Writing your dissertation may be one of the hardest part of your studies. However, this guide will help you cover everything you need to know about writing your dissertation in 2021.
Let’s dive right in!
DRAFTING, STRUCTURING AND PLANNING A DISSERTATION OR THESIS
Definition, topic and specificity of a dissertation thesis
The dissertation thesis is an academic work that proposes a theoretical and/or applied approach based on recent research in the field in which the author makes a contribution to the field by extending the research on a particular topic, criticizing existing theoretical models and proposing new models and approaches in the chosen field of research.
The distinctive feature of a dissertation thesis is the scientific, argumentative investigation of a precisely specified research topic; it is not a journalistic or literary approach, but an approach that uses methodologies specific to a research approach – arguing the choice of research topic, synthesis of theoretical positions on the problem studied, critical analysis of existing research in the field, proposal of new research topics or hypotheses, experimental, empirical validation of the hypotheses of the research questions, formulation of scientific conclusions, etc. In this respect, reading peer-reviewed articles in research journals specific to the research field has a double benefit: 1) it is an important source of information and 2) it familiarizes the reader with the approach and rigors of the research field.
The following aspects are important when choosing your theme:
- what previous interests you have had in the subject – theoretical or practical;
- how familiar you are, conceptually and factually, with the field in which you wish to do your dissertation;
- research skills – what methodology you have mastered and how you can work with it;
- your coordinator’s interest and competence in the field;
- the relevance of the chosen topic.
Title of the dissertation thesis
The title of an undergraduate thesis or dissertation should inform the audience about the central theme of the research. It is generally established at the beginning of the thesis writing process with the coordinating professor and may be changed during the course of the writing process if the thesis develops in different directions or elaborates on aspects discovered during the writing process.
The title will refer to:
- the specific research topic and possibly the research hypothesis;
- specific theoretical or methodological approach;
- the results or impact of the research.
Possible errors in the wording of the title:
- very general;
- very long;
Size of the dissertation thesis
In general, an undergraduate dissertation is between 50 and 70 pages, and a graduate dissertation is between 40 and 60 pages. It is important that the minimum number of 50 or 40 pages is respected, but an undergraduate thesis may exceed 70 pages and a dissertation may exceed 60 pages (if the topic is suitable or if you feel you cannot cover the topic in the recommended number of pages).
The size of the thesis depends on:
- the intricacies of the field to which the thesis belongs;
- the theoretical apparatus required for the argument;
- the type of research methodology used;
- the volume of collected and analyzed data.
Structure of the bachelor or dissertation thesis
The canonical structure of a dissertation consists of three main parts: introduction, body and conclusions. It is desirable that the introduction and the conclusions of the thesis contain between 5 and 10% of the word count of the whole thesis. Thus, between 80% and 90% of the text of the thesis is allocated to the chapters that make up the body of the thesis. For example, of the total 70 pages of the dissertation, it is recommended that between 4 and 7 pages be devoted to the introduction, 4 and 7 pages to the conclusions, and the remaining 62 and 56 pages respectively be allocated to the body chapters.
This canonical structure should be detailed by constructing an outline of the paper, including chapter and sub-chapter headings. The outline of the thesis a) facilitates the delimitation of the scope of the research (so that too broad topics that cannot be covered by the thesis can be avoided), and b) represents the work plan whose research questions guide the scientific approach, helping to outline the red thread of the thesis.
In the following we characterize the canonical structure of an undergraduate or graduate dissertation paper.
A) The following elements must be specified in the Introduction:
1) the importance of the research topic and the motivation of the student in choosing the topic,
2) the research questions on which the research is based,
3) the methodology used,
4) the chapter structure of the thesis/dissertation and a brief description of the chapters,
5) the bibliography and documentation on which the research will be built.
B) The body of the dissertation must be the elaboration of the arguments subsumed by the research questions. The proposed arguments must be methodologically, and, where appropriate, empirically, grounded. A distinctive feature of a research endeavor such as that represented by the thesis is the scientifically rigorous elaboration of the proposed arguments.
Arguments can support two types of results:
a) negative results – the function of which is to expose and criticize the shortcomings of an approach or theory
b) positive results – the function of which is to propose original approaches, to theorize certain aspects that have been elided, insufficiently or inadequately analyzed.
C) In the Conclusions, the student should indicate the results of the research, answers to the proposed research questions, limitations encountered during this research, if any (e.g. the novelty of the topic and the lack of books on this topic) and how this research opens new directions of study that could be explored in the future. The student should also indicate the significance of the results obtained from a theoretical and/or practical perspective in the context of research in the field. Where appropriate, students are encouraged to evaluate and present the impact of their research at both theoretical and practical levels.
The expression must demonstrate mastery of the language of the field. Appropriate formulations are:
- the proposed approach analyses, treats, describes, explains, etc.;
- Chapter 1 aims to present, analyse, observe…etc.;
- data has been measured, extracted, verified.
The role of the coordinator
In order to better capture the role and limitations of the coordinator’s involvement in the development of the dissertation, we will break down and summarize the main tasks and limitations of the coordinator’s involvement:
The main tasks of the coordinator are to:
- indicate the approach and structure of the work;
- analyze and validate the theoretical approach and argumentation, as well as the methodological approach;
- support in the selection of relevant bibliography;
- support in data analysis and interpretation;
- give feedback on progress;
- formulate critical opinions;
- evaluate the work as it progresses and make suggestions on the progress needed for its completion and public presentation.
