DEPARTMENT OF POLITICAL SCIENCE

POSC/UAPP 815

PUBLIC MANAGEMENT STATISTICS

APPLIED STATISTICS FOR SOCIAL, POLITICAL, AND POLICY SCIENCES

Fall 1997

H. T. Reynolds

Course Overview and Philosophy* Course method* Readings and software* Requirements*
E-mail* Desktop computers* Suggestions and attendance* Important Numbers * List of topics
Do you have a question?

Statistical thinking will be as necessary for efficient citizenship as the ability to read and write.

-H. G. Wells

Thou shalt not sit with statisticians nor commit a social science.

-W. H. Auden

Nothing in education is so astonishing as the amount of ignorance it accumulates in the form of inert facts.

- Henry Brooks Adams

When you can measure what you are speaking about, and express it in numbers, you know something about it; but...when you cannot express it in numbers, your knowledge is of a meager and unsatisfactory kind.

-Lord Kelvin

Get your facts first and then you can distort 'em as much as you please.

-Mark Twain

There are three kinds of lies--lies, damned lies, and statistics.

-Benjamin Disraeli

OVERVIEW

The following quote from the New York Times raises an important point:⁽¹⁾

A year ago [1996], after a furious debate within his own party and among his own advisors, President Clinton signed the most radical revision of the nation's welfare policy in history. Last week the President declared the change and unqualified success.

"I think it's fair to say the debate is over," Mr. Clinton said., "We know that welfare reform works."

His evidence? A 1.4 million-person drop in welfare rolls over the past year.

As the article points out, a single figure such as the size of welfare enrollments only shows that the number of people receiving public aid has fallen, not that welfare reform works. After all, the decline could have resulted mostly from improved economic conditions. What is significant, however, is that Clinton's argument represents just one of countless examples of the use of data and statistics not merely to understand social problems but to advance particular ideas, policies, causes, and even ideologies. Consequently, however one views the topic, statistics--the analysis, presentation, and interpretation of data--dominates the way the nation makes its public policies

Naturally social and policy scientists rely on statistical analysis to make their points. But these techniques have seeped into popular culture as well. Every newspaper carries stories containing assertions purportedly resting on "hard" facts: "Evidence proves that capital punishment deters crime." "Data show that CFCs deplete the ozone layer." "Scientific studies reveal that the American economy is becoming less competitive." "Blacks disproportionately receive death sentences for capital crimes." "Laboratory experiments point to the harmful effects of television violence [or pornography or tobacco advertisements], which should therefore be outlawed." "Census reports imply that..

One can go on in this vein indefinitely, but the point is clear: everyone--common citizens along with social scientists and policy analysts--are constantly bombarded by claims and counter-claims, all seemingly backed by hard numerical evidence, about public policy and current events. These arguments have political, moral, economic, social, and practical dimensions. But what often unites them is their reliance on statistics of one sort or another.

In light of these considerations this course has four broad objectives:

First, it provides a one-semester introduction to statistical reasoning and concepts. I hope, in other words, to present enough information to enable everyone to evaluate intelligently assertions that rest on statistical methods and inferences. By the end of the semester, for example, you should know enough to analyze critically a claim such as "There are positive correlations between the generosity of welfare benefits, undesirable social behavior, and dependency on public assistance."
Second, you will learn to use several statistical methods to summarize and interpret some interesting data. My goal is to encourage you to think about what (if any thing) a batch of numbers says about the "real" world.

Third, you should be able to locate, retrieve, and analyze data from a variety of different public and private sources.

Finally, the course offers a foundation for those who want to pursue the subject in more depth. Most disciplines--sociology, political science, law, psychology, public administration, and economics--expect their practionners to use at least elementary statistics in their research. This level of expertise normally takes two or three semester to acquire, and 815 tries to lay a solid foundation for more advanced study.

