©
  University of Idaho
  All rights reserved.
  Psychology Dept.
  University of Idaho
  Design - P&D

As you can see here, there are four or five major methods that we commonly use to examine whether something is good or not. As you can see, each of these is listed here. Let's talk about the first one, what is called systematic or naturalistic observation and we begin our discussion on slide three.

In systematic or naturalistic observation, what you are going to do is observe others. Or, observe some organism and then make some kind of inference about what they are doing. Usually this procedure involves counting some kind of behavior and is often conducted in field settings, not in the lab. An example of systematic observation might be standing on the corner of a major intersection and observing the number of individuals who are wearing their seatbelts. Another observation might be observing behaviors that go on at a particular meeting (such as an AA meeting or a political event). You make some kind of count of something that happens (such as, when people stand up and go "Yeh!"), and then you make some kind of inference about how supportive they are about a particular topic.

What are some advantages of this technique? As we can see in slide four, when you use systematic observation you know that the situation is real and that it's not some kind of artificial lab type of thing. You can also observe behaviors you can't do in the lab. An example might be observing somebody who might be exposed to high amounts of radiation. Or you might observe something in the forest, such as watching bears. lions, or whatever it might be. The big advantage that it provides you with a big picture of what is going on out in the world.

So, what are some disadvantages? As you can see on slide five, if the observer is not being very careful they can distort information. In addition, the results can actually change if the subject becomes aware that they are being observed. Let's give you an example. Let's say you are trying to identify the number of people in a bar who are actually designated drivers but who drink and drive from the premises. So, what you do is go into a bar, you observe individuals who are calling themselves designated drivers and how many drinks of alcohol that they actually order and consume. Now, while you are doing this you are sitting at a table just watching this going on. Then someone at the obsevee's table says, "Hey, why's this person over there watching us for? Maybe it's a psychology experiment. Let's screw it up!" And, so they start to drink much faster than they normally would. Or, they stop drinking altogether. Now, this might be good if they are stopping all together. However, it doesn't give you very accurate information. So from the standpoint of a good experimental design, this has some weaknesses. So what about the next technique?

The next technique as we see in slide six is the case study. The case study is basically is a major investigation of one particular unit. The unit can be anything. It can be an individual, a university, a town, a city, etc. Generally it is a detailed study of one particular individual or one particular thing. Let me give you an example. Let's say you are a physician and you are in your own little medical office and you are doing fine and all at once somebody walks in with pink, yellow, and green polka dots all over their body. And they go, "Doctor...Doctor! I have pink, green, and yellow polka dots all over my body!" What is the first thing you're going to think of if you are a physician? "I'm going to be famous!"

So the first thing that you do is the standard things. You look in their eyes, ears, and all the other stuff. But then, since you can't find anything wrong, what you begin to do is take a detailed history of this individual. You examine where they have been, what they've done, what they've drank, everything you can think about. Ultimately, after you find out all about this, you take all the information and you write a report. Then, you submit it to a journal and you call it "Steve's Disease."

So, we have Steve's Disease, and Steve's Disease is out there in a major journal somewhere. At a different time and location, another physician comes along, another person walks into their office with have pink, green, and yellow polka dots all over their body. They go, "Doctor...Doctor! I have pink, green, and yellow Polk dots all over my body!" And the doctor goes, "Hey, I remember this, I saw it in a journal a little while ago." And so, what they begin to do is look back into the journal, and there it is, Steve's Disease!

This physician does the same thing that the first physician does; They take a major detailed history, they do a lot of blood work, and on, and on, and on...! They write it up and put it in a journal. They also send it to the CDC, because now you have two people with this disease. As we continue on a variety of other people begin to come into different clinics with Steve's Disease. Each of them starts to show the same symptoms.

Over time the Center for Disease Control begins to start saying, "Hey, there's something going on out there with Steve's Disease. We need to do some more investigation of this. A similar thing could occur with a university researcher. So, what they begin to do is look at all the commonalities of Steve's Disease. What they find is that a person who lives in one particular area or who lived in one particular area and engaged in one particular behavior ends up with Steve's Disease while other people who did not do the behavior do not.

These are the classic techniques that one finds when one is first trying to identify some kind of new disease out there. The classic example of newer diseases where this procedure was followed is HIV and SARS, and other types of diseases. So, what is the advantage of the case study?

As we can see in slide seven, you can develop a lot of understanding from examining a lot of these experimental findings. In addition to that, this is a technique we usually use when we have no idea about what is going on.

