This chapter will provide an introduction to the methods used to conduct social science research. The approach of the chapter is to provide guidance in principle rather than instructions in detail. Therefore, it will start with an overview of how social science research uses ideas and facts to generate knowledge. Then, it will move to a brief description of the standards of scholarly research. This will be followed by an examination of how theories and data are used in social science research. The chapter ends with sections on how to structure a research paper and how to attribute and document sources.

Important concepts will be introduced and defined throughout the chapter. However, it will be useful to define the following key terms from the start:

Method: the systematic application of reasoning to the analysis of data.

Analysis: "1. The breaking up of any whole into its parts so as to find out their nature, proportion, function, relationship etc." (Webster’s Deluxe Unabridged Dictionary, 2d ed.).

Theory: A set of ideas that explain observed phenomena in terms of underlying relationships to other phenomena.

Data: Qualitative and quantitative observations of the world.

Empirical: Having to do with the analysis of data

The aim of scholarly research is to increase our knowledge about the world. But what is knowledge? Is it simply the some total of factual information we have about the world? Astrologers have very detailed information about the movement of planets and stars but they ask questions and find answers that are very different from those of astronomers. This is largely because astrologers have a different understanding of how observable events (or phenomena) are related to each other than astronomers do. Given the same information, astronomers and astrologers will claim to know entirely different things about the world.

Therefore, one can define knowledge as a body of facts, and ideas that interpret and explain them. The importance of this definition is that it points to two approaches to gaining knowledge. Theoretical research focuses on ideas about how the world works and empirical research focuses on observing the world as it works. Though individual projects may be purely theoretical or empirical, social science research as a whole depends on the interaction of theoretical and empirical research to expand our knowledge of the world.

The methods of social science research emphasize the use of both theory and data (ideas and facts). Theories explain how two or more general classes of phenomena (e.g., income inequality and political violence) are related. In most empirical research, data is collected to test the generalizations that theories produce. Theories that survive the testing process (i.e., "hold water") are used to explain particular events or to predict the outcome of future events. Table 16-1 illustrates the role ideas and facts play throughout the research process.

Table 1: Interplay of Facts and Ideas in a Research Project

There are several substantive standards for research in social science in general, and IR in particular. Though they are often labeled and categorized in different ways, the following list captures the main criteria:

Interest: The research topic should be positive, significant, and general. A positive is non-normative and deals with the world as it is not as it should be. A significant topic is not trivial and does not leave the reader asking "So what?". Research should also focus on events or classes of events that are likely to occur again and thus are general in nature.

Originality: Research should extend the existing body of knowledge. This can be done by adding new theories or data, by refining existing theories or data, and/or combining existing theories and data in new ways. It is incumbent on researchers to review the literature on their topic to ensure that they are not "reinventing the wheel".

Cumulation: Though it should be original, research should also build on existing theoretical and empirical research. Researchers are expected to use the information and ideas of previous researchers as a foundation for their research. In doing so, researchers must be candid and critical about what they take from others. That is researchers must properly attribute their sources to ensure there is no doubt which ideas are original and which are not. They must also scrutinize the ideas and cross-examine the information they use to ensure that the foundation of their research is solid.

Method: Research must have a discernable methodology that is appropriate, transparent, and persuasive. There are many different methodological approaches from which a researcher may chose. The choice of an appropriate methodology is often governed by the topic and the existing literature on the topic. Whatever methodology is chosen, it must be transparent to the reader how the research is conducted. That is, it must be apparent to the reader how each step of the research was conducted. This standard is that readers should be able to reproduce the research themselves (thus this standard is often described as reproducibility). Finally, the research must be persuasive in that it must support the conclusions of the researcher’s analysis and there must have been the possibility that the research could have pointed to a different conclusion.

Results: Though the results may be qualified or ambiguous, research is not done until it comes to a conclusion. From the beginning of the research, it should be clear what the research is trying to explain, or what question it is seeking to answer. At the end of the research, the researcher should be straightforward about presenting the explanation or answer produced by the research.

Theories explain observed events in terms of unobserved relationships to other phenomena. The events that theories explain are called dependent variables and the phenomena that are used to explain the dependent variable are called independent variables. Theories tell stories about how independent variables affect dependent variables. They also make predictions, called hypotheses, about what will happen to the dependent variable given different values of the independent variable.

Theories are very useful for analyzing particular events because they suggest likely cause and effect relationships. In so doing, they point out what data is important and what data is not. This filtering function simplifies the data collection task and prevents the researcher from being overwhelmed by data. Theories not only tell the researcher what data to collect but also what different the data means. Thus, by providing a framework for interpreting the data, theories also simplify the analysis task.