Needless to say, not everyone can become a statistician. That's neither a realistic nor desirable goal. But it is possible to learn enough basic concepts and methods that will allow one to ask intelligent questions and to know what constitutes reasonable responses.

Go to top

METHOD:

Like many academic subjects statistics has to be learned "actively" by describing and analyzing data. Hence, as noted below, "assignments" or homework accompany nearly every class. Intended to clarify various ideas discussed in class these assignments should be completed carefully and promptly. (The due dates will always be clear.)

Go to top

WHAT YOU NEED:

Depending on your interests and budget pick one (or a combination) of the following. I highly recommend both the textbook and the MINITAB software, but you can succeed without them.

- Minimal but satisfactory:

* Class notes and data.

* McKenzie, Schaefer, and Farber, The Student Edition of MINITAB for Windows to be used as documentation in University computing sites. (These places may have manuals that you can also use, but by far the best strategy is to buy the Student Edition, even if you don't use the software that comes with it.)

* A statistical calculator

* A subscription to the New York Times, available at the Newark Newsstand. Even if you are a "distance" learner you should subscribe to the Times.

- Intermediate: More expensive but much better

* Class notes and data.

* McKenzie, Schaefer, and Farber, The Student Edition of MINITAB for Windows for use with a home or office computer

* Quantitative Applications in the Social Sciences (Sage): Jacoby, "Statistical Graphics for Univariate and Bivariate Data" (No. 117); Henkel, "Tests of Significance" (No. 4); Schroeder, Sjoquist, and Stephan, "Understanding Regression Analysis" (No. 57); and Hagle, "Basic Math for Social Scientists" (No. 108).

* A statistical calculator

* A subscription to the New York Times, available at the Newark Newsstand. Even if you are a "distance" learner you should take advantage of the Newsstand's many, many excellent services.

-Advanced: most expensive but the best choice by far.

* Class notes and data.

* McKenzie, Schaefer, and Farber, The Student Edition of MINITAB for Windows for use with a home or office computer.

* Agresti and Finlay, Statistical Methods for the Social Sciences 3^rd Edition.

* Quantitative Applications in the Social Sciences: Jacoby, "Statistical Graphics for Univariate and Bivariate Data" (No. 117); Henkel, "Tests of Significance" (No. 4); Schroeder, Sjoquist, and Stephan, "Understanding Regression Analysis" (No. 57); and Hagle, "Basic Math for Social Scientists" (No. 108).

* A statistical calculator

* A subscription to the New York Times, available at the Newark Newsstand. A student subscription only adds about $15 to the cost of books.

It seems to me that investment in a first-rate text such as Statistical Methods for the Social Sciences makes sense even for those people who do not intended to "do" statistics during their careers. After all, for the rest or your lives most of you will be discussing, if not formally evaluating, arguments that rest on at least some statistical methods. Having a good reference book ought to help greatly.

Go to top

REQUIREMENTS:

Since I believe that statistics has to be learned interactively by doing, not observing, I want you to put your knowledge to work. Thus, the main course requirement consists of daily or weekly assignments. During each class (more or less) I will hand out a set of "problems" to "solve." These assignments, most of which involve analyzing various types of data, should help you understand the concepts and methods described during the class. Some, but not all, will be graded. Those that are not formally graded are "scored" as follows:

* + = satisfactory

* 0 = acceptable but you should look at my comments to make sure that you fully understand.

* X = Resubmit. Something has been done or expressed incorrectly. Correct the indicated errors and resubmit. Again the point is to make certain that you understand the material.

Since clear thinking and orderly work go hand-in-hand, I want you to get organized, write notes and intermediate calculations on scratch paper, and turn in neatly written sheets. Attach only the graphs, figures, or tables explicitly requested. Presenting reams of computer "print out" doesn't help anyone.

You can't expect to do well in the course if you don't conscientiously do the assignments on time. Missing assignments or a couple of "Xs" will lead to trouble.