So, what are some disadvantages? As we can see in slide eight, it can be very inaccurate if done poorly. Often times, you will also get biases from the individual actually doing the case study. Everyone has some particular bias no matter who they are. Finally, you can also get political pressure to achieve some particular result or draw some particular conclusion. Consequently, it has some major weaknesses in that regard.

So, what is the next major technique? As we can see in slide nine, this is the survey method. The survey method basically involves giving questionnaires or interviews to measure something in a population. You can measure all sorts of different things. You can measure attitudes, you can measure behaviors, or you can measure opinions. You have all probably taken a survey of some kind or another. It asks you about how often you have engaged in X? Who did you vote for? Or, what did you think about this or that?

The advantage of surveys ( which we can see in slide ten) is that you can get a lot of information about a wide variety of things. You can also get very sensitive information provided you maintain strict confidentiality procedures and other types of things. This is a common technique we use in the Social Sciences to get information about high risk sexual behavior, and all sorts of other things.

What are some disadvantages of surveys? As you can see in slide eleven, the first thing we must make sure of is that the sample must be representative of the population. Let's say that I am going to do a political survey yet all the people I contact are only Democrats or only Republicans. As a result, I make some kind of conclusion. But, As we know, this conclusion can be wrong. Thus, you must make sure that you look at all sorts of demographic groups or some other types of target audiences to make sure they actually represent the population you will be discussing and drawing your conclusions. Time of day, who is working, age, sex, race, and many other variables all can have a major impact in the results of your particular survey.

There are also other problems as well. These are listed on slide twelve. The big thing about surveys today (and this really has a lot of researchers very concerned) is that people lie. That is, they lie about their results. The questions can also be biased. You can load a question any way you want and you can get an answer any way you want. A lot of unscrupulous survey companies actually do that. Probably, the most concerning aspect of this is direct marketing. Basically what these people are doing is that they are marketing some kind of product in the guise of some kind of particular survey. As a result of that, people are becoming very leery about providing information on surveys. In addition, you have identity theft worries, and other things out there that people are concerned about today.

What is the next type of method that we use? This method is shown in slide thirteen and is called the experimental method. The experimental method is the most commonly used method by all the sciences. What it does is evaluate variables. A variable is anything that varies over time. For us, in this class, there are two types that are important.

The first of these is shown on slide fourteen and is called the independent variable. The independent variable is the variable that is actually manipulated by the experimenter. It's the amount of money that you are given, or the volume of noise you are presented with, or anything else that you can vary over time. What we do is manipulate this independent variable we see what happens to the next variable called the dependent variable. (Shown in slide fifteen). The dependent variable is actually the variable you observe changes in. It would be things, such as, your stress level, your heart rate, or whatever it might be. Let's give you a couple of examples of how this works.

Let's use the example of noise and heart rate which we have shown on slide sixteen. What I am going to do is vary three levels of noise. Low levels, medium levels, and high levels. I want to see what these levels of noise do to your heart rate. Does it stay the same? Does it increase? Or, does it decrease? So, at low levels, what I have observed is that your heart rate stays at eighty. At medium levels it basically goes up to ninety. And, at high levels it goes up to one hundred twenty. Basically, we can conclude that if you are holding everything else constant, by changing the volume of noise you are actually manipulating the heart rate.

As we can see here on slide seventeen, the thing that you vary is the noise and that is your independent variable. The thing that you observe changes in is your dependent variable.

What are some of the advantages to the experimental method? As we can see in slide eighteen, the first thing you can do is reach precise conclusions. That is, X causes Y to occur. That means when you are doing an experiment you have to hold lots of other extraneous things out there constant that could influence the outcome of a particular study. An example might be the time of day. Another example might be the position of the sun. Or, the persons age. All these things can have influences, even your blood sugar level.

So, what are some disadvantages of the experimental method? As we can see in slide nineteen, the first major problem we might have are ethical issues. That is, we can do experiments on people with all sorts of different things, such as, radiation, etc., but they are not very ethical to do. The next thing that is a problem with some experimental technologies is that it can be artificial. The lab is not the real world. It can be very, very close but it is not the real world. And finally, there are just some things we cannot do experiments on very well. Such as, attitudes, opinions, and assorted things like that.

Well, what's the next major method that we have? As we can see in slide twenty that is the correlational method. The correlational method is basically a way to estimate the extent that two variables are related to each another. However, just because two variables are related does not mean that one causes the other.

