production date 2/5/00

Introduction to Behavioral Science Research

The Scientific Method and The Need for Data Analysis

Table of Contents	Objectives
Why Study Statistics	Do all students need this?
Charles Peirce	Learn the 4 methods of knowing.
Basic Aim of Science	Learn the Aim of Science.
Definition of Scientific Research	Learn the two components of Scientific Research.
Computer Projects	Do your first Computer Project
Additional Information	Discover interesting Web Links
Questions/Test	Take the End of Chapter Test.
Report	Send a Chapter Report to your Instructor.

Why Study Statistics?

This chapter explains why the study of statistics is so important beyond the author's opinion discussed in the Navigation section. The line of reasoning begins by stating that: there are several ways that humans come to believe and/or understand certain truths. In other words, there are several methods of knowing. In this chapter, we organize those methods of knowing around a simple classification system developed by the American philosopher, Charles Peirce. The best of these methods of knowing is the method of science. The scientific method uses research and mathematics to support the "truths" it reports. Statistical analysis or statistical inference is the primary mathematical tool used in scientific research. Without a grounding in statistics, understanding the results of scientific investigations and discovering truth would be nearly impossible. Beliefs and prejudices would remain, but good methods of testing ideas would be lacking. By taking this course you are beginning to build a basic analysis tool that will assist you in many areas. You will learn the skills necessary to collect information and ask the data, through statistical analysis, to reveal the truth concerning your questions.

Social science is not immune to ideas being supported through unfounded beliefs and prejudice. When two opposing ideas are believed by different groups, the groups often become radicalized. They argue with one another, trading verbal barbs and insults. At some point, the scientific method, and statistical analysis of relevant data will be asked to clear up the confusion created by the two.

As an example of a current radicalized position, reading teachers are today exposed to, and trained in, two very different methods of teaching reading to children. These two reading methods are the Whole Language Approach and the Phonetics Approach also known as Direct Instruction. The essential difference between the Whole Language Approach and Direct Instruction lies in what is believed about how children learn. Often the public fights waged by believers from each of the reading approaches, make both camps look ridiculous and/or mean spirited. In California, first Direct Instruction was used, then Whole Language Instruction was mandated, and recently, Direct Instruction with phonics was again mandated, replacing Whole Language. This short QuickTime movie, with sections taken from NBC television, illustrates the problem. A 20/20 television program that aired in 1995 also pointed out educational changes from Whole Language to Phonics. The Whole Language Umbrella group was so distressed with what they perceived an an unfair presentation on the 20/20 program that they produced a very pointed rebuttal. As you watch the movie clip above, or read the rebuttal, think about how data analysis is being used to substantiate the position being taken. Most of the time in arguments such as these, opinions are being expressed without the support of data. If a research study is used, it is often not addressing the points made by people on the other side of the argument. Of course, statistical analysis of human learning is the key to unraveling the differences in these two camps, allowing us to more efficiently teach our children to read. There would need to be agreement on the design of the experiment, but the results from one or more experiments could help these two sides understand the truth of how one best teaches children to read.

Many other researchable examples could have been used to illustrate the importance of the skills you will begin to learn in this environment. Milgram's Obedience to Authority study was presented earlier as an example of a important experiment.) Of all the skills you will, or have had, an opportunity to learn, your ability to statistically analyze data will be at or near the top of any list of useful skills you create. Skill in conducting and understanding statistical analysis is important (probably critical) to your future in any professional field. For example, if you are planning to become an educator, you will be called upon to make critical decisions about how well students perform on tests, the effectiveness of learning materials, how students are progressing in their studies, etc. If you become a psychologist, your practice may involve deciding if a treatment plan is working for one or a group of patients. Your choice of the tests that you use in your assessment battery will be guided by the test's statistics. As a social worker you need to understand probabilities, and data concerning the eligibility of a client for a program. An environmental researcher may wonder if there is a relationship between the concentrations of toxins and deformities in certain animals. Finishing a degree in Leisure Studies, you will know that sound knowledge based on research is critical for the professional management of leisure services. Research based information can help solve leisure problems. It can also lead to improved service practice. All of these decisions in the social sciences are guided by the statistical understanding you will gain in this course.