One might object that, by driving the collection and interpretation of data, theories bias and limit the analysis. To a certain extent they do, but, as a practical matter, researchers must limit the data they collect and analyze. Whatever bias a theory may impart on the research, it is preferred to the bias that results from basing research on ad hoc explanations, or on data that happens to be the most readily available. However, researchers must examine the "track record" of the theoretical perspective they select and be aware of the limitations it imposes on their analysis. For this reason, researchers usually employ more than one theoretical perspective when interpreting an event.

Finding Theories: Choosing a theoretical perspective or perspectives upon which to base one’s research is perhaps the most important part of the research. The selection and development of a conceptual framework of analysis is also the most intellectually challenging task a researcher faces. For this reason many students try to bypass this step and go straight to analyzing data. Unfortunately, in not making a choice, they have chosen to develop ad hoc explanations based on unsystematically collected data. This is an approach that is acceptable in journalism but not in social science.

Therefore, once a topic is assigned or selected, research must begin with a survey of the literature to identify the relevant theoretical perspectives. The Pol-Mil Factors Handbook presents theoretical perspectives on many different topics, and thus, it is a natural starting point. Students should also conduct a search of the scholarly literature on their specific topic to identify additional theoretical perspectives. Students should search academic journals, libraries and publishers for the latest scholarly articles and books on their topic. The emphasis here is on scholarly works because these will have more developed theoretical frameworks, as well as discussions of and references to the larger literature on the topic.

Using Theories: Once the various theoretical perspectives have been identified, the researcher must choose which one or ones to use. The researcher may choose to refine a theory, combine it with others, or develop a new theory. The choice is generally governed by the theory’s record at producing results and its applicability to the exact question the researcher is interested in answering. Most importantly, the theory’s dependent variable must be the same as that of the researcher. If the researcher is trying to predict how a war will end (the researcher’s dependent variable), then theories that explain when wars start (the theories’ dependent variable) are not applicable.

The goal is to create a conceptual framework that defines key concepts and specifies cause and effect relationships between them. Concepts are general characteristics or aspects of events under study. For instance one might describe events in Chechnya as a war, an ethnic conflict, a rebellion or a counterinsurgency. Each of these descriptions is a concept that needs to be explicitly defined before its relevance to the particular events can be examined. A good theory about the cause of ethnic conflicts will provide a clear definition of ethnic conflicts and well as of the causes. It can not be overemphasized that it must be clear what aspect of the event the researcher is trying to explain (the dependent variable), and what aspects of the event the researchers is using to explain it (the independent variables).

Once the conceptual framework has been established, the researcher is ready to engage the data. This part of the process is more concrete and systematic, but also more tedious and frustrating. The researcher must identify, find, evaluate, and analyze the data. Each of these steps is described below.

Identifying Data: The researcher must determine what data to collect to measure each of the key concepts in the conceptual framework. If the conceptual framework is well thought out, then it will be a fairly straightforward task to identify the ideal types of data. Of course, the ideal data is rarely available, but the researcher must start out with a good idea about what the best data would be in order to identify what the next best data is.

Finding Data: Given time and resource constraints, most student researchers use existing data, as opposed to conducting primary research to collect new data. Research libraries, such as UNC Chapel Hill’s Davis Library, are one of the most valuable sources of data available. They either contain or have electronic access to an overwhelming supply of data. More importantly, they have a staff of reference librarians who are professionals at connecting researchers with information.

Unfortunately, students often attempt to collect all their information off the internet. While the internet is an invaluable research tool, the websites on it are a problematic source of information. Most high quality texts, articles and data are copyrighted, and are not available to the general public over the internet. The best sources of information on the internet usually require some kind of subscription or access fee.

Fortunately, libraries have started buying access to these services for their patrons. Troy State University offers its student remote access to a number of on-line subscriptions such as InfoSeek and EBSCO host at https://tsulib.troyst.edu/html/login.html. North Carolina Public Libraries offer cardholders access to a wide range of subscription services in a service called NC Live. A description of these services can be found at <http://www.nclive.org/>. UNC Chapel Hill cardholders have access to NC Live and a number of other services described on their website at <http://www.lib.unc.edu/>. NC Live and most of UNC Chapel Hill’s services require the researcher to physically be in the library.

Evaluating Data: Researchers have a responsibility to critically evaluate the origin and quality of the information they use. This task has been greatly complicated by the explosion of information available on the internet. The low cost and relative anonymity of web publishing allows not only for a diversity of ideas and viewpoints but also for a great deal of dubious or false information, and outright propaganda. Because web publications have typically not been through a review process, researchers must be much more critical of web publications than of traditional publications.