There is a statement that I want you to memorize. If you do not get anything else out of this class you at least need to remember this... As you can see in slide 21 “Correlation does not imply causation." Repeat, por favor, "Correlation does not imply causation."

What are some examples of correlations? As we can see in the first one on slide twenty-two, the rooster crowing causes the sun to come up. This is a classic example of a correlation. So, the rooster crows in the morning, "cock-a-doodle-doo", and the sun comes up. The rooster crows again the next morning, "cock-a-doodle-doo", and the sun comes up. It's a perfect relationship. The rooster crowing and the sun comes up. Thus, it is perfectly obvious to anyone that the rooster crowing must cause the sun to come up.

Well, as we know, that is wrong. You go and chop off the roosters head and the sun still comes up. It is a perfect example of a correlation but it does not mean that one causes the other. Next example, ice cream and drowning.

As ice cream consumption goes up drownings go up. As ice cream consumption goes down, drownings go down. So, it is perfectly obvious to anyone out there that ice cream causes drownings to occur. As you know, there is a problem with that. Because, the time you eat a lot of ice cream is in the summer. When do people usually drown? In the summer! Again, we have another variable out there that is actually controlling or having an impact in this particular type of example. Again, it's a perfect relationship but it does not mean that one causes the other.

Next, slide twenty-four. Your genetics cause alcoholism and other disorders. It is perfectly obvious that, as you see in families, many people who are alcoholics or drug addicts also had parents or siblings who were alcoholics or drug addicts. So, many, many, individuals have basically said that your genetics causes alcoholism to occur. That may be true, but basically the methodology they are using is a pure correlational technology. It is not an experiment. As a result of that you have to make sure you understand that there are weaknesses with that particular argument. It does not mean that it is not true. It just means that the material and technique you are using to draw that conclusion is a correlational method. So, you have to be very cautious when you make statements like that.

Let's take a look at some other examples. The classic example is shown in slide twenty-five. Birth control pills cause breast cancer. Again, we have individuals who take birth control pills and individuals who develop breast cancer. Does that mean that birth control pills cause cancer? Again, the technique they are using is a correlational method. If one uses an experimental approach where you are holding other things constant, then one can make that argument. However, at this time we have to remember that these are correlational modals. You see many, many, many other types of examples out there in your local newspaper; something causes Y. Something causes this. Something causes that. The question you have to ask yourself is, "What kind of methodology did they use?" Did they use an experiment; did they use a correlation, etc? Again, you need to be careful when you are evaluating material and the conclusions they are drawing. So, question the types of models and the types of designs people are using when they draw those conclusions.

Now, correlations are expressed as a number. As we can see in slide twenty-six, the numbers range between +1 and -1. The closer the number is to zero the less relationship there is. Or, the closer the number is to + 1 or -1, the more relationship there is.

As we can see in slide 27, a 0.9 has a greater relationship than a 0.4; or a 0.3 has less of a relationship than a 0.8.

So, again, as we can see in slide 28, the number only tells you how much the variables are related.

As we see in slide 29, what does the plus or minus sign tell you? s

This is shown in slide 30. Basically, in a positive correlation (that is, a +1 to 0); one variable increases another variable increases.

A classic example, as we see in slide 31, and in the previous slide 30, is beer drinking and the probability of puking. As shown in slide 31, the more you drink the higher the probability that you are going to puke.

What about a negative correlation? What can it do? Well, as you can see in slide 32, in a negative correlation, as one variable increases another variable decreases.

A classic example would be that as beer consumption increases your driving skill decreases. This is shown in slide 33.

What are some advantages to the correlational method? Well, as we can see in slide 34, you can use them with lots of different variables in different situations. They are also more precise than case studies or observational methods. And, you can study them or study variables that you can't do experiments on. However, as we can see in slide 35, there are some disadvantages as well. The most important one is that you cannot draw cause/effect relationships. Remember, correlational methodologies do not account for other variables which may actually control the behavior, and several of these I have shown you. Remember, as we see in slide 36, correlations do not imply causation.

Well, that concludes the section on research methods. Again, it is designed to provide you a little bit of an overview of the different methodologies that we use in psychology. It is designed to give you some general information you can use to evaluate behavior. Make sure you keep those things in mind before you start to draw precise conclusions.

In our next section we are going to start to talk about epidemiology and looking at some different variables statistics that we actually have out there of people with substance use and abuse. So until then, we hope you enjoy your day and we look forward to talking with you again soon.