Statistics is a tool, like a scalpel. You will need to practice your use of statistics and develop your analysis skill, like a surgeon needs to practice using the scalpel developing surgical skill. Using a scalpel, a skilled surgeon can find, and often remove or repair basic plumbing problems in the body. A skilled social scientist using statistics can discover the truth. Statistics is the core tool for one of the methods of knowing proposed by Charles Peirce.

Charles Peirce

The famous American philosopher Charles Peirce stated that all humans arrive at what they believe through four different methods. He termed these the Four Methods of Knowing. They are:

Tenacity
A Priori
Authority
Science.

Tenacity

The method of tenacity says that humans hold firmly to the truth, and know it is the truth because they hold firmly to it. The more often they repeat something (even though it may not be truthful) the more truthful it becomes. This is not a valuable method of discovering the truth, but humankind uses it more often than we like to admit. Indeed, many of our worst beliefs were derived and held through the method of tenacity.

When I was a student at the University of Nebraska, working as a mental health technician at the State mental hospital, many of the "truths" held by our patients appeared to be held through the method of tenacity. One of the most memorable patients believed that the environment was full of hurtful radiation (probably not a entirely incorrect belief). He protected himself by wrapping his body in tinfoil. His belief that tinfoil could protect him from these dangerous rays was his own. As a mental health worker, I was supposed to convince him that his behavior was inappropriate. But the more I told him that his belief was ridiculous, the more he stated that tinfoil protected him. The more he stated his belief, the stronger it became. It was more than a bit humorous, and at the same time quite sad, to see him hoard tinfoil from any source. Alone in his room he would fashion the tinfoil into protective bands that he wrapped around his body. In my final days working at the hospital, he had enough tinfoil wrapped around his body and hidden under his clothes, that when undressed only the skin on his feet, hands, and face was exposed. Otherwise, he resembled the Tin Man from The Wizard of Oz.

A Priori

According to the a priori method of knowing, through free communication and rational thought, humans can arrive at the truth. Things are true because they are self evident. It just stands to reason that they are true. One example of a belief held by the a priori method is that women can stand more pain than men. The line of thought goes like this: (1) Women go through childbirth. (2) That process looks quite painful to men. (3) Even a macho man knows he couldn't take that much pain. (4) Therefore, women must be able to stand more pain than men.

Certainly communication through discussion is a valuable tool. However, the truths we have learned by exclusively using this method are relatively few. In the area of psychology and intellectual assessment there is a famous example. On the Wechsler Intelligence Scale for Children, this question is asked. "What is the thing to do if a boy (for girls use girl) much smaller than yourself starts to fight with you?" Many psychologists in thinking about this item decided that it was probably biased against minority children. The reasoning went like this. Since many minority children live in lower income neighborhoods, and in these neighborhoods children are exposed to more violence, and because of the violence in their lives, minority children will fight back because they need to do so to survive. Therefore this question is unfair to them. They will answer incorrectly that they should in some way fight back. One of the national associations of psychologists was so convinced that this item was unfair to minority children that, without data, they proclaimed that minority children were unfairly penalized by this item -- forcing down their scores on the intelligence test. However, after data was repeatedly collected concerning scores on this item, it was shown that there was no difference in the answers given by children whether they lived in poverty, or affluence. There was no difference in answering the question by Black, White, Asian or American Indian children. A truth that was thought to be quite self evident was actually false.

Authority

The method of authority is the method of established belief. If an authority says something is so, then it must be. You are using this method right now. You may be taking notes from this text, or printing it and using your highlighter to mark important parts. You believe that what is written here must be the truth. Perhaps you have read something that is interesting enough to include in an e-mail message to one of the listservs, or to another student in class.
How much do you trust the following statement? There must really be a Charles Peirce, and he must have said these things. Besides this text, what other information do you have that Peirce existed?

Using the method of authority as a belief system is as good as the authority. Often knowledge and beliefs from authorities have the weight of tradition and public sanction behind them.