Focus Students Note: assignments should be sent or taken to the Political Science office, 347 Smith Hall, University of Delaware, Newark DE 19716. Include a self-addressed, but not stamped, envelope so I can return them.

Go to top

E-MAIL AND THE INTERNET:

Whether as an "on campus" student or a "distant learner" (participant in FOCUS) you will be extensively using electronic mail and the world wide web. You thus have to have access to a desktop computer. Fortunately, numerous choices abound. For example, both Delaware's main and peripheral campuses house public "computing sites," places where students can find computers and (sometimes) consultants. Most of these machines are connected to the internet, which permits you to send and receive "e-mail" and access the course "web page." Moreover, the price of these devices has fallen in recent years to the point where even very sophisticated machines are within the budget of many students. If at all possible, I suggest that you either buy one or try to arrange for regular access to a friend or coworker's. But in case you can't there are plenty of machines on the various campuses, especially in Newark, that you can use for free.

If you have any choice, try to use "windows" (IBM or IBM clones) machines because I will not discuss Apple systems at all.

Go to top

DESKTOP COMPUTING AND STATISTICAL PROGRAMS:

One of the advantages of computing is that they take the drudgery out of statistical calculations. Since I want everyone to analyze as much data as possible and to concentrate on concepts, not calculating formulas, we will be spending a great deal of time using statistical program "packages," computer software that bundles innumerable procedures into a unified system.

Although dozens and dozens packages exit, most of them have the same features and work in the roughly the same way. So if you learn how to use one, you can easily transfer your knowledge to others. Still, in order to provide some unity and continuity to the course I rely mostly on MINITAB and to a lesser degree SPSS. (As just noted, once you learn one you should be able to use the other with little or no problem. But for learning statistics MINITAB is in my judgment far superior. So I stress it above other program packages.)

Go to top

OFFICE AND GETTING HELP:

My office is in 434 Smith Hall where I will be on Tuesday and Thursday, approximately 2 to 4 pm. In addition, I am more than willing to meet at other times of mutual convenience.

An excellent way to stay in touch and ask questions is via E-mail. My address is htr@udel.edu.

If you are having trouble doing the assignments, please let me know immediately. When seeking assistance, though, please keep these points in mind:

- You should make an initial effort to do the problems. That way we can determine what you do and do not know.

- Bring or send printout (if applicable). During the first couple of weeks I will show you how you can mail and print copies of your computer calculations. Having this "diary" will help us pinpoint any mistakes you may be making. Do not try to reconstruct what you have done from memory; bring a "hard copy" of your work.

- Write down specific questions. Writing may help clarify the situation in your own mind.

- Do not be afraid to ask questions. Since the subject matter is cumulative, it is important that you understand everything as we go along.

ATTENDANCE:

Regular students note: Attendance is always required, even though some or all of the classes are being video taped. These tapes may be available for review purposes, but you are nonetheless expected to attend each and every session. If work or research obligations present a problem, you might wish to consider the Focus section of the course.

SUGGESTIONS:

This class does not assume any knowledge of statistics, computers, or research methods. (For some it may be too elementary, and they are encouraged to take a more challenging course.) What is important is a good attitude: the material can be useful to your professional development if you keep an open mind and make an honest effort to learn.

Experience tells me that the students who experience most difficulty learning statistics or computer programs such as MINITAB or SPSS are often somewhat disorganized and sloppy in their work habits. It is important to be neat and systematic. Read directions carefully before undertaking any assignment. As an example, do not sit down at a PC without having first read the assignment, gathered the materials you need, put them in the order in which they will be used, have blank scratch paper available, and list what it is you are supposed to do. Leave extraneous or unnecessary items at home. That way you will work faster with fewer errors.

Go to top

IMPORTANT NUMBERS AND ADDRESSES:

Office: 434 Smith Hall - Phone 302-831-1940

Political Science Department: 347 Smith Hall - Phone 302-831-1940

Fax: 302-831-4452