This course is generally all about how we know things. More particularly it is about how we know things in politics. This is very very important. If you understand the methods by which people find out things, you are several steps ahead of everyone else when you face a problem and require finding the knowledge/information to address that problem. And that is what life and work is all about--solving problems. So we are giving you a skill in this class that will have valuable uses way beyond the time when all that you learned about South Carolina politics or the United Nations is no longer accurate. We will be learning skills that apply way beyond just political problems. These skills can earn you money, big bucks! Ok, enough motivation.

Generally speaking, all knowledge comes to us in four ways: sensory experience, reason (or logic), authority, or intuition (or faith). Let's talk about each. 

1. Sensory experience. This is the basis of scientific knowledge. It depends on the notion that we can use our senses in ways that allow us to agree on what we observe. Things like the meter reads 3.5, or the color is red, or it has a salty taste, or they answered "no" to a question, or it feels rough, or it smells sweet. To the sense that we can all agree on what we saw, felt, smelled, touched, or heard, we have created knowledge about something, knowledge that can be transferred from one person to another without disagreement. This is what we call "intersubjective transmissibility." Sounds pretty straight forward. But, if you think about it, it is not so simple. We often disagree about what was observed because some of the senses are subjective. This is why clearly laying out how the observations are to be made is so critical in science (later you will learn that this is called "operationalization.") 

2. Reason or logic. Facts can be put together to get to other facts. The rules we use for this operation are called logic. Someone who is illogical violates these rules and reaches conclusions that we do not accept. Most of us have developed a sense of what is logical or not. Unfortunately, we make a lot of mistakes, as any of you who took a logic course well know. We commit many fallacies without knowing it. Examples include appeals to popularity (all the kids are going, daddy, so can't I go?) and appeals to pity (if I don't pass this class, my mom will be really mad at me). Generally, all logic falls into two kinds, inductive and deductive.         

Deductive logic starts with something that is pretty close to universally true about a class of things and then draws conclusions about some member of that class that must be true if the generalization was true. You can something of this form of argument in what politicians often do in campaigns. They often campaign against politics and other politicians. They start with the generalization that most of us hold to be universally true--that all politicians are crooks. This means of course that their opponent, who is a politician, is also a crook. And what about them? They display themselves as an outsider, as a nonpolitical politician, so the generalization does not necessarily apply to them. The problem we should have with this is two-fold. First, the generalization is not universal, and second, running for office does make them a politician, especially if they win. If the premises must be true and the logical form is correct, then the conclusion must be true. If either of these conditions do not hold, we cannot be certain about the truth of the conclusion--it may or may not be true.         

Inductive logic goes in the opposite direction. We make a number of observations about specific things and then conclude that if this observation holds true about all these specific things, it probably holds true for most all of the other things that have the same characteristics as those we originally observed. Wow, what a mouthful! 

Let's try an example. We observe that this businessperson is a Republican. Then we observe that another is, and another, and so on. We may take a sample of businesspersons and find that 80% of them are Republicans. So far we are just in observable facts. (You will note that the senses are involved here too as we must agree on the observation that each is a Republican.) Then we make the inductive leap. We say that if 80% of this sample are Republicans, then it is highly likely that around 80% of all businesspersons are Republicans. As you can see, all survey research is based on inductive logic. 

3. Authority. Both of the first two sources of knowledge play important roles in science. Authority does as well, but in a different way. If you think about it, most of what you get in college is by this route. How do you know what hormones are produced by glands in the body? By the authority of the biology or anatomy professor. There are almost innumerable facts that you accept as fact on the basis of the authority of those who teach you in your classes. Think about the next lecture you listen to--of even reading this one--almost all of the content is based on the authority of the person delivering it. 

In court cases, lawyers hire authorities to present facts, or dispute facts, as "expert witnesses." A few years ago I had such an experience. With my inexperience in playing this role in courts I was at a disadvantage, but the lawyer who hired me thought that I might have even more authority since I was not a "professional hired gun" who makes his living this way. Presumably, my inexperience in legal testimony enhanced my authority in scientific matters because making a living through legal testimony might undercut objectivity. 

