|Faculté des sciences de la société
|Département de science politique et relations internationales
|Introduction to the methods of political science
You have to make a number of choices when you do research, and the researcher has to make five choices:
- ontological and epistemological: that is, having a certain conception of society, one almost touches on the field of philosophy and less on the field of practical research.
- have a conception of science: science is part of society, there is no one way of conceiving society and science. It is possible to assimilate this choice to the notion of paradigm.
- find an adequate mode of explanation: for a phenomenon that we want to study, we go from the most general and abstract plane to the closest plane to the objects studied. It depends on the conceptions of society and the science that one must have.
- fit into a theory: refers to the previous choice.
- choose a methodology: each of these choices depends on the prior choice you make. The way society is perceived determines the methodological choice made on a research project. A whole series of choices are made regarding techniques.
Distinction between the (post-)positivist paradigm and the interpretative paradigm[modifier | modifier le wikicode]
Paradigm[modifier | modifier le wikicode]
It's a concept from Thomas Kuhn. He tries to develop a theory on science based on the notion of paradigm, i.e. on how society develops from a paradigm. It is a theoretical perspective that is shared and recognized by the research community of a discipline that is based on previous achievements of the discipline and that guides research in terms of the choice of facts to be studied, the object, the formulation of hypotheses and the implementation of scientific research tools methodology. This is linked to the formulation of a theory, but more generally, it is the way of approaching a social theory that makes it possible to define the theoretical and methodological tools to be used to promote its theory.
The paradigm is a vision of the world, a reading grid that precedes theoretical development. Kuhn makes a fundamental distinction between normal science and scientific revolutions.
- normal science: long phase where a given paradigm in Western history has dominated.
- scientific revolution: paradigm shift.
If we abandon the idea of Kuhn's historical development and apply it to social science, today there are several paradigms that must be chosen and included.
There is a coexistence of opposing paradigms that can be characterized in different ways. Paradigms characterize social science research, methodological choices stem from the choice of paradigm in which one is inscribed.
A theory in the general sense of the term and methodology are intimately linked; we cannot think of one without the other. These paradigms are found in the theoretical pole of which there are four paradigms:
- positivist ;
- understanding ;
- functionalist ;
Ontology[modifier | modifier le wikicode]
It is a way of conceiving and developing science to study society as, for example, political phenomena.
According to Charles Tilly, there are four "ontologies", that is, ways in which researchers have approached the phenomenon to be explained, designed and developed science is to design and develop reality..:
- Phenomenological individualism: individual consciousness is the only place of social life, observation is not the best technique to follow, because we cannot see in individual consciousness. It is a question about the social reality that lies in the individual consciousness about how men construct their world.
- Methodological individualism: these are individuals as fundamental, even unique social reality, we centre ourselves in their behaviour and not in what they think. We must find the meaning of things in individuals, not in the consciousness of individuals, but in the behaviour and facts of individuals.
- Holism: the social structure, self-sustaining watchmaking, Durkheim is an example, we must analyze phenomena as a single whole (systemic approach). We cannot understand society if we do not consider all the different parts as a whole. The systemic approach in social science goes in this direction, for example. We must consider society as a whole, in other words, it is a generalizing social paradigm.
- Relational realism: social ties constitute the fundamental element of social life. There are several ways of classifying these paradigms and conceptions of society.
We will not study a phenomenon in the same way if we think that the essence of this phenomenon will be found in people's consciousness and in their relationships or we will find it in the whole of a phenomenon in which a person fits.
Sociological traditions[modifier | modifier le wikicode]
According to Collins, four sociological traditions can be distinguished:
- tradition of conflict: it is through conflict analysis that social phenomena can be explained. Society is inherently conflictual, Marx's theories are famous examples.
- utilitarian-rationalist tradition: human beings are rational.
- holistic tradition: "durkheimian".
- micro-interactionist: interactions must be analyzed at the micro-relational level.
When we talk about method, we cannot make abstractions, the reflections are not only about theory; we have to think about paradigms, ways of conceiving society, but at the same time, there are various forms in order to understand approaches.
We will distinguish between two major paradigms:
- positivist paradigm: empirical, objectivist, explanatory. Here, the term "positivist" has no negative connotation.
- interpretative paradigm: humanist, subjectivism, understanding.
These paradigms are general conceptions of the nature of social science that allow us to understand and know social reality.
There is an opposition between Durkheim and Weber (approach of facts or understanding). These are various ways in which we can know social reality. These two approaches show us the difference between quantitative and qualitative approaches.