The coordinator is not responsible to:
- search for and provide students with the complete bibliography of undergraduate or dissertation work (but only to guide the student towards relevant work in the field). Students must identify, inventory and read scholarly works, scientific articles appropriate to their own discipline, which can be found both in libraries and in international databases (e.g. J-Stor, Ebsco).
- to provide administrative information;
- to synthesize, instead of the students, the theoretical part of the thesis or dissertation, but only to direct the students in this respect;
- to design, instead of the students, the analytical part of the thesis or dissertation, but only to direct the students in this respect. Essentially, the students have to bring something new to the field of study in which their scientific work falls, and this is only supervised by the coordinator.
Drafting your dissertation
The first step in drafting the thesis or dissertation is to familiarize yourself with the research area and terminology. The prerequisite for research is the framing of the topic in the research field (international relations, computer science, political science, economics, law, management, chemistry, business administration, etc.) and a good working knowledge of terminology. Depending on these references, contact will also be made with the coordinator in order to establish:
a) the title,
b) the general structure of the thesis,
c) the research plan/design.
After this stage, the student and the coordinator will identify the main bibliographical sources corresponding to each component of the general structure of the thesis. A timetable for the completion of this first bibliographic list will then be agreed upon.
Once the first bibliographical list has been completed, the theoretical approach and the practical part (if applicable) will be discussed, the relevant bibliography of both segments will be completed and the detailed structure of the thesis will be formulated. In the next stage, the student will prepare and submit to the coordinator the synthesis of the theoretical part of the thesis; this stage is followed by the analysis and reformulation, based on the coordinator’s suggestions, of the theoretical part of the thesis.
Once these stages are completed, the focus will be on the development of the case study and the implementation of the applied/experimental part of the thesis.
At this stage, the relevant data for the case study or those obtained from the implemented application are collected and analyzed, followed by the interpretation of the data in the theoretical context presented in the first part of the paper, and then the formulation of conclusions.
The collection, analysis, interpretation and formulation of conclusions is carried out in full methodological rigors, and the final stage is carried out by establishing with the coordinator the methodological aspects and the conclusions formulated.
Planning the writing of the thesis
In order to meet time and quality standards, it is worth considering the following milestones:
- The choice of topic and coordinator should be made during the second semester of the last but one year of study, or at the latest in the first semester of the last year.
- Formulation of the title and the first structure of the dissertation to be completed within a maximum of one month after the choice of the coordinator.
- Completion of the bibliography for the theoretical part: 1-2 months.
- Consultation with the coordinator and analysis of the theoretical part: 2-3 weeks – ideally in parallel with the bibliography.
- Approaching the case study or application and collecting, analyzing, interpreting and drawing conclusions: 2-3 months.
- Consultation with the coordinator and analysis of the case study or application – 2-3 weeks.
- Verification of the final version – 2-3 weeks.
- the practical part also depends on factors outside your control – so make sure you plan ahead;
- the final drafting of the paper will take at least two weeks;
- set a timetable for consultations with the coordinating professor in advance – at least one meeting per month is necessary.
Stages in writing the bachelor or dissertation thesis
In order to make your work more efficient, we suggest the following algorithm for writing your dissertation thesis:
Selection and review of the bibliography
- use various ways of searching and selecting the bibliography – libraries, online databases, accredited sites, etc.;
- study different methodological approaches specific to your topic;
- check that you have sufficient bibliographical sources available – discuss your final selection (based on author/concept list, theoretical approach) with the coordinating teacher.
Browse the bibliography
- build a correspondence between the theoretical approaches summarised and/or employed in your paper and the researchers who proposed and used them;
- construct a correspondence between the theoretical concepts used and the researchers who theorized, analyzed, critiqued them;
- draw up a concept map of the work from the bibliography.
Analysis and writing of the theoretical part
- present in an undistorted way all the theoretical approaches employed;
- present proportionally all the theoretical approaches employed;
- in the presentation of the theoretical approaches, do not omit to present criticisms and point out their shortcomings;
- try to exemplify in an original way the use of the theoretical approaches presented.
Drafting and detailed formulation of the methodological approach
- select with the coordinator the bibliography for the methodology part and start writing this part;
- present a comprehensive framework of the different ways of approaching the case study or the application of your research topic;
- analyze the different methodological approaches by presenting the strengths and weaknesses of each;
- argue your methodological choice;
- discuss with the coordinator the methodological approach you opt for and revise the methodological part if necessary.
Conducting the research
- on the basis of the adopted research plan, collect, analyze and interpret the data;
- specify the limits of the methodological approach or problems in carrying it out.
Data collection and processing
- select the most feasible data collection methods from the options set out in the methodological framework;
- collect data in full and strict compliance with methodological protocols;
- check at least twice the accuracy and thoroughness of the data collected with the coordinating teacher and start processing the data.
Writing the practical part and data analysis
- by using the assumed methodological tools, use the data collected to refute or confirm certain research hypotheses;
- use textual formulations as well as images, graphs, tables to present relevant research results and to support your research approach and the interpretation you present.
Interpretation of data
- is one of the most important parts of your approach and the one that is of most interest in evaluating your work;
- validate your interpretation of the data with the coordinating teacher;
- avoid interpretations that are too general or too vague;
- formulate your interpretations in relation to the research themes/hypotheses and research objectives.
