After my meeting with LTC Sanborn I began to think of ways in which I can broaden our research question to produce more generalizable and interesting conclusions.  A primary resource in this was “Disposable Income, and spending on fast food, alcohol, cigarettes, and gambling by New Zealand Secondary School Students”(Darling et. al 2006) This study which examined the relationship between the source of high schooler disposable income the amount they spend on the products listed above. This article gave me the idea to expand our research question to the include fast food and tobacco products. This helps expand the nature of our study as previously it could only include Cadets who are of legal age to drink. With the addition of fast food and tobacco, however, I believe we will have a far more inclusive population that perhaps includes a large majority of cadets. Second, I am considering adopting methodology of the study to our current research question. Darling, et. al used data collected through in person interviews of students. This may help our data collection as I believe cadets will be more likely to participate in the study if they are approached in person. Lastly I propose our new research question. What is the relationship between the source of cadet disposable income and their propensity to consume alcohol, tobacco, and fast-food.

Help Received: “Disposable Income, and spending on fast food, alcohol, cigarettes, and gambling by New Zealand Secondary School Students”(Darling et. al 2006)

D. Chase Simpson

During our first discussion on what would be the research question for our group project, Ryan Elliott, Joe Gruber and I began to experience the difficulties with collecting data. Joe’s original idea was to test which class of cadets would be more willing to disregard VMI superstitions. Joe had the idea to use the willingness for a cadet to pass between limits gates as a test of this. My first thought was than an unbiased sample would be very difficult to collect for numerous reasons. First, undetected observation would be extremely difficult because passing between limits gates is not the convenient route to most any point on post and therefore it would be very likely that cadets may have avoided the route simply to save time. Therefore, it would be required to inform the cadet in question to the nature of the study and to gain information through survey data. This could potentially have a biasing effect as cadets may change their behavior when they know they are being observed. Second, splitting the sample by class may also have a undesired effect, especially in regards to rats who may not be familiar at all with the superstition in question.

In the first chapter of, “Essentials of Political Analysis” Phillip Pollock defines what a concept is in political science and describes the methods through which a concept can be measured. Pollock sets out two critical terms in political research, “conceptual definition” and “operational definition.” Pollock defines the conceptual definition saying it, “clearly describes the concept’s measurable properties and specifies the units of analysis to which the concept applies.” Complementary to the conceptual definition is the operational definition which, “describes the instrument to be used in measuring the concept”. These two definitions can be particularly difficult to formulate for more abstract concepts such as liberty or democracy. The principle qualifications for a property of a concept is that, “they must be concrete and, they must vary. As Pollock states a property of a concept cannot be a concept itself.

In, “Conceptualization and Measuring Democracy: A New Approach” Michael Coppedge and John Gerring demonstrate the difficulties of defining abstract concepts using the example of democracy.  Early in the piece Coppedge and Gerring concede that there is no chance for total agreement on the definition of democracy. Instead, they attempt to resolve some of the more commonly argued points. But as they state, accepting that there will be no consensus on definition concedes that there will never be a universal measure for democracy.

The argument put forth by Coopedge and Gerring is critical to the understanding of concepts in political science. Their work, while specifically on democracy, can be extrapolated out to almost all abstract concepts in political science. Just as there is no definition for democracy there is no concrete definition for autocracy, liberal, conservative, communism, socialism, or any other “-ism” in the political world. This present a challenge for any researcher in the field. Therefore, in order for any research to be accepted by the majority of the Political Science community it must begin with a logical and consistent conceptual definition.

Equally important to the concept being measured is the explanatory mechanism in social science research. In “Nuts and Bolts for the Social Sciences,” Jon Elster, describes the role of casual mechanisms in a strong theory. The casual mechanism is critical for social science research because it provides an explanation for why two variables are correlated. Correlation without an explanation as to why these two variables may be connected provides a very weak causal explanation and provides little to no value to the discipline. There is the commonly cited example of the correlation between ice cream consumption and murder rates. Stating that as ice cream consumption increases so does the murder rate. There is no causal argument between the two, however, and they are both most likely caused by a third variable such as higher temperatures.