These paradigms are developed around three questions:
- ontological question: does social reality exist and what is its nature? concerns social reality and its nature.
- epistemological question: is this social reality knowable? can we know her?
- methodological question: if this reality exists and is knowable, how can we know it?
Each of these paradigms provides different answers to these three questions.
Ontological Question[modifier | modifier le wikicode]
Does social reality exist? - Ontology.
Society exists, we can observe it, but it can only be known probabilistically, observation depends on the theory itself (post-).
This is the positivist position, we are just trying to know the position, the ontological definition that the two paradigms give.
Reality does not exist as an objective fact, but it is constructed; reality does not exist as such, it is a social construction.
Everyone has their own social reality in their head. The world we can know is the world that is built by the meaning that people give to the world.
Epistemological question[modifier | modifier le wikicode]
If reality exists, is it knowable? - Epistemology: ways of knowing.
On the one hand reality, on the other the researcher. To know the world, we must try to detach ourselves, because the social world exists and is real. In other words, it is a dualism between researchers and reality.
Being able to control all the factors, you have to get as close as possible. There is the idea of experimentation and experience. We try to replicate what we do in the hard sciences by applying them to the social sciences. It is a willingness to experiment by manipulating reality.
We want to explain the facts that are there, objectively, we do it by moving away. We're looking for cause-and-effect logic.
Law in the humanities; we must find a law. Provisional (post-) law, the falsification of the hypothesis (a good hypothesis must be capable of being subjected to empirical tests and falsified).
We deny dualism and we deny objectivity, because reality is constructed, everyone gives its meaning, it cannot be objective. There is no division between researcher and research. The world is not objective, it is by definition subjective.
We are not looking for laws, but we are looking for meaning. The goal is not to experience reality and arrive at laws. We seek to understand and interpret by bringing out the profound meaning of the phenomena observed.
We want to understand, to do it better we must participate in reality.
Statements of (probabilistic, provisional) possibility, ideal types (caricatures of reality) one cannot make laws, one can try to have some abstraction. This generalization is made by statements of possibilities and "ideal types", that is, a kind of caricature of reality where the essential traits are highlighted.
In one framework we aim to explain phenomena and in the other to understand them. The methods of data collection and analysis will obviously be different.
Methodological issue[modifier | modifier le wikicode]
How can reality be known? Methodology: what are the tools?
We want to manipulate all the explanatory factors. The researcher intervenes on reality through experience, trying to reach the objectives of explanation and generalization of a law.
Reality can be observed from the outside.
We have hypotheses, we start from a theory, and we try to test them with the observation of objective reality. Ideas are started and then tested to find confirmation or empirical verification of ideas in the field.
Since we are aiming for generalization, we prefer these methods in the technical sense.
Social reality is analyzed by variables. For example, we are interested to know if the social origin influences those who are more attentive to the course. What interests the researcher are not the individuals, but the variable.
Motivation is to better understand the deep motivation of actors to behave more in one way than another.
It is a question of interpreting the facts observed.
We try to start from reality to generate theories, in the end we want to arrive at a theory. We start from empirics to try to generate theories. We start from the subjects to go back to a theory, that is, to generate a theory.
Qualitative techniques are preferred.
The unit of analysis is individuals, the term subject becomes important. We are interested in a set of characteristics, that is, in the whole individual.
In his book, Corbetta speaks of three paradigms:
- positive paradigm: no longer exists in the social sciences, no one thinks that the social sciences should be like the exact sciences.
- post-positiviste: is more nuanced, it is the criticism of positivism notably made by Karl Popper. This critical reformulation has nuanced the pure positivist paradigm in several ways. Social reality is external, but can only be known probabilistically. On the other hand, empirical observation depends on theory.
- interpretive paradigm
From these two paradigms flow two ways of doing social science research that are radicalizations of these positions:
- quantitative research
- qualitative research
Quantitative and qualitative research[modifier | modifier le wikicode]
The Chicago School will produce systematized studies with the first efforts to study social phenomena in a quantitative way without putting qualitative studies aside. In the same university, we meet two schools. In the 1940s and 1950s, quantitative research dominated, particularly with the use of polls in elections. In the 1960s, in the United States, there was a resurgence of the qualitative approach with a return to the qualitative perspective by producing historical analyses.
General approach[modifier | modifier le wikicode]
|Relation theory - and the research process
Logically sequential phases follow (some phases must arrive before others). e.g. First you have to have a hypothesis to test it. What is important is the degree of structuring. It is the shaping of research. We have a highly structured research design with sequential phases.
