Introductory Quantitative Methods

This course introduces the graduate student to basic methods of empirical inquiry in the social sciences. The overwhelming majority of studies that test hypotheses, empirically fit models, produce predictions, or estimate policy impacts are based upon some form of quantitative or statistical analysis. This course will provide a basic introduction to statistical methods for political scientists and policy analysts. The course will provide a solid foundation in statistical inference, enabling the student to become a competent producer of basic statistical research. In addition the skills acquired will enable the student to become a somewhat more sophisticated consumer of more advanced research methodologies.

The level of mathematical treatment is somewhat more advanced than an undergraduate class, but will still be moderate (high school algebra). We will not dwell upon derivations. [While it is desirable to have had some prior coursework in statistical inference, this course does begin with the basics. Statistical inference seems to require several repetitions to sink in, so Statistics 101 (or its equivalent) is considered a prerequisite. The student taking PS 400 without having had a prior stats class is will find that the class will be somewhat more difficult, but can be survived. Please discuss this with me if you have not had any prior coursework in statistics.]

The basic organization of the course is: (1) a general introduction to research methodology, (2) descriptive statistics, their use and interpretation, (3) the essential elements of probability, (4) the foundations of statistical inference, (5) and finally an overview of selected hypothesis tests.

Course Requirements

The course requirements are two in-class open-book examinations and a large number of assignments (~15).  The 'midterm' will be late in the semester due to the desire to spread the more mathematical elements across two tests rather than place it all in the final. No seminar paper is required, however, the assignments will contain a modicum of written work in which clarity and sophistication of presentation are critical elements of the assignments. In addition, the assignments will be required to be turned in as Web pages. I will provided detailed instructions on how to create web pages and how to accomplish this task.
 

1. Statistics (3rd Ed.) by David Freedman, Robert Pisani and Roger Purves. Norton. (1998)
2. The Elementary Forms of Statistical Reason. by R. P. Cuzzort and James S Vrettos. St. Martins (1996)
3. Selected Pages on the World Wide Web (See Course Outline Below)
   • Most notably Hyperstat Online by David Lane at Rice University
   • Also, for those inclined to mathematics and puzzles, Cut-the-Knot

Additional Features

A few other articles and books will be placed on reserve in 316 Woodburn.  In addition there are several Web assignments in the course outline.

Also, my class notes and presentation slides will be available via the Web.  These notes are under development and will change constantly throughout the semester.  I expect them to change almost daily.  I strongly recommend against wholesale printing of them!

• PowerPoint Version of Notes (recommended if you wish to print them)
  • Part I
  • Statistical Decision Making
  • Part II
  • Regression Slides
  • Web Page Design

A Word about Computers

All of the assignments for the course may be accomplished with either Microsoft Excel, Corel's Quattro Pro or Stata. In addition, all are required to be posted as web pages. Some assignments may be spreadsheets, while others will require the use of a Web page editor. Instruction will be provided on how to create web pages for those who have never done so.

If you wish, you may elect to use your own graphics, spreadsheet, or statistical software. I will be, however, be unable to provide support except for these three packages. To facilitate the assignments, I will provide some basic instruction in Excel, Quattro Pro for Windows and Stata. I have found spreadsheet software to be quite easy to use and also provides a powerful graphics 'toolkit'. Supplementary instruction on its use will be provided. I encourage you to become comfortable with computers to whatever degree you can. They are becoming increasingly important and useful, and this trend will intensify substantially over the next few years. Basic computer literacy will be an increasingly critical component of job placement.

Basic Computer Literacy and recommended software expected by the end of this class:

•     Email (recommend either Eudora or Netscape Mail)
•     Web Browser (either Netscape 4.7 or MS Internet Explorer)
•     Spreadsheet (Either Excel, Lotus, or Quattro Pro)
•     Web page editor (either MS Frontpage or Virtual Page or even Notepad!)

    If you are uncertain about any of these, please let me know.

As noted above, the assignments will require that you be able to create web pages. Instruction on this will occur in class.

Course Outline with Readings (* = Recommended)

This is an initial reading list. I may add more as the semester goes on.
 

Students with Special Needs

Social Justice Statement

All of the statistics assignments for the course may be accomplished with either Excel or Quattro Pro for Windows and Stata 6.0 or NCSS 6.0. If you wish, you may elect to use your own graphics, spreadsheet, or statistical software. I will, however, be unable to provide support. To facilitate the assignments, I will provide some basic instruction in Excel or Quattro Pro for Windows. I have found these spreadsheet packages to be quite easy to use and also provides a powerful graphics 'toolkit'. Supplementary instruction on its use will be provided. I encourage you to become comfortable with computers to whatever degree you can. They are becoming increasingly important and useful, and this trend will intensify substantially over the next few years. Basic computer literacy will be an increasingly critical component of job placement.

Please note that all assignments need to be 'published' as web pages or files that are linked to a Table of Contents page. Please keep in mind the following:

• Make your assignment readable.
• Please put your name on the top of the page, so that when I print it I will know who's it is!
• Always print preview your web page. I will have to print them to give you feedabck, so try not to waste too much paper. It is easy to have spreadshheets print too wide and be unwieldy as a result
• See the Samples Page

Description of Assignments

The following is a description of the assignments for PS 400. Several of them require that you select your own data to make some point. The Statistical Abstract of the United States (SAUS) is available in the main office with Linda to assist you in finding data. You might also try a few data sources available on the PolyCy Guide 

Using your own favorite Web browser and Spreadsheet (i.e. Internet Explorer and Excel) find data for that you can use for the next assignment. This data should be for a number of cases (e.g. 50 states,) and should measure some political variable of interest to you. I would like for you to acquire this data from the Internet, so I suggest that you use the Statistical Abstract of the United States. In order to facilitate this, I have set up some basic instructions for capturing data from a Web Source. Please follow them for this assignment. The end product will be a data set saved on disk for use with Assignment #2.