- The conclusions should provide an overview of a) the significance of your work in the context of related research, b) the limitations and added value of the research, and c) the directions for further research that can be explored based on your work;
- conclusions should explicitly state what the answers to your research questions are.
THE RESEARCH PLAN (DESIGN)
What is the research design?
The research design is the investigative strategy adopted to provide a scientifically qualified answer to research questions or hypotheses. As can be seen from this definition, the main factor guiding the research design is the research questions or hypotheses. The nature of the research questions or hypotheses proposed to explain a particular phenomenon determines, to a large extent, the adoption of a particular research strategy. For example, if we want to investigate the county-wide or national impact of a retraining program, then it is advisable to adopt a quantitative approach; if, on the other hand, we want to investigate the marital practices of a marginalized community, then we will opt for a qualitative approach. The difference between the two approaches can be seen in terms of the information collected: in a quantitative approach, the information is precise, replicable and easily usable for generalizations, but characterizes certain superficial aspects of the phenomenon studied, whereas in a qualitative approach the information targets deeper layers of the phenomenon studied, but is imprecise, difficult to replicate and ineligible for generalizations.
Developing research questions and hypotheses
In general, research questions and hypotheses are the consequences of problems or difficulties in evaluating and analyzing certain phenomena within certain theories. This characteristic is not specific to the socio-human sciences, but is the main driving force behind the evolution of hard sciences. For example, the main pillars of contemporary physics, relativity theory and quantum mechanics, were developed following the perception of explanatory and predictive shortcomings of Galilean-Newtonian mechanics with regard to certain phenomena, such as the perihelium of Mercury or the photoelectric effect.
In the social sciences, too, the detection of inconsistencies between the explanations and predictions that a theory implies and the phenomenon under investigation is an important factor in generating research questions and hypotheses. For example, Kenneth Waltz’s theory of structural realism led him to predict in 1993 that Germany and Japan would soon become nuclear states and that NATO would collapse as a result of the collapse of the Soviet Union. Both predictions turned out to be false, so investigating the causes of these inconsistencies and developing alternative explanatory and predictive models became research topics.
Another important factor in generating research questions is the in-depth investigation of a phenomenon, the causes that produce it and the accepted explanation for its occurrence. Analytical scrutiny of a phenomenon often produces a significant change in the way the phenomenon is perceived, evaluated and interpreted. For example, without being the corollary of a particular political theory, it is widely assumed that there is a growing and pervasive disinterest in civic and political activism in many Western countries. Some studies of this phenomenon, however, such as Pippa Norris’ book Democratic Phoenix: Reinventing Political Activism, challenge this assumption, based on a laborious survey of indicators of political engagement in 193 countries.
- The sources of the research questions and hypotheses discussed above do not constitute an exhaustive list of factors that generate the research questions.
- As we emphasized, the development of research questions and hypotheses takes place after a thorough understanding of the theoretical framework within which the phenomenon is being assessed.
In addition to the research questions and hypotheses, the adoption of a research plan is influenced by the resources available to undertake the research, whether these resources are of an economic nature, such as available funds, or of a more abstract nature, such as evaluation and analysis time. Suppose the Ministry of Education wants to find out the impact of a curriculum focused on developing an awareness of European identity on pupils’ tolerance and asks for an evaluation of this impact in three months. Although such an investigation is usually managed quantitatively, the short evaluation time will lead to the adoption of qualitative methods, e.g. focus groups in different parts of the country and appropriate analysis of the results. As mentioned above, the financial and human resources available to carry out the research also influence the adoption of a research design.
Another factor determining the choice of research plan is the researcher’s training and skills. For example, a political scientist with statistical skills will be inclined to focus on a type of problem that calls for an experimental or quasi-experimental research design, an economist will be inclined to develop mathematical models of the phenomenon being studied, a game theorist will analyze conflicts using game theory tools, while an anthropologist will prefer participatory observation and a historian archival analysis and/or document analysis.
Certain disciplines, such as political science, international relations, etc., lend themselves to different methodological approaches, using both qualitative and quantitative tools, so it is not surprising to find different methodological approaches in these fields, even within the same study, which are capable of capturing the phenomenon under investigation more accurately.
Functions of the research plan
The main functions of the research plan are:
- to operationalize the research questions into a rigorously articulated strategy of data collection and analysis and
- to ensure that these procedures are adequate to provide scientifically valid answers to the research questions investigated.
Of the two functions, the greater weight is given to the second function, although the process of operationalizing the research questions is a key element in the development of the research design.
Operationalization is the process by which research questions or hypotheses are given a practical dimension, i.e. they are formulated in such a way that they can be implemented, tested and verified, or experimented with. In this sense, the task of elaborating and defining the concepts that constitute the theoretical tool of analysis becomes crucial.
A first step in the operationalization process is therefore to specify the theoretical framework within which the analysis is carried out. Without insisting on the importance of the theoretical framework, it should be noted that the analysis of any phenomenon, be it social or physical, can only be carried out in a theoretical context. In the absence of a theoretical framework, we cannot talk, strictly speaking, of phenomena, much less of research questions or hypotheses. This constant awareness of the theoretical lenses through which we analyze a phenomenon requires, on the one hand, the identification of the limits of research and, on the other, the problematization of the significance of the results obtained and the distortions inherent in the use of theoretical tools.
