CHARLES HUGHES: Given that ABA is based on science, and in fact is a science, I thought it might be helpful to talk a bit about what makes something a science. First, let me explain why I believe ABA is a natural science, just as biology and physics are. The basic principles of any physical or biological science are discovered, not invented. They exist independently, regardless of whether we humans have recognized them or not.

So the force of gravity has existed as long as the physical universe has existed, long before an apple fell on Isaac Newton's head and he started thinking about why things fall down, and why we just don't float away. Two, electricity existed before Benjamin Franklin flew a kite in a lightning storm. He didn't invent electricity, he discovered it. And then we started figuring out ways that we could use it.

So the principles of behavior, whereby living organisms tend to repeat behaviors that are reinforced by the environment, or decreased if they are punished by the environment, were in effect long before BF Skinner and others discovered them. So if a behavior resulted in our early ancestors finding food, that behavior was repeated. Conversely, if a behavior resulted in your neighbor winding up in the stomach of a tiger, that behavior decreased, certainly for your neighbor. And once scientists observed and identified scientific principles, they usually began to figure out ways of applying these principles to solve problems through the use of scientific experimentation.

So what is science? Well, that is a very big question, but I think some of the words in this slide get at the essence of what it is. Words such as systematic, whereby step-by-step procedures are applied to study phenomenon so that we can organize that knowledge in a way that helps us understand it better. We are, in essence, trying to understand how the natural world works. Now underlying this presumption that the basic processes of the world and beyond are overall orderly, no matter where you are.

Finally, scientific knowledge is independent of personal, political, social, and economic reasons. Now this is a tough one. What I mean is that to be able to keep science from being impacted by human biases is tough. It's probably impossible to do, so we should try though to minimize it. Now think about the fact that Galileo was condemned by the church for his observation that the earth rotated around the sun, instead of the other way around. Unscientific beliefs have often put scientific findings on hold for hundreds of years because they conflicted with current non-scientific beliefs or biases.

So there are three basic types of scientific investigations. These three investigation procedures allow us to understand natural phenomena, but at different levels. Now the first and the lowest level is when we describe an observed phenomenon. The more systematically we observe something without bias, the more accurately we can describe it. The more accurately we can describe it, the better we can organize and classify the phenomenon. We can then use that data to begin to make hypotheses about why, and under what conditions, the phenomenon occur, and then begin to test those hypotheses.

Now another type of investigation is prediction. After repeated and systematic observations of a phenomena occur, and the data from these observations is analyzed and summarized, scientists often discover that two events, or situations, seem to occur together. That is when one event occurs, another event also occurs with regularity. This is called a correlation. In a sense, we can say that x predicts y.

Now, in your text the authors use the example that the occurrence of winter predicts when certain birds fly south. So flying south is correlated with the onset of winter. Now remember at this point we are not controlling or manipulating these variables, we're only observing that they occur. Thus we cannot say for certain that x causes y, only that x might predict that y will occur.

While some correlations seem to have a causal relationship, we need to be careful. Many times, this leads to faulty assumptions. So for example, because of some correlations it was believed that vaccinations caused autism, a conclusion that has not been borne out. This whole thing got started when somebody noticed that autism typically is diagnosed around two to three years of age, and that is when most kids get their measles and rubella shots.

Another example, in the field of psychology, the pseudoscience of phrenology, which is predicting behavior based on the bumps on a person's head, came about in the 1800s because a doctor named Gall noticed that an executioner, and a boy who liked to hurt animals both had a prominent bump above their ear. Because of these very limited correlations, he made the leap to being that we could explain and predict behavior based on where bumps on the head were located. Now we often laugh when we hear these conclusions based on spurious correlations, but they're still happening and we must guard against making causal predictions from them. So to summarize, we can have two things or more things that are highly correlated, but there may be other variables yet unidentified that will predict a phenomenon. So we need to look more closely at correlations, by using controlled experiments, to see if y occurs only when x occurs.