The theory precedes observation according to a linear logic.
There is no idea that we first need to have a hypothesis, we can start with the field. There is a lower degree of structure. It is a more open and interactive approach, not separate moments. The whole process is much more flexible.
The theory emerges from observation, but not necessarily all the time, because this approach is more flexible. The two phases can follow one after the other in a more cyclical logic. It's a back and forth between theory and empirie.
|Function of literature
It is fundamental for defining theory and formulating hypotheses. This supports the logic of linearity and sequential phases.
It is auxiliary, in the research process it is less important compared to other aspects. At the beginning it is possible to settle for a less important knowledge of the field compared to quantitative research, because there is no need to formulate hypotheses.
They are operationalized, make it possible to move from the theoretical level to the empirical level; they make it possible to translate theoretical concepts empirically.
Orientative concepts, they are open, in construction, the relationship between research theory is interactive, so concepts too. They must give direction to the analysis and may change during the research. They are under construction allowing a better definition of the concept.
|Relationship to the environment
The researcher intervenes in reality and tries to change something, especially in experimental studies. The researcher is external, he intervenes through methodological tools.
There is no idea of manipulation, the researcher is part of reality. We want to manipulate the variables in their natural state without intervening. This poses the methodological problem of how to study an object without modifying the studied environment, especially since people react to the way they are studied.
|Psychological interaction between researcher and object of study
Detached and neutral. Axiological aspect, we try to neutralize our values to see this distance of interaction, I am not part of "this" reality.
We try to put ourselves in the jackets of the people whose behaviour we want to explain. Empathic identification is sought from the perspective of the object under study. This helps to understand the actions of others through empathic identification.
|Physical interaction between researcher and object of study
The researcher doing the analysis should not be the one collecting the data.
It is much rarer for a researcher to try to analyze data produced by someone else.
|Role of the subject under study
Action, helping groups to engage in a process of transformation.
Data collection[modifier | modifier le wikicode]
Precedes the search itself. First, drawing, then research. Before you can go into the field, the research drawing must be final.
You do not need to have a pre-established research design, it can change depending on the interactions you have with the subject being studied. As the concepts are open, the way in which we will study a certain reality is constructed during the research. Idea of non-linearity of research.
Especially in the individual approach, a survey is conducted so that it is as representative as possible. The idea is to start from a sample and draw broader conclusions. The research results are supposed to be generalized to the universe from which individuals were drawn.
Representativeness is not essential, it is the singularities that interest us. We seek to study singular cases that are not statistically representative. We do not want representativeness, because we are trying to identify specific cases.
The goal is to have an objective representativity by a controlled instrument which makes it possible to arrive at a matrix of data. It is through standardization and standardization that statistical representativeness can be achieved.
We do not tend to standardize, we vary according to the subjects we have before us, according to the research objectives. We will not apply the same instrument of collections to all subjects, we vary it according to the specific interest and what we want to bring out of the analysis.
|Nature of the data
The main opposition lies in objectivity as opposed to subjectivity. The data must be as objective as possible.
The data have depth as opposed to superficiality; the data must be as deep as possible.
Data processing[modifier | modifier le wikicode]
|Object of the analysis
Analysis by variables and impersonal. The main unit of analysis is the individual, but above all the characteristics of these different individuals.
It is at the centre of the analysis, it is a holistic approach. We want to capture and seize individuals as a whole.
|Objective of the analysis
We look for correlations between variables, they are covariations.
Data processing is a paradigm of understanding.
|Mathematical and statistical techniques
They want to collect data in a standardized way (a large number of cases need to be worked on)
Results[modifier | modifier le wikicode]
|Presentation of results
The aim of the operation is to find relationships between the variables, the information is presented numerically. (However the narration is also present). We're looking for correlations or covariations.
Relating certain attributes, they try to construct a narrative. It is a holistic approach to understanding the person. It is based on excerpts from interviews; it is a narrative perspective such as, for example, reconstructing a person's voting habits.
We're trying to establish causal models, laws. e.g. those with a high academic level are more likely to participate in politics. Their priority objective is to generalize.
Causal logicWe will talk about causes, we will look for cause-effect relationships.
It is a schematic representation of society that allows us to go beyond the specific framework studied. They do not necessarily want to generalize their results.
The idea is to keep all their characteristics in a classification logic.
|Scope of the results
We want to infer the result to a broader set than we can characterize as nomothetic, that is, we are looking for laws.
We tend towards specificity, that is to say we tend towards an idiographic description.