Sometimes even the best authorities tell us untruths that we believe. In Lippincott's Silent Reader for the second grade, written by my dearest aunt and published in 1923, Aunt Ethel stated that bananas with brown spots may be eaten raw, but if they don't have brown spots they should be baked or boiled before they are eaten. Of course, as she told me in 1991, everyone knew green bananas were poisonous. As another example, at one time the church held and taught that the earth was the center of the universe and unmoving. Most of us know that because Galileo published a book in 1632 that discussed Copernicus's views that the earth rotated and revolved around the unmovable sun, that he was placed under strict house arrest. He remained in his villa, unable to publish anything, until his death in 1642. As Jacob Bronowski (1973) noted in his textbook The Ascent of Man, Sir Isaac Newton was born in England on Christmas day of that same year. While Galileo discovered the scientific method, it was Newton and other scientists who followed him that have made the scientific method the best for discovering the truth .

Science

The scientific method of knowing states that through research and empirical science we will come to know the truth. Science believes that nothing is obviously truthful. We must support our beliefs through observation, measurement, and experiment. This is what is called empiricism. Both proper research methods (research design) and numbers and their manipulation (statistics) are involved. This course and learning environment are concerned with the latter. You will learn how statistics (a core skill in the scientific method) enables you to make decisions that lead to truth.

To understand science more fully, it can be contrasted with the other three methods of knowing. The three methods (tenacity, a priori, and authority) can be combined together and referred to as the common sense methods. In the following paragraphs five major ways that science differs from common sense methods of knowing are discussed.

(1) Use of Conceptual Schemes

Science uses conceptual schemes and theoretical structures in a very systematic way. Common sense methods use theories in a very loosely structured manner. For example, one belief that was held by common sense was that illness was caused by sin. Humans paid for their sins by being struck down with disease. Science, of course, discovered that illnesses are caused by germs and viruses. In scientific thought the virus is identified, its impact on the body is studied, and what treatments are useful or useless are investigated. Scientific investigations of this nature are so valuable, that when studying the effects of a potential drug treatment, researchers will actually withhold a potentially life saving drug from a subject. Society believes that the knowledge gained through this type of experimentation is valuable enough that some subjects are given placebos instead of the drug. Of course, as soon as the efficacy of the drug is known, all subjects will be provided the treatment if it is useful.

This HIV QuickTime movie was produced by a Silicon Graphics Molecular Modeling program. Using data obtained about the structure of the Human Immunodeficiency Virus (HIV) virus, this computer software package built a molecular representation of the virus that causes AIDS. The movie simply rotates this model is space so that researchers have a better visual representation of the virus. You can click and drag the movie controller at the bottom of the movie to interactively rotate the virus molecule. The ability to visualize what is being studied greatly assists in understanding what is being studied. Chapter 3 introduces statistical visualization tools. If we were still pre scientific just imagine the treatment that current HIV sufferers would be subjected to by others.

(2) Use of Empirical Tests

Scientists systematically and empirically test ideas about the truth. These ideas, depending upon the detail of their development, are called theories and/or hypotheses. Common sense notions are often believed because they agree with a previously held position. The point is, that science demands one or more well constructed tests to hold that something is true. Common sense notions can often continue to be supported by believers even in the face of contradictory evidence. In psychology, the notion that people reveal their unconscious thoughts when asked to tell what they see while looking at an ink blot has great appeal. However, scientific evidence is lacking to support the notion that people project their beliefs, fears and pathologies into the stories they tell when shown inkblots. This lack of empirical support has not changed the fact that the Rorschach is one of the most commonly given and interpreted (or misinterpreted) personality tests. For the Rorschach to be scientifically supported there would need to be research supported by a statistical analysis of the data indicating that certain stories associated with certain inkblots indicated certain personality traits.