The clarification of the theoretical framework implies:
- defining the significant concepts in the research questions or hypotheses, concepts known in the literature as variables and
- identifying the units of analysis.
- In the following we will outline these two constitutive elements of the operationalization process, noting, beforehand, that variables and units of analysis determine each other.
- In general, variables are defined as those characteristics or features that differ from one unit to another. In order to operationalize the research questions and hypotheses it is beneficial to distinguish between two types of variables: independent variables and dependent variables. Let us detail what independent and dependent variables are.
Independent variables are those autonomous factors that determine or influence certain characteristics or phenomena – represented by the dependent variables. For example, if the characteristic we want to investigate is wage income, then one factor influencing wage income is education level; we will therefore consider education level to be an independent variable (obviously, in relation to the wage income variable).
Dependent variables are those characteristics whose manifestation is determined or influenced by certain factors – represented by the independent variables. In the previous example, as can be seen, the dependent variable is wage income.
a) Dependent and independent variables are correlative concepts (each is defined by reference to the other)
b) Dependence and independence of variables are relative; an independent variable relative to a particular characteristic or phenomenon under investigation may become a dependent variable in another study. For example, the independent variable level of education (relative to the variable wage income) may become a dependent variable in another study.
Analysis units are the entities possessing the characteristics under analysis. In this sense, we speak of the units of analysis as the supporting objects of the characteristic or attribute under examination. The circumscription and explicit identification of the units of analysis play a decisive role in clarifying the scientific approach in general and the research objectives in particular. Without claiming to be exhaustive, we present below some types of units of analysis frequently encountered in the social sciences.
Units of analysis can be:
a) human individuals,
b) groups or collectivities (such as family, household, etc.),
c) institutions and organizations (such as universities, town halls, banks, firms, political parties, etc.),
d) social interactions (marriages, divorces, arrests, demonstrations, email exchanges, quarrels, etc.),
e) social artefacts (articles, books, newspapers, TV programms, jokes, etc.).
Each of these units is the seat of a characteristic or variable of interest. For example, individuals are the unit of analysis for characteristics such as age, height, weight, gender, profession, etc., a collectivity such as a family is the unit of analysis for characteristics such as number of spouses, number of children, average income, etc., institutions and organizations are the unit of analysis for variables such as funding regime, field of authority, profit, ideology, etc., marriage is the unit of analysis for a characteristic such as type of ceremony, newspapers or TV programs for the presentation and analysis of a political theme or event, etc.
The unit of observation and the unit of analysis do not coincide. The observation unit is the entity from which the information is collected. The unit of analysis is the entity to which we can attribute the characteristic under study. For example, if the variable of interest is the average income of a family, the unit of observation is the individual (the data is collected from family members), but the unit of analysis is the family (the family has an average income, not its members).
A second step in the operationalization process is to identify indicators that allow the implementation, verification and testing of the hypotheses and questions of research questions.
The second function of the research plan or design is to validate and verify that the conclusions and answers offered in the study are articulated in full scientific rigors. We will not insist on the importance of this step, but we will mention the main risk involved in circumventing it: that of missing the objective or task of the research by producing inadequate or unsubstantiated answers, in short, pseudo-answers. In order to prevent this, we need to focus, firstly, on establishing the appropriateness of the methods and procedures used in identifying a relevant answer and, secondly, on checking the appropriateness of the relationship between the concepts/variables involved in the research questions and the indicators identified as empirical (observable and measurable) expressions of them.
Establishing the appropriateness of methods and procedures will prevent us from investigating phenomena with inappropriate instruments, for example, applying a questionnaire (typical of certain societies and cultures – such as the Western one) to an indigenous tribe in the Amazon basin, or investigating with statistical means the relationships within such a tribe. This in extreme cases. In most cases, the research design will allow us to choose the appropriate methods to achieve the research objective. For example, if we aim to investigate a population-wide phenomenon captured in the study by a quantitative variable, we will not use non-random sampling techniques such as snowballing or focus groups in the research design. At best, however, the appropriateness of methods and procedures will prevent us from hunting mosquitoes with a cannon (using methods that are too strong for the research objective) or digging up the garden with a scalpel (using instruments that are too fine for the research purpose).
Checking the appropriateness of the relationship between concepts/variables and indicators involves ensuring that we have not operationalized another concept, that we have not operationalized aspects superficiality of a concept and, as a result, you have not captured the analytical significance of the concept or that you have operationalized the concept too crudely. To mitigate these risks, it is beneficial to adopt some validation and verification procedures. For example, if the study involves an experimental or quasi-experimental design (which we will discuss a little below), then validation should be done recursively experimentally. If the study is of a different nature, then the validation and verification of indicators must be done both top-down, starting from the proxies of the operationalized concept to the indicators, and bottom-up, starting from the indicators to the concepts and variables.
1) The functions of the research design/plan are:
- to operationalize the research questions into a rigorous data collection and analysis strategy and
- to ensure that data collection and analysis procedures are adequate to provide scientifically valid answers to the research questions under investigation.
2) Operationalizing research questions involves:
- specifying the theoretical/conceptual framework which is achieved by:
- defining the independent and dependent variables
- identifying the units of analysis
- identifying indicators that allow the implementation, verification and testing of hypotheses and research questions.