Some examples: quantitative and qualitative research[modifier | modifier le wikicode]
Often when we talk about methods, we almost always discuss the different approaches and almost systematically we find ourselves in opposition between professionals. Although these two approaches are treated here as distinct, it is not necessary to leave with a fixed idea of the two approaches, because there are also certain characteristics that belong to one or the other or to both.
Research has been done on individual engagement in social movements, trying to explain why individuals participate in social movements.
Two questions emerged; perhaps through these two approaches there would be different answers:
- why do people go into social movements? what are the explanatory factors? idea of the quantitative approach
- what are the processes and mechanisms that engage people in movement? how do we get involved? idea of the qualitative approach
1st step - first we will look at explanatory factors. Readings that highlighted three main types of factors:
- socio-demographic characteristics: depends on values.
- integration into prior social networks: individuals become part of movements not necessarily because they share values, but because they are inserted into prior social networks; these are links that the person may have with other people regardless of their values and socio-demographic characteristics.
- perception of external reality: people who feel they are effective and have an impact if they mobilize and mobilize.
This has allowed the construction of a theoretical framework that allows hypotheses to be formulated.
2nd stage - definition of the research drawing and hypothesis tests :
- sample survey: sending surveys and questionnaires according to strict criteria. It is a logically sequential process.
- statistical analysis: data collection and construction of a data matrix for statistical analysis. Based on the results, we tried to generalize them by looking for correlations between the level of integration and participation, etc.
- generalisations: the objective was to go from the individual to the general by applying research to all persons involved in social movements.
- explanation: the idea is to explain the degree of commitment and to highlight a cause-and-effect logic and to draw correlations through relationships between two variables. In this case, regression coefficients were highlighted. Thus we have a measure of the relationship between the variables.
- conclusion: the relationship exists not only in the sample, but also in the general population.
Rather, the focus is on the activists' pathways, processes and mechanisms that lead to engagement:
- in-depth interviews with activists: there is proximity through the study of personal cases and this through the analysis of life stories. An attempt was made to reconstruct the lives of some of the committed members in order to determine what led them to become involved. The idea was that commitment was linked to people making connections between their different spheres of life. It is a holistic approach considering the fact that we are interested in people's subjectivity and having an empathic approach in order to understand the reasons for their commitments as well as, for some, their disengagements.
- through interviews a theory was generated: the hypothesis is that the commitment comes from a dysfunction between the spheres of the individual. An Activism Theory has been generated - it is more likely that people will remain engaged in a movement over a long period of time if the various spheres of life are linked together, there needs to be some coherence.
- excerpts from interviews were used to present the results: a story was written to show that when activists had a holistic vision, then they were more likely to remain engaged in the movement. While people who disconnect spheres have a greater propensity to disengage. Thus the extracts make it possible to corroborate a discourse.
- conclusion: we have some ideas that guide us and show how this approach aims to generate a theory.
What is the best method?[modifier | modifier le wikicode]
Methodologists have three different positions on whether quantitative rather than qualitative research is better or vice versa:
- A rather rigid one that says that quantitative and qualitative methods are incompatible: in view of fundamentally different logics that rely on completely different ontologies and epistemologies, then they cannot be recorded because they rely on very different theoretical elements (does reality exist as an external object?)
- Subsidiary to qualitative methods: they say that some aspects of qualitative analysis are useful, but quantitative continues to be the "best" method; the quantitative method is considered the method par excellence because it is more scientific and the qualitative method has a subsidiary function. Qualitative methods have a subsidiary function, i.e. there are certain aspects that can be addressed in some interviews. Quantitativist researchers in particular advocate this approach.
- Full legitimacy of both methods: it depends on the research, on the method we are going to study. Neither is less scientific than the other (qualitative is no less than quantitative)
Combining methods is not an easy task due to the differences reviewed between these two paradigms. However, it depends a lot on the research question, on what is being studied; there are some questions that are more likely to be appropriate to one of the two approaches. The two main methods lead to knowledge that can be different. Finally, these two methods can be combined in an approach that some call "triangulation of methods" which makes it possible to bring several approaches.
Annexes[modifier | modifier le wikicode]
- Introductory Essay to Charles Tilly’s Writings on Methodology By Y. Sekou Bermiss
- "How I Work" by Charles Tilly
- Charles Tilly : un riche héritage pour la sociologie politique par Hubert Rioux - Politique et Sociétés, vol. 30, n° 3, 2011, p. 99-122.
- Historical Concepts, Social Ontology, Macrohistory by Daniel Little - University of Michigan-Dearborn; SSHA, Pittsburgh, October 2000