Controlled investigations provide further information about the level and strength of relationships between variables. Controlled experiments, and the information they yield, are the most useful for any scientific field of study. Controlled experiments manipulate events and variables for the purpose of establishing functional relationships. That is, a functional relation exists when y reliably occurs when x is used or presented. Now we do that by designing experiments that control, what I mean by control is eliminate other possible variables and their effects, so that we can isolate the impact of the variable in question. So we conduct experiments designed to answer questions such as whether the use of a particular math curriculum reliably results in better math performance. Or whether tantrums decrease when, and only when, a particular extinction procedure is used. And we can say with confidence, that other variables did not cause the change in the behavior.

Now before we move into ABA as a science, I want to cover some basic assumptions and strategies that all empirical sciences share. In your text these shared perspectives are called attitudes of science, and there are six of them. The first one is determinism. Determinism assumes that to a large degree the area of phenomena being studied, whether it is particle science, human behavior, gravity, is orderly and lawful. Now it has to be lawful in order to control and predict phenomena.

In other words, we can establish functional relationships such as, when we heat a gas, it expands. Or when gravitational pull increases, more thrust is needed to escape it. Or when a behavior is reinforced, it increases. And so on. Now if the universe was not orderly to a great degree, we really would know nothing. Because we would not be able to predict anything.

The next attitude of science is empiricism. That is, we should objectively and without bias observe the studied phenomenon. Scientists need to proceed without individual prejudices and opinions, but as I noted earlier that's easier said than done, given that we are only human. So that is why we have things such as double blind studies, especially in the area of medicine. It's why we have more than one observer, or person, who measures the change in behavior. It's why replications of the experiment are performed by a variety of people, and so on.

Now as I mentioned before, experimentation is the basic strategy for establishing functional relationships. It is possible that a functional relation exists between two variables such as variable x and variable y, but the only way to establish it, is to control for the impact of other possible variables. But one, or even two replications of a study, is usually not enough to have full confidence about a functional relationship. So scientists rely on multiple replications of these experiments to make sure the results are reliable. Replication of results is a cornerstone of the scientific process. Many times we read about groundbreaking discoveries such as cold fusion, only to find out that nobody else can replicate the findings.

Because science is about providing an explanation of why, how, and under what conditions phenomena occur, scientists hold to something that's called the law of parsimony. That is, they initially examine the simplest explanation over more complex explanations. Thus, the simplest, or the most parsimonious, or frugal explanations need to be ruled out before more complex explanations are considered. When you hear someone say that a theory is elegant, that's what he or she is talking about.

The last attitude of science presented here is philosophic doubt. This makes me think of an expression I heard in one of my statistic courses back in the day when I was a doctoral student. That statement is, in God we trust, everybody else needs data. Now what that's getting at is we should not accept things as fact, unless there is some form of proof. And we need to keep a healthy level of skepticism about that proof. So scientists need to continually question what is considered fact, at any point in time. If they did not question and test and, if warranted, set aside their current beliefs, especially about their own research, we would still be feeling the bumps on people's heads to explain their behavior, and think that the sun revolves around the earth.

The quote on this slide shows that BF Skinner, considered by most to be the father of behaviorism, was a strong advocate of continual questioning about what is held as truth. Out of these attitudes of science, comes a definition of science. One that holds for ABA, as well as any natural science. Take a minute to read it over. Now as you can see, this definition contains such concepts as description, prediction, and control that lead to parsimonious explanations, and a healthy skepticism about what is currently known.

So that's enough scientific excitement for now. This is a good time to complete an activity to evaluate your memory and understanding of some of the content we just covered in this segment. Directions for what to do, and where to find this self-evaluation activity, are at the bottom of the video. Now when you've completed the activity, and read the provided feedback, you could come back to the video and continue with segment two. Or, you could do something a bit more pleasurable to do, and come back to the video at a later time.