(3) Use of Scientific Control

Scientists control variables in studies which may change the results of their investigations if these variables are not controlled. Lay persons using common sense methods of knowing seldom control for these extraneous variables. We may believe that minority children have a better chance of coming into contact with the legal system than majority children. Why that happens (what causes this and what the truth is) depends on whether or not one controls those extraneous variables. Common sense notions may lead one to believe that race determines whether or not one respects the law. Obviously, one could only hold this belief as true if the extraneous variables of environment, family income, and other extraneous social variables were not controlled. The scientific method would demand that these variables be controlled. The scientific method would lead to the conclusion that economic and/or educational disadvantage plus a feeling of alienation and not skin color are precursors of delinquent behavior.

(4) Use of Consistency or Repeatability

Scientists look for relationships which are consistent over time. A scientist must replicate his or her study. Just seeing something happen in a particular way once doesn't mean it always happens that way. Lay persons using common sense methods often link together two events which are related in only a single instance. This practice can certainly lead to "truths" which are incorrect. Several years ago, I heard of this referred to as the Sauce Béarnaise Syndrome. The explanation went like this. A rural family had traveled to the big city to see the sights. At lunch they decided to eat at this fancy-smancy restaurant. The mother and child had a hamburger and fries, while the father wanting to show his sophistication, had a sandwich he had never heard of before. When he ordered he wasn't even sure of how you pronounced it's name. His platter came to him with the sandwich covered with a goop which the waitress said was béarnaise sauce. Unbeknownst to the father, he was also getting the flu. He ate the new sandwich (secretly wishing he had ordered a hamburger) and the next day had a fever and cough. He linked the two events (sandwich + sickness) and immediately knew he had food poisoning - thus the Sauce Béarnaise Syndrome. Of course, a good scientist would eat at that restaurant several times and only conclude he was getting food poisoning if he got sick each time. Joking aside, scientific papers are written in enough detail so that others can conduct either similar or identical studies. To believe in our conclusions, we must see others obtain the same results under the same conditions.

(5) Use of Metaphysical Explanations

In science, explanations which can't be tested are automatically ruled out. Untestable explanations are called metaphysical explanations. (Note: Stanford University has a Metaphysics lab that you might find interesting to visit on the web.) Science is concerned with things which can be publicly observed and tested. While an oversimplification, this difference is dramatically pointed out by contrasting behavioristic and psychoanalytic schools of thought. A behaviorist measures everything. It is easy to count when a specific behavior has been rewarded (positively reinforced). On the other hand, psychoanalytic theories were driven by conflicts between the Super Ego, the Ego, and the Id. All of these concepts were difficult to measure. Do you know how much ego strength you have? Also, explanations like It was God's will are not scientifically acceptable as a cause of an event.

If something can not be observed, then it can not be measured. If it can not be measured, then it can not be given a score. If it can not be given a score, then it can not be statistically analyzed. If it can not be statistically analyzed, then it is not science.

Summary of Science vs. Common Sense

Does it have	Science	Common Sense
Systematic definition of theory	yes	no
Systematic test of theory	yes	no
Control outside variables	yes	no
Metaphysical explanations ruled out	yes	no

One chooses science over common sense because in the long run, science has led to truths and corrected much of what was held in error as the truth by common sense methods of knowing. Scientific thought is self correcting and common sense is not. With science you have to see the same results repeatedly and from differing researchers. Most scientific writing is subject to peer review. Having other scholars in your area review what you have done before it is ever published helps keep untruths out of our scientific knowledge base.

Basic Aim of Science

The basic aim of science is the creation of theory. A theory may be defined as a set of interrelated ideas that present a systematic view of why things happen by specifying certain relations among variables. For example in the social sciences researchers who believe in behavior learning theory (often simply called behavior modification) state that if a reward follows a behavior, the frequency of that behavior will increase. The corollary of this statement is that if a punishment follows a behavior, the frequency of that behavior will decrease. These are two interrelated ideas in behavior learning theory about why the frequencies of certain behaviors change. The purpose of scientific inquiry is to explain or predict phenomena. In this example, the scientist would like to explain why or predict whether certain behaviors will occur and/or change the frequency of their occurrence. To study science or create a theory which you hold to be true using the method of science, you must conduct scientific research.

