asks?
Respondent's ability to introspect
And here we come to the second critical point, namely the respondent's capacity for introspection . I pose the issue with a rather provocative question: how can we measure certain social characteristics (such as the one mentioned above in relation to research on political participation) without taking into account that the capacity for introspection—like the ability to remember and select salient events—is not the same for all individuals? In other words: who can assure us that survey results correspond to the same question asked, the same stimulus induced?
But, these doubtful elements aside, there's no doubt that quantitative methods have at least some eminent practical utility. Moreover, they have become widely used in recent years in the social sciences, thanks to large databases and the so-called information technology revolution that has impacted all fields of research with software and special programs for processing data with unprecedented efficiency and effectiveness.
Bibliography:
- By Uncle S., Agustoni A., Maretti M., Research methodology, Milan, Einaudi, 2012.
- Corbetta P., Social research: methodology and techniques. See The reference paradigms, Bologna, il Mulino, 2003.