The real crux of the issue here, just as in court cases, is whether the authority is credible. Are they really an expert in the matters about which they are speaking? On tv you see actors portraying experts all the time--to lend authority to their claims that this or that works better than something else. Political endorsements of candidates by well known and well respected figures is another example of knowledge by authority. But be careful--a sports star may not be an authority on politics any more than I am an authority on biology or religion. Make sure that an authority's credentials are relevant to the subject. 

4. Intuition, or faith knowledge. Have you ever had a feeling that something was true or a premonition that something was going to happen? This is intuition. It is not based on observation, on logic, on authority. Rather, it is based on some internal guide that is not observable to anyone else. Sometimes your feelings turn out to be right, but you really have no way of knowing beforehand. 

You may try and use inductive logic to lend credibility to your feelings. It usually goes something like this. You observe that your premonitions in the past have usually been correct, maybe 80% of the time, so you project to future premonitions. This is usually quite shaky because we are often selective in what we remember as success and failure in the past. Moreover, even if we are objective in collecting past data, unless the current intuition involves exactly the same kind of event, past behavior may not be relevant. Just because you are good at predicting who will win close political contests does not make your prediction about auto accidents or an afterlife correct. 

This kind of thing is outside the realm of science. Religious faith may be true, but it is outside of science. They are not necessarily incompatible; rather, they are different kinds of truth. This is, of course, the problem with what is popularly known as "scientific creationism." It is an oxymoron. It may be true, but it is not demonstrable thru the methods of science. It is not open to change as are all scientific theories. It ultimately rests on faith that some great mover is behind all that happened which we can observe.  The same can be said of the most recent incarnation from the religious right, "intelligent design." 

The question of the relation of faith knowledge to scientific knowledge has been fraught with political controversy. While those in the scientific world see no great problem because these areas of knowledge are different and not incompatible, many religious fundamentalists see scientific knowledge as contradicting their own religious beliefs. So for a long time now they have sought to use political power to give their own nonscientific views equal billing as science or to exclude scientific knowledge that they feel is in conflict with their own faith knowledge from public schools. 

This question has arisen in the area of teaching evolution perhaps more so than in any other subject area. A few years ago it arose in a political decision in Kansas to exclude teaching evolution in public schools. Reprinted below is an editorial by a scientific authority, Stephen Gould (Time, August 23, 1999, p.59). It is an excellent article that reinforces the point I am trying to make about the differences in these areas of knowledge. So now you have two authorities on this!

 Every fact that we establish as a scientific truth must be established thru the scientific method. That is true in ALL the sciences, not just political science. For example, in my role as an expert witness in the summer of 1999, I was critiquing the work of a marketing psychologist, who was the expert on the other side. The basis of my critique was that his work violated the principles of the scientific method, not that I knew more about the intricacies of "mall intercept" studies or product recognition research than he. In fact, he knows far more about the details of those kinds of studies than I do. But nevertheless, in doing a research project that establishes some fact as scientific, he must follow the proper steps. Failing to do so renders his findings as little more than conjecture, as something the courts should not pay attention to in rendering a verdict.     

So let's look at the steps. They will form the outline of the course. We will talk about each in turn and do problems associated with each step. On some steps we will spend more time, like the analysis step, which involves learning how to use a few statistical tools. On others we will only spend a little time, like the first step of problem selection. Sometimes researchers combine steps and do them in a little different order, but they are all essential. If it claims to be scientific, it has to be established following these steps.