3) Validation and verification of the procedures and methods used in the research involves:
- determining the appropriateness of the instruments of analysis for the purpose of the research
- verifying and matching the relationship between the concepts/variables involved in the research questions and the corresponding indicators.
Types of research plans (designs):
Research plans can be classified according to different criteria, and even if the same criterion is adopted, there is no single classification; at best, there are widely accepted classifications. For example, depending on the methods of data collection and analysis we distinguish between quantitative and qualitative research designs. In the following, we present the most commonly used research designs.
1) Descriptive research plan/design
2) Correlational research plan/design
3) Experimental research plan/design
4) Comparative research plan/design
5) Case study plan/design
1) Descriptive research plan/design
In a descriptive research design the interest of the investigation is to capture a particular phenomenon, more precisely, to determine the characteristics of the phenomenon. In this sense, descriptive research design is said to answer questions such as
“what are (the characteristics of the phenomenon)?”, “what is (the feature of the phenomenon)?” etc. In the descriptive design, the variables are not controlled by the researcher, which means that the study only provides a picture of a phenomenon, but does not help to identify and verify its causes. The traditional ways of implementing quantitative descriptive design are: i) through observation and ii) through surveys and/or polls. In qualitative research, the implementation of descriptive design is often carried out through case studies.
2) Correlational research plan/design
A correlational research design involves investigating relationships between two or more variables. Even in this design the researcher does not control the variables, so the study does not explain the causal mechanism between the variables, but only whether or not there is a correlation between them. For example, we suspect that there is a relationship between gross domestic product and energy consumption (both per capita) within EU countries, so to test our hypothesis we adopt a correlational research design. The limitation of this design, however, will prevent us from establishing what causal relationship exists between the two variables. All we can prove, within this design, is the existence or non-existence of such a correlation.
3) Experimental research plan/design
In an experimental research design we test the causal relationship between variables. Therefore, in an experimental design, the researcher controls the variables and establishes the existence/non-existence of causal relationships between variables. For example, we will adopt an experimental research design if we want to determine the impact of an educational program, or the effects of a drug. Obviously, in such a design, the researcher manipulates the variables and conditions of the experiment. Experimental design is the most rigorous way to test hypotheses and determine causal relationships.
4) Comparative research plan/design
Comparative research design or plan involves investigating the relationship and/or causal relationships between variables by analyzing similar cases. There are two broad types of comparative research designs: similarity-based and difference-based. In a similarity-based comparative design, we consider cases that are as different as possible in terms of a particular characteristic suspected as the cause (the independent variable) and as similar as possible in terms of other characteristics (other intervening variables – background variables). In a difference-based comparative design, we consider cases that are as different as possible in terms of background variables and as similar as possible in terms of independent variables. Obviously, in both designs we compare the effects of variations of some variables (background or independent variables) on the dependent variable. The logics of the two designs are, we believe, easy to grasp. In the first case it is an elimination logic: if the background variables are kept (as much as possible) constant, but we vary the independent variable as much as possible, if the phenomenon (surprised by the dependent variable) occurs in some cases and not in others, then we can exclude the background variables as possible causes of it and consider the independent variable as the cause. In the second case, we have a logic of concordance: if the background variables are (as) different as possible, and the independent variables as similar as possible, if the effect or phenomenon is present in all the cases analysed, then the independent variable is most probably the cause.
5) Case study plan/design
The case study design involves the in-depth investigation of an event, a state, a community, institutions, institutional policies, etc., generically, a situation or case over a longer period of time.
In general, the case study design has two major functions: a) a prospective and descriptive function, whereby the case study becomes the foundation for hypothesis formation or investigation of a more general phenomenon etc. and b) a control or examination function whereby we test the proposed hypotheses or theories employed in explaining a more general phenomenon instantiated by a case, often deviant. The case study used for the purpose of disproving a hypothesis or theory is known as a critical case study.
DATA COLLECTION AND ANALYSIS METHODS
These methods are traditionally divided into quantitative and qualitative research methods and techniques. Each of these methods contains data collection or gathering techniques and data analysis techniques. Before analytically presenting the main quantitative and qualitative methods of data collection and analysis, we will list some significant heuristic approaches involved in the investigations underlying undergraduate and graduate dissertation work.
Approaches to the research topic
The exploratory approach is used when dealing with completely new questions; either in research areas where little data and information exists, or in a completely new way of approaching the subject (when an author tries to propose a new paradigm for dealing with a subject or when the phenomenon under investigation has not been approached in a scientific way before).
The descriptive-empirical approach is used when the methodological approach to data collection is a priority. It involves the systematic description of a phenomenon, social mechanism or process, based on data collected using various qualitative or quantitative methods. The purpose of this approach is to formulate observations and conclusions about phenomena that are not directly observable or to empirically test theories or theoretical models (see descriptive research design).
The comparative approach is used when we aim to identify similarities and differences in societies, communities, groups, institutions or organizations, or in particular phenomena or processes, offering the possibility of classifications based on a number of variables (see comparative research design).
The historical-interpretive approach
The historical approach is used when looking at processes. Historical analysis should focus on the context in which the facts under analysis took place, the meaning created by that context and their transformation over time. Specific working methods can be used such as oral histories collected through interviews, historiography and document analysis.
Quantitative data collection methods
The main quantitative data collection methods and techniques are observation, survey and experiment.