Definition of Scientific Research

Scientific research is a systematic, controlled, empirical and critical investigation of ideas (hypothetical propositions) about the presumed relations among phenomena. While you shouldn't memorize this formal definition, it does point out two important components necessary for you to learn in order to conduct scientific research.

The first component states that scientific research is systematic and controlled. What is being talked about here is how one designs research. Research design answers questions concerning what possible extraneous variables might effect research studies and how they might be controlled. To gain this knowledge you must take one or more classes in research design.

The second component is the fact that scientific research is empirical. This is the focus of this course. Chapters 3 through 15 will teach you how to statistically treat data in order to discover the truth. Statistics is concerned with the empirical analysis of data in order for you to draw conclusions about the relations among phenomena. As noted in the beginning of this chapter, statistics is as important to scientific thought as the scalpel is to surgery. You will begin to learn about its use in this material.

A scientist who was conducting scientific research on the statement above, that a behavior followed by a reward would increase in frequency, would need to design an experiment where a behavior was defined, a reward was provided after the behavior was displayed, and the frequency of the behavior could be counted over time. The design aspects of this experiment would involve the definition of the behavior, perhaps if there were multiple observers, training each of them to recognize when the behavior was displayed with equal skill, making sure that the reward was actually a reward for the multiple subjects involved in the experiment, etc.. The statistical component would include the counting and collection of both the time measure, and the behavior frequency counts. The statistician would then need to know how to appropriately use statistical techniques to analyze the collected data.

The goal of this material is to start you on your way to becoming a scientific researcher. The material in this environment concentrates on statistics and not experimental design.

Computer Projects

Becoming familiar with Statlets

Project 1

This is a simple project where all you are required to do is read the introduction to the Statlet manual, and, if asked, alert your instructor that you have done so. Read all the directions for completing this first project before attempting to do so.

(1) click the Statlet button in the Navigation bar. The initial Statlet's page will appear. It is in this page that you can request different versions of Statlets (Java 1.0 or Java 1.1) and different forms (Menu or Standalone Applets). To understand what this is all about, you must read the Introduction section in the user manual.
(2) To accomplish this, click the User Manual link in the upper left corner of the introductory page. A new browser page will appear with the following links.
Table of Contents
Introduction
Chapter 1: File - Creating, Opening, and Saving Data Files
Chapter 2: Edit - System Preferences, Data Editing
Chapter 3: Plot - Scatterplots, Business Charts, Tables, Prob. Distributions
Chapter 4: Summarize - Statistics, Tabulation, Crosstabulation
Chapter 5: Analyze - One Sample, Two Samples, Multiple Samples, Sample Size
Chapter 6: Model - ANOVA, Regression, Time Series
Chapter 7: Quality - Pareto Analysis, Capability Analysis, Control Charts
Chapter 8: Window - Manipulating Windows
Chapter 9: Help - Online Help
Glossary
Index
(3) Click the Introduction link, and read the introductory material. You will be tested on this information in the online quiz given below.
(4) After you have read the introductory material, close the extra browser window by clicking in the close box -- returning to this page.
(5) If asked, report your level of understanding of this material by using this submission form available if you have an active internet connection.

Additional Information

As you have probably noticed, this material is referred to as a learning environment. It is far more than a statistics textbook. You always have available many tools and sources of information to assist your learning. Near the end of each chapter, you will find links to web resources for important topics, issues or people mentioned in the chapter. You will also find links to forms that allow you to search the internet for those same (or other) topics. These forms will introduce you to the major search engines on the internet.

In this chapter, you can find more information concerning Charles Peirce, Sir Issac Newton and Jacob Bronowski.

Try a search using AltaVista using this form. AltaVista provides a very helpful index of Web sites and newsgroups. If you have an active link to the Web, the form will actively search AltaVista for web sites or newsgroups using any search term. Select whether you want to search the web, or news groups (Usenet). Type your search term (i.e., psychology) in the space provided and click the Submit button. To return to this page, simply close the additional opened browser window.

Questions/Test

This link allows you to take a computer scored end-of-chapter test.

Report

Please send a report indicating your understanding of this chapter to your instructor. You will need to know both your and your instructor's e-mail addresses.