  1. Problem selection--what is important enough for me to spend my time and money on studying?         

This step, more than any other step, is normative. It allows values to get into our research. Note the word "important." That is a value question. Sometimes you may do something because you are told to by your teacher or boss. Or you may do research because some group with funds are willing to give you a grant in that area. Is it any wonder why many more research projects are done on the weapons and strategy of war than on peace? The military industrial complex can make a lot more money on war than on peace. College professors search lists of possible grants and design research to get those grants. If they don't, they may not be given tenure and promotions and salary increases. To put it another way, those with resources play a powerful role in deciding what does and what does not get done in research. Another example--much more research was done in the 1960s on how to modernize textile mills than was done on the cause and prevention of byssinosis, or brown lung disease. Why? Textile mills were more interested in profits. In fact, they tried to discourage research on byssinosis by not allowing researchers access to medical records or the mills. 

2. Theory--what do we already know?

Before designing a study, one should first find out what is already known. This involves looking at the literature in an area. Once that was mainly looking at scientific journals in libraries, but today much can be done on the Internet. The problem with the Internet is that it is wide open, that any idiot can place information there, whether it is correct or not (like Wikopedia). In other words, it carries less authority. So we must make more evaluations ourselves--we must be more discriminating. 

What does theory tell us? That depends on whether it is normative or empirical theory. By definition, empirical theory is an idea or body of thought that attempts to describe, explain and predict. Describing, explaining, and predicting are the overall goals of science. Normative theory gets into prescribing, telling us what we should do. That is not what this course is all about. 

Looking at theory helps us in several of the next steps, developing ideas about what the key concepts are, how these concepts can be measured, how they are related to other concepts in hypotheses, what we might do to extend what is already known, and how we can gather data to go about that task. Sometimes we might replicate a study exactly, or we might change just one thing to see if it makes any difference. Other times we can design a new study to answer related questions. But looking at what has already been done is essential in deciding either how we can do it better or how we can build on what is already known and extend knowledge. It is an incremental process, like sea creatures building great reefs--each builds on the other, and sometimes great parts fall away and have to be rebuilt. Such is the nature of the scientific research process.

3. Concepts--what are the key ideas involved in the problem and theory?

All the sciences, social and behavioral and life and physical, are filled with concepts. Examples are nearly endless: energy, entropy, living, normal and abnormal behavior, psychosis, fear, serenity, community, power, effectiveness, culture, power, influence, alienation, confusion, and so on. Some concepts are quite concrete, like weight or age. Others are quite abstract, like the legal concept of "trade dress," which is defined as the "total image of a product or good or service... ." In political science, political culture is an abstract concept while voting participation is more concrete. As you will see when we go deeper into this area, we will have to be very careful to use accepted concepts where possible and develop new ones when necessary, but avoid using terminology that makes readers think we are talking about something else. If you talk about alienation, but are using a different definition for alienation than other researchers, you may end up adding more confusion than insight to the body of scientific knowledge. I know that most of you will not be doing original research, but you will certainly be evaluating research done by others. One of the first things to do is see if they are using concepts that are generally accepted and defined as others have done.

4. Operationalization--how can we measure these concepts?

This is a huge step that is fraught with danger. It is where much research fails. It involves much creative effort. How do you get from the concept to something that can be observed using the five senses in a way that anyone else can follow the same steps and make the same observation with the same results? If you like, you can think of this as measurement rules. the rules must be very precise. If not, someone else will apply them and perhaps get different results when measuring the same object. In surveys, for example, if I change the wording in a question, I can get two different answers from the same person. We will spend a lot of time on this. It is absolutely critical.

5. Hypotheses--how does one variable affect another?

Oops--you see there is a new term in this one--variable. What is this? To use the terms that we have introduced up to this point, a variable is a concept that has been operationalized. We have rules of measurement to apply to the properties that some object has and the result is something more concrete--a variable. Hypotheses are relationships that you expect to exist among variables because of what you understand about theory. For example, the theory of balance of power suggests that nations do what is necessary to keep any one nation from overpowering them and threatening their survival, regardless of ideological considerations. So we might use this theory to explain alliances (a variable) that exist between nations that fear a powerful (another variable) neighboring nation and have a hypothesis like the following: The greater the military expenditures of a nation relative to neighboring nations, the more likely neighboring nations will have formal military alliances with each other. We will also spend a great deal of time on this topic. I want you to be able to critique and write well constructed hypotheses.