In observation, the researcher passively records some data about a phenomenon without actively intervening or investigating the observed phenomenon. This data collection technique is rudimentary, but not irrelevant or useless. However, its major limitations stem from the passivity of recording observations: no information is solicited, no factors are manipulated, therefore the information collected is precarious.
The survey is by far the most important and frequently used data collection technique. We will not dwell on the different survey typologies but will try to clarify what a survey is and what information is collected. A survey is a method of collecting data from individuals selected according to rigorous sampling procedures by questionnaire. The sampling procedure ensures that the individuals surveyed form a representative sample of the population from which they come.
The questionnaire, therefore, is the investigative instrument of the survey. The questionnaire takes the form of a series of questions generally covering one or more of the following characteristics:
a) socio-demographic characteristics (sex, age, marital status, etc.),
b) socio-professional characteristics (profession, length of service, etc.),
c) social characteristics (background, social status, etc.),
d) economic characteristics (salary, income, expenditure, etc.) and
e) attitudes, perceptions of phenomena, events, institutions, etc.
At this point in the discussion it is useful to introduce the distinction between a survey and a survey: broadly speaking, the difference lies in the predominantly objective nature of the survey, focusing on the collection of socio-demographic, socio-professional, social or economic characteristics, whereas the survey is geared more towards the identification of individuals’ attitudes, perceptions and intentions, and therefore the survey focuses on more subjective components. Of course, this difference is not the only distinction between a survey and a survey and does not mean that the other component is completely absent: in a survey, attitudes and perceptions can be quantified, just as in a survey socio-demographic and socio-professional characteristics, etc. are present.
The experiment is the most scientifically rigorous method of collecting information. In an experiment, the researcher not only records the characteristics of interest of a phenomenon, but also manipulates its determinants. Controlling the factors translated into controlling the independent variables defines the experiment. Details of the nature and function of the experiment in the social sciences can be found in the experimental design section of the Research Design chapter.
Qualitative data collection methods and techniques
This method of data collection is indispensable when dealing with sensitive, high-stakes social issues (so time spent among subjects and gaining trust); it is useful when correcting/completing data collected by other methods (e.g., to reveal the contradiction between candid responses and the equally candid facts that contradict them, or to avoid behavioral change generated by more invasive methods of observation); respectively, it is problematic when limited time resources are involved or when issues need to be put into a broader context (than that observable on the ground) to be more adequately understood.
Field notes can be seen as the inevitable and indispensable bureaucratic element of any qualitative research, and they make the difference between the amateur and the professional researcher. When collecting and coding fieldnotes it becomes obvious to everyone that the analysis and data collection is done simultaneously by establishing the criteria of relevance (what facts we have followed and what facts we have ignored, what questions we have asked and what questions we have omitted, etc.) and by organizing the information with the help of codes used in the indexing of paragraphs.
The looser the structure of the interview, the greater the likelihood of gathering data of a qualitative nature with a high degree of novelty, i.e. the greater the degree of control, the greater the quantitative nature by obtaining confirmations or verifying frequencies. Unstructured interviewing is also known as ethnographic interviewing because it is the most common form of interviewing used to collect qualitative data.
All the issues mentioned with regard to the unstructured interview also apply to the focus group. Both produce data with a strong ethnographic character, both have an acceptable degree of validity, and both have the same limitation: when it comes to representativeness or if measuring frequencies is an important issue, they need to be complemented by quantitative methods. An advantage over personal interviewing would be to be able to study group dynamics, conflicts, ways of negotiation and the cultural forms in which they are expressed, etc., if this is what we intend to do. The focus group is also more useful than the interview if we are studying the subjects involved in a situation, as they will complement and correct each other, activating each other’s memory of the situation under discussion.
Mathematical and quantitative data analysis methods
Quantitative methods of analysis are made up of the multitude of mathematical techniques used to examine data (regardless of their source or nature, quantitative or qualitative). The most important mathematical techniques used are (I) statistics, (II) game theory and (III) network analysis.
(I) Statistical techniques
Statistical analysis techniques are usually divided into two main parts: methods and techniques of descriptive statistics and methods and techniques of inferential statistics.
1) Descriptive statistics contain those analytical procedures by which certain characteristics of the data under analysis are determined and presented. The characteristics determined summaries certain aspects of the structure of the data and are presented in the form of indicators of (A) central tendency, (B) scatter or dispersion and (C) data distribution.
A. Indicators of central tendency:
(a) The arithmetic mean is that value obtained by dividing the sum of all values by the number of values
(b) Median is the value of the statistical individual at the center of the ordered statistical data (series), i.e. the value with respect to which the number of lower values equals the number of higher values.
c) Mode is the most frequent value of a statistical data (series).
B. Dispersion indicators
Without going into the mathematical details of the definition of scatter indicators we will mention the most relevant and frequently used ones:
a) The Gini index is the average of the differences of all pairs of two different values of the data (series).
b) Mean deviation from a given value of: the average of the absolute differences (in mode) of all the values of the data (series) from a given value of which may or may not belong to the statistical data (series).
(c) Variance: the mean of the squared differences of all values from the mean of the values of the statistical data or series.
(d) Standard deviation: the square root of the variance.