Generally, researchers must evaluate both the sources and content of their data. To evaluate the data’s source, the researcher must identify the author of the data and how the author originally acquired the data. The researcher must also critically examine the reliability and objectivity of both the original author and the original author’s data. Researchers are free to use data from biased or unreliable sources, as long as they acknowledge and describe the bias or unreliability, and qualify their analysis to account for the imperfections of the data.

When evaluating the content of sources, one of the biggest problems researchers face is sorting out propaganda from scholarship. Bodi (1995) lists the following indicators of propaganda and scholarship that students and researchers can use to evaluate both print and web publications:

Analyzing Data: Once data has been evaluated in terms of itself, it must be analyzed in terms of the overall research question. If the data is qualitative in nature, then the researcher must subjectively appraise the data and use logic to derive conclusions. As stated before, the theoretical perspective adopted provides an invaluable starting point for logical arguments. A detailed discussion of how to construct logical arguments is beyond the scope of his chapter. However, useful discussion of using logic can be found at Purdue University’s Online Writing Laboratory at http://owl.english.purdue.edu/Files/123.html .

Statistical methods are usually used to analyze quantitative methods. Again a detailed discussion of these methods is beyond the scope of this chapter. However, Johnson and Joslyn (1995) provide a thorough introduction to the range of statistical methods used in political science research. In addition, Middleton (1997) provides detailed instructions on using Microsoft Excel to conduct statistical analysis.

The structure of the report, usually a paper, is a thing apart from the structure of the research. That is, the sequence of the paper does not follow the sequence of the research. While you can, and should, write large parts of the paper as you conduct the research, you must write the paper as a report of your results, not as narrative of your research. The main difference between the sequence of the research and the paper is that research will end up pointing to one fundamental conclusion while your paper must start and end with that conclusion.

Introduction: In terms of presentation, this section is the most important. It orients your reader to your question, methodology and subsequent conclusions. You want to lay out your paper for your reader by explaining what you’re doing, and how you are doing it. It is essential that the writer and reader know where the paper is going from beginning to end. Generally, one should be able to read the first and last paragraphs of a paper (or the first and last pages of a thesis) and have it make sense. To that end, the introduction as a whole should answer the following questions:

What is the question, hypothesis or story you are telling? By the end of the first paragraph, the reader should know what your paper is about. You need to be very clear exactly what it is you are trying to explain or understand. In technical terms, what is your dependent variable?

What is the state of the art on this subject? Describe the literature as a whole paying particular attention to seminal works and the most recent works. If there is more than one school of thought on your subject, then describe the debate. Your purpose is to point out what has been done and what has been left undone. It may be that no one has looked at your particular problem or it that you think there is a shortcoming in their approach. Your review of the literature should lead into the answer to the next question.

What are you going to do to improve on the literature? Having pointed out the shortcomings of the literature, you must now show how your going to, at least partially, correct them. Explain your methodology and how it improves on existing works. The reader should have a good idea of what it is you are going to do.

Where are you going from here? Your introduction should always end with a road map of the remaining paper or thesis. Explain the structure of your paper/thesis and the logic behind it.

Where will it end? You should always give a sense of what your conclusions will be. You are not trying to surprise your reader.

Methodology/Theory: It may be that your methodology or theory is particularly involved and requires a detailed explanation. If so you should briefly describe your methodology in the introduction and go into detail in a separate section.

Substantive Sections: Here you do the work. It is important that as you present information, you also analyze it. Never assume that a quote, a fact, a graph or figure speak for themselves. Always explain the significance of the data and the implications of the analysis for the overall point you are making. Don’t wait until the paper’s conclusion to point out the major points.

Conclusions: There should be nothing new in the conclusions. The reader should be convinced by now. If, when writing the conclusion, you see a new argument for or against you point, you need to go back and put that in the body. What you are doing here is summing up what you’ve found and suggesting broader implications of your findings.

Reference List: All papers require a reference list or bibliography. One should list only works that have been cited in the paper/thesis. There is a school of thought which advocates a "Reference List" of works cited and a "Bibliography" of a wider list of relevant works. However, I argue that if the work is so relevant, it should have been cited; or, if it isn’t worth citing, it’s not relevant enough to be listed. If there are interesting works on tangential subjects, it’s better to comment on them in footnotes.

Bodi, Sonia. 1995. "Scholarship or Propaganda: How Can Librarians Help Undergraduates Tell the Difference?" [online] Journal of Academic Librarianship. 21: 21-26. [Accessed may 5, 1999] Available on EBSCOhost.

The Chicago Manual of Style, 2d Edition (CMS). 1993. Chicago: University of Chicago Press.

Johnson, Janet Buttolph and Richard A. Joslyn. 1995. Political Science Research Methods. Third Edition. Washington, DC: CQ Press.

Middleton, Michael R. 1997. Data Analysis Using Microsoft Excel. Belmont, CA: Duxbury Press.