6. Data Gathering--how should we collect information to test our hypothesis?

When you to out and identify the groups of objects to which your hypothesis applies, the population, and then select a sample out of this population, and then take the measurements on each object in the sample (called the unit of analysis), you have engaged in data collection. This involves a lot of very technical stuff, including defining the population, finding the actual units of analysis in the population, deciding how many to select, deciding exactly which ones to select, when and where to select them, and keeping track of the process (this can involve training hiring, training, and managing interviewers). We will go through all these things and will actually do a survey.

7. Analysis--what do we do with the data once it is gathered?

Once the measurements have been made, you have raw data. These data have to be coded so that they can be analyzed. We will learn some easy statistical techniques for analysis: tabulation, crosstabulation, analysis of variance, scatterplots, correlation, regression. This can be fairly complex if we had to do it by hand, but we now have sophisticated computer packages that make it very easy--we will be using MicroCase. 

What is the point? We want to see if the data support or fail to support our hypothesis. How certain must we be that it does support the hypothesis? Here we will learn about significance tests, which allow us to calculate the probability that we can have a relationship like we found in a sample when  no relationship exists in the population. We will also have to learn to control for effects that other variables may have on the original hypothesis. 

For example, it is well known that women are more likely to vote Democratic than men. Is that true of all groups of women? Or are there other factors that might affect that relationship? Perhaps women are more likely to be poor, and that is why the vote more Democratic. So we might see if the relationship holds for all income groups, looking at each group separately. If that makes no sense now, it will later.

8. Theory reformulation and Reports--how do we draw conclusions from the data and then write it up in a report?         

After you draw some conclusion about the hypothesis, you must ask yourself how this finding affects the original theory. The data may simply support the theory. It may qualify the theory, finding that the theory seems to apply more to some subgroups that to others, or under some circumstances more than others. For example, a USCA student once looked at regional culture and hypothesized that urban northerners were less likely to be polite and helpful than urban southerners. After doing a very interesting telephone survey in which he posed as someone needing information, he found that gender had more to do with politeness and helpfulness than region. So regional culture, at least among those in urban areas, may be less important than gender socialization, in so far as the behaviors of politeness and helpfulness. The theory is modified.     

If you think about it, you will see that even in solving everyday problems, we often use a process that is roughly analogous to the scientific method. You are presented with a problem, say, making yourself financially independent. You ask around to see what we already know about making money. You find a theory in talking to expert counselors at your high school that education seems to be the key. This involves two concepts, education and financial well-being. As in many everyday problems, the concepts are pretty concrete to start with. You operationalize them by looking at whether a person quits after high school or gets a college degree. The hypothesis is that those with college degrees will make more money than those who quit after high school. Then you take a very unscientific sample of friends and family members and see who is making what. You analyze the data in a very rough way, noting that most all of the college grads make much more, except for Uncle Al, who became an auto mechanic and seems better off than about half of the college grads. You conclude that with few exceptions, people do make more with a college degree, supporting the theory. That may explain why many of you are here!

Assignment: Go to any scholarly journal. It can be a political science journal like the American Political Science Review or Public Opinion Quarterly, or it can be a journal from another field. Find one article that illustrates the steps of scientific research. Almost any piece that does original research should work--but a book review or a critique of someone else's research will not work. Then read the article and briefly describe what the researcher says about each step in the process. That means you should have 8 short paragraphs. Note that often they use different terms for the steps than the ones I used. For example, often "theory reformulation" is called "discussion." Often "problem statement" and "concept formulation" and "theory" will be combined in the introduction. Sometimes their "problem" is simply to improve on someone else's research, so the problem statement is implicit in their introduction, where they engage in a discussion of what someone else found. Remember that each step answers a different question. You will be expected to present the article and explain how the steps of the scientific process are followed to the rest of the class.