C. Frequency distribution and indicators of the shape of the distribution
The frequency distribution of values is generally presented in tabular form, known as a frequency table, to which a graphical form is associated to better capture the structure of the data. As far as the graphical form is concerned, we can opt for a pie chart or a bar or rectangle graph called a histogram. If you are targeting a broad, non-specialist audience and are not comparing indicators, then we recommend using the pie chart as it is more intuitive. If you are addressing a specialist audience and/or comparing indicators, we recommend using histograms.
a) Skewness indicators: measure the symmetry of a data distribution with respect to the central values, particularly the mean.
b) Kurtosis indicators: measure the difference in height of a data distribution from the normal distribution.
Inferential statistics are those data analysis techniques whereby we generalize to the population level the results obtained from a representative sample. The most common generalization techniques are:
- a) Estimation of means: starting from an average value obtained at the level of a sample we determine with a certain level of confidence the confidence interval on which the parameter lies at the population level.
- b) Estimation of proportions: starting from a proportion obtained at the level of a sample we determine with a certain level of confidence the confidence interval over which the parameter lies at the population level.
- In inferential statistics, there are a number of hypothesis testing techniques known as statistical significance tests (Z, t, χ2 (hi-squared) ANOVA, etc.)
- (c) Statistical significance tests: used to determine in probabilistic terms whether the difference between two or more quantities of which at least one was obtained at sample level is real or due to sampling fluctuation. The proper interpretation of these tests is in terms of conditional probabilities: ‘if we adopt the null hypothesis (i.e. assume that the difference between magnitudes is due to sampling fluctuation), the probability of obtaining the observed difference between magnitudes is x%’.
- Significance tests aim to check the statistical significance of the difference between two or more quantities. However, the arsenal of quantitative methods of analysis is not exhausted by these techniques. Quantitative methods of analysis have been enriched by powerful mathematical techniques, such as multivariate analysis in particular, for establishing and testing the association of two or more quantities.
- (d) Variable association: used to determine both the association of two or more variables and its strength for each type of variable combination (nominal, ordinal, quantitative). The basic methods used concern the association of qualitative dichotomous (coefficients φ, Y and Q), categorical (coefficients C, V, λ, τ), ordinal (coefficients τ, γ and d), quantitative (correlation coefficient r, linear regression, etc.) variables. Advanced research methods are grouped under multivariate analysis and represent an extension of basic techniques and methods by including a larger number of variables and investigating the relationships between them. Multivariate analysis techniques are classified into:
- (i) dependence techniques (path analysis, log-linear analysis, multilinear regression, MANOVA, etc.) used to explain or predict a dependent variable(s) in terms of independent variables.
- (ii) Interdependence techniques (factor analysis, cluster analysis, etc.) used to deduce the structure of the data under analysis by revealing relationships between variables, cases or objects.
Game theory tools imported into data analysis
Game theory is a mathematical tool for analyzing interactions between different actors. The instrumental virtues of game theory can be seen in its diverse range of applicability: economics, biology, ethics, political science, international relations, etc. The fundamental assumption of game theory is represented by the rationality of actors, more precisely by the principle according to which the choice of an action is determined by the maximization of benefits. In addition to this assumption, the distinctive feature of the interactions studied in game theory is the strategic way in which players or actors choose actions, namely that the choice of actions by each player is influenced by the actions available to the other players. Situations in which each player’s action decisions are influenced by the decision choices of others are described in the literature as strategic.
Strategic games are determined by three elements: the crowd of actors/players, the crowd of actions/strategies of each player (the crowd that forms the player’s agenda), and the preference relationship by which each player ranks the crowd of all action profiles. Based on these elements we define the fundamental concepts and analytical tools of game theory: best response, dominated strategies (strong, weak), dominant strategies, Nash equilibrium, perfect subgame equilibrium, Pareto optimal, iterated strategy elimination procedure, reverse induction, etc.
The classification of strategy games can be made according to different criteria, such as the number of players (two-person games, n-person games), the information of the players about the actions or action options of the other players (complete information games, incomplete information games), the information of the players about the previous decisions of the other players (sequential games, simultaneous games), the distribution of benefits or utilities (zero-sum games, nil-sum games), the probability of choice of actions or strategies (pure strategy games, mixed strategy games), etc.
Network analysis is a mathematical approach to the various relationships between the entities that constitute the units of analysis of the research. The relationships on which the network analysis focuses are obviously the relationships relevant to the research objective. While statistical techniques and quantitative approaches address the properties of the units of analysis (socio-demographic characteristics, socio-professional characteristics, social characteristics, economic characteristics, attitudes, perceptions, etc.), network analysis addresses the connections that exist or are established between the units of analysis. In this sense, we can say that network analysis is an approach complementary to statistical techniques. The relationships between units of analysis constitute a network, and the study of this network is mainly carried out using graph theory. Thus, the properties (symmetry, transitivity, etc.) of the relationships (edges, in graph theory) between the units of analysis (nodes, in graph theory) are captured mathematically, models of the networks formed by these nodes and edges are developed, the properties of these networks (density, centrality, etc.) are analyzed and their dynamics are established on the basis of these properties.