The difficulties in trying to prove causality are demonstrated in Donald Campbell and Laurence Ross’s study, “The Connecticut Crackdown on Speeding: Time series Data in Quasi-Experimental Analysis” In the late 1950s the government to reduce traffic fatalities introduced harsh penalties for speeders. When examined at a single data point, the difference between 1955 and 1956 it appears that there was a significant decline in the number of traffic fatalities in the state. Also in comparison to the average of the other New England states it appears that there was a significant decline in fatalities after the implementation of the traffic laws. When compared to each northeastern state individually, however, Rhode Island experienced an even sharper drop in traffic fatalities, without the introduction of strict traffic laws. As argued by Campbell and Ross, this comes together for an overall inconclusive study on the effect of the speeding laws far from the certainty with which the governor declared the laws a major success. The inconclusive nature of the statistical analysis requires, if a relationship is to be determined, a reexamination of the casual mechanism as part of an overall improved theory

In, “the importance of research design in political science,” Gary King and Robert O. Keohane present a model for effective research design and theory development. King and Keohane make clear that their proposed methodology is not exclusive to quantitative or qualitative research and instead argue that the differences between these two methods are minimal regarding “the logic of inference.”  King and Keohane begin their paper with what they believe to be the fundamental problem with social science saying, “Social science statistics focuses too little on research design, and its language seems arcane if not impenetrable. The numerous languages used to describe methods in qualitative research are diverse, inconsistent with jargon and methodological advice and not always helpful to researchers.” In the face of these traditional differences King and Keohane argue that there can be a common language and method of inference between both quantitative and qualitative methodology. King and Keohane emphasize the need for social science research to pursue generalizable inferences about how the political world operates. To merely describe events or seek absolute truth and law is to miss the capabilities of social science research. Lastly, King and Keohane emphasize the mode of inference they provide is based on the idea of “triangulation” this methodology is about bringing about the most data sources possible in relation to the theory presented.

In, “The Message Overwhelming the Message: Ideological cues and perceptions of Bias in Television News,” Joel Turner sought to explain the persisting belief that major cable news networks, CNN and Fox News, are ideologically biased despite there being no empirical evidence of this.  Turner theorizes that it is not the news being reported that causes viewers to perceive bias but instead the network affiliation. To test this Turner was able to utilize a unique methodology for social science research by conducting a controlled experiment. While social science research is generally constrained to the use of real world scenarios, turner could use a fake news studio to produce replicated news stories and then present these stories to unknowing participants to achieve a high degree of control to his study. This strength in research design and methodology certainly bolstered his conclusion that the news content of the major networks was not biased but it was instead people’s perception of the news network that made them believe in considerable bias.

In “Political Science in the United States: Past and Present,” (1985) David Easton sets out the historical development of political science research. Easton begins by defining Political science as “the study of the way in which decisions for a society are made and considered binding most of the time by most of the people. Easton then goes on to claim that political science has passed through four distinct phases of development: the formal, traditional, behavioral, and post behavioral periods respectively. The formal stage of political science was, according to Easton, centered around the role of political institutions. Political scientists during this period believed that the key to understanding how decisions were made in society could be found in understanding how political institutions operated. Later, the researchers of this period began to focus on the informal structures within political institutions. Prominent legalist thinkers included future United States President Woodrow Wilson. The second period of political science research, the traditional period, began to move the center of political research from the formal institutions and towards the influence of political parties on decision making in government. This period, still quite immature methodologically, focused on “mere description and the collection of information about political processes than to overarching theories about how they operate. During the third phase of political science research, The Behavioral stage”, the modernization of the discipline truly began in earnest. As stated by Easton, “Without question, this is the central transformation that has occurred in Western political science in this century. Behaviorism believed that human behavior could be measured and even predicted through empirical testing. This idea rapidly increased the rigor of political science research. While behaviorism greatly benefited political science research, its shortcomings, predominately its pursuit of “law-like regularity” in human behavior triggered what Easton called a postbehavorial revolution. It is important to note, however, that the this did not signal a rejection of behaviorism but instead a return of traditional political science thought in the context of more rigorous methodology and empirical testing

As I finished Colonel Foster’s research design course in the spring of 2017 I feel that I have gained a strong foundation in conducting literature reviews, developing theories and formulating hypothesizes. I have created what I believed to be a strong research design but have so far lacked the experience of conducting true statically tests of my theory. This is my primary goal for the course but also the source of my largest anxieties in regards to the course materials. I feel that I need continued education on operationalizing variables, utilizing computer models to run statistical tests, and developing proper controls for my tests. I also hope to develop my abilities for gathering enough data to conduct a proper test of my theory. As a dearth of adequate data was the primary road block for my research this summer. Also, as shown in my first quiz grade it would appear that I still need to figure out the difference between inductive and deductive approaches