Graphical representations of social networks are called sociograms and are, as long as the network is small, a powerful heuristic tool. In sociograms, nodes are represented by flat geometric figures – circles, squares, triangles – and edges by lines – straight or curved – connecting these points. Depending on the properties of the relationships, the lines can be simple or directional, which is marked in the sociogram by an arrow. For example, marriage is a symmetric relationship, so the line representing the marriage relationship between two nodes or two actors is simple, and the graph illustrating a network of symmetric or non-directional relationships is called a non-directed graph. Non-symmetric directional relationships such as an advisory relationship or a helping relationship are marked by an arrow, and the graph containing such relationships is called an oriented graph. Helping learning relationships in a seminar group can be represented by a graph or sociogram in which the edges are directed and express the asymmetry of the relationship. Network analysis establishes, among other things, the structural properties of nodes and distinguishes the pre-eminence of some nodes, which explains, for example, how misinformation circulates among students in an exam. The results of network analysis have an explanatory function as well as a predictive function.
Network analysis is useful in contexts where the attributes of actors are sensitive to the structure of which they are part. For example, the type and amount of information an individual holds depends on the information exchange structure in which they are embedded. Phenomena such as influence, diffusion, contamination, learning, in general, transfers or exchanges between different entities, are studied more accurately and efficiently with network analysis.
Common analysis methods and techniques
We conclude the research methods section by stating that some analysis techniques are common to both methodological approaches. One such common technique is secondary data analysis, which we will focus on below.
Secondary data analysis
Secondary data analysis involves a scientific investigation based on information collected in other studies or research. A significant part of the data collected for different purposes is publicly accessible, so that, having these data at our disposal, we can make or test different hypotheses, we can answer research questions in a qualified way. Thus, secondary data analysis is the use of qualitative or quantitative tools to approach information collected under different theoretical circumstances (for a different research purpose) in order to analyze phenomena in the social sciences. Two advantages stand out in this approach: a) the elimination of the financial and time costs involved in the data collection phase and b) the use of professionally collected data. In this respect, we strongly recommend that students and masters students make extensive use of secondary data analysis. Of course, the approach also has some shortcomings, the most important of which, of a theoretical nature, is the fact that the data being operated on have been collected for other purposes, so some information relevant to the research may be missing. Another disadvantage is the reduced accessibility of data that have been rigorously collected by the relevant institutions. However, the scientific value of research using secondary data analysis is significantly superior to research where data are collected in a more scientifically permissive manner.
We recommend the use of secondary data analysis as a methodological tool in the development of the thesis/dissertation. The advantages of using secondary analysis are a) the elimination of the financial and time costs of the data collection phase and b) the use of professionally collected data. The range of availability of data collected by specialized institutions is large enough to allow research to be undertaken as part of a bachelor or master thesis.
STYLISTIC AND GRAMMATICAL ASPECTS OF WRITING A DISSERTATION THESIS
Writing style is a very important aspect of an undergraduate or graduate paper, as the ability to communicate the results of research in a clear, concise and understandable way is essential in academia and beyond. As we pointed out in Chapter I, a dissertation should be written in an academic language, using specialist terms specific to a given field or topic, without, however, abusing technical terms that may make the paper difficult to read. Also, technical terms used throughout the thesis should be clearly defined either in the footnotes or in a glossary at the end of the thesis if there is a longer list.
A colloquial, informal, journalistic and subjective style should be avoided in an undergraduate paper. An attempt should be made to maintain an objective, rather impersonal tone, with the author’s opinions being expressed without value judgements, but only scientific arguments. The use of the passive rather than the active voice is therefore recommended in such cases, as is the use of the first person plural (‘we consider that…’) instead of the first person singular.
Characteristics of the academic style of writing an undergraduate paper include the following:
- clarity: the sentences used should be concise, complete and precise; it is preferable for sentences to be shorter rather than encumbered by a complex structure of subordinate sentences; the paragraphs of a section or chapter should form a whole, following a logical structure, with each paragraph discussing a single basic idea/argument. For better structuring of the content, it is recommended that each paragraph begins with a key sentence, which summarizes or synthesizes the argument that will be developed in the body of that paragraph through sub-arguments and/or examples.
- the use of relevant examples and comparisons to support and reinforce the arguments used by the author;
- precision: avoid ambiguous terms, vague or indeterminate expressions such as “some studies show that…” (which studies?), “some authors believe that…” (who are these authors?), “according to some opinions…” (who are the proponents of these opinions?), etc. It is also not advisable to use excessive parentheses, which can distract the reader from the main argument, and to use overgeneralizing phrases such as “everyone knows that…”.
- grammatical correctness and rigorous application of punctuation rules, quotation marks and other punctuation that contribute to the neat appearance of an undergraduate/dissertation paper.
The correct use of verb tenses is also an important stylistic aspect of a paper and must respect certain formal rules, which relate to the grammatical agreement of tenses, the subject matter or the type of chapter in the paper where they are used. Thus, the present tense is used mainly in the introduction, conclusions and recommendations sections of the paper, as well as in the body of the paper, when discussing current events/theories/situations, when formulating definitions of concepts or when the text refers to generally valid observations.
The use of the past tense is justified in the following cases: when the nature of the research is historical, referring to past events/situations/figures; when reference is made to theories/concepts/methods/studies carried out previously, or sometimes when discussing the work/opinions of an author/authors who are no longer living – although in such cases, if these works or theories have relevance to the present, the present tense may also be used; also, if the paper refers to research results (usually of a practical or experimental nature), it is advisable to use the past tense to describe them.
We hope that you have found or guide helpful for writing your dissertation. Do you think that university programs should dedicate a class on how to write your dissertation?