 Goals:

1. Use Windows 95/98
2. Learn to capture data from an Internet data source
3. Learn to open a spreadsheet program.
4. Learn how to move around inside the spreadsheet and examine the layout of the software.
5. Learn to use the Button Bars.
6. Enter data in spreadsheet. (Identify the source!)
7. Learn to convert text data to numeric data.
8. Save spreadsheet data and return to it later.

Assignment # 2 Measures of Central Tendency.  Using the data from Assignment #1, calculate the average for that variable. Using the same data, calculate the measures of dispersion used in class. Specifically, calculate the range, standard deviation and variance. Perform the calculations rather than using the @Functions! Then make the spreadsheet presentable. Use labels. Explore the fonts. Make the page you print pretty! Print your results! This exercise is designed to get you used to Windows and a Spreadsheet.  For a step-by-step demo - try the spreadsheet tutorial for calculating the standard deviation (Quattro Pro only).  Click "Open it" on the file Dialogue Box when it appears.  This requires Quattro Pro 7 or later.

Goals:

1. Use Windows 95
2. Find and execute Excel or Quattro Pro within Windows.
3. Learn how to move around inside the spreadsheet and examine the layout of the software.
4. Learn to use the Button Bars.
5. Enter data in spreadsheet.
6. Save spreadsheet data and return to it later.
7. Learn to construct formulas.
8. Learn to construct labels.
9. Learn to visually enhance your presentation.
10. Learn to print output.

[One hint: Use the computer to put your name in the upper right hand corner. Why have it print all that other stuff, and still have to hand write your name?!]

Assignment #3 Statistical Exercises

Complete Problems   #1, #8, #10 in Ch 2, pp 73-75 of Statistics for Social Data Analysis.  You may use Quattro Pro as your calculator.

Assignment #4 Stem-and-Leaf Plot

Produce a Stem-and-Leaf Plot using data that you provide (Identify the source! Always identify the source!). This can be done on a word processor or on a spreadsheet. The plot itself is sufficient for the assignment, as long as the data is clearly identfied. Note how this plot preserves the data as well as producing a picture of the distribution.
Goals:

1. Learn how a stem-and-leaf plot shows the shape of the distribution while preserving the data.

Assignment #5 Frequency Histogram

Again with data you collect produce a Frequency Distribution or Histogram. This can be done with Quattro Pro. Turn in the picture produced by the computer plus a paragraph describing the distribution you observe.  A step-by-step spreadsheet example can be found here.

Goals:

1. Use spreadsheet to input data.
2. Produce a histogram using the software.
3. Successfully print the picture.
4. Interpret a frequency distribution in terms of measures of central tendency and dispersion.

Assignment #6 Pie Graph

Using Quattro Pro and data you select, produce a pie graph. I suggest that you provide data which makes some point. Provide a paragraph which articulates how the graph demonstrates your case.

Goals:

1. Use spreadsheet to input data.
2. Produce a pie graph using the software. Use Titles, Labels and Legends where relevant.
3. Successfully print the picture.
4. Interpret the graph.

Assignment #7 Bar Graph

Using Quattro Pro and data you select (different from #4), produce a bar graph. I suggest that you provide data in order to make some point. Make it different from the data used in #7 above. Provide a paragraph which articulates how the graph demonstrates your case.

Goals:

1. Use spreadsheet to input data.
2. Produce a pie graph using the software. Use Titles, Labels and Legends where relevant.
3. Successfully print the picture.
4. Interpret the graph.

Assignment #8 Line Graph - Time Series Plot

Using Quattro Pro and data you select (different from #4 & #5), produce a line graph which shows data across time (Time series data is not strictly required for this assignment, however, it is usually what is needed for this type of graph to be preferable to a bar graph.). Again, provide data in order to make some point. Provide a paragraph which articulates how the graph demonstrates your case.

Goals:

1. Use spreadsheet to input data.
2. Produce a time-series plot using the software. Use Titles, Labels and Legends where relevant.
3. Successfully print the picture.
4. Interpret the graph.

Assignment #9 Converting Nominal Dollars to Real Dollars (Controlling for Inflation)

Using Quattro Pro and a time series data set that involves money over time (i.e. per capita income for the period 1960-1995) convert the raw (nominal) dollar values into real dollars. See the sample spreadsheet on Converting Nominal Gas Prices to Real gas Prices for this assignment.

Goals:

1. Use the Web to input data to a spreadsheet. The data needs to be a data series in current (Nominal) dollars and the Consumer Price Index (CPI). You may use a specialized CPI or PPI if you desire, but explain or justify.
2. Convert Current dollars to real using the CPI.
3. Produce a time-series plot using the software. Use Titles, Labels and Legends where relevant. Plot both series.
4. Successfully print the picture.
5. Interpret the graph. Note differncs (or similarities in both series)

Assignment #10 Statistical Exercises

Complete Problems  Ch 6, pp 117-119 #2, #3, #4, #8, #9 of Statistics for Social Data Analysis.  You may use Quattro Pro as your calculator.

Assignment #11 T-Test

Develop a hypothesis about whether some sample is different from some population? Collect data for which you have a population mean, and then define a sub sample from it. For instance, collect data on all 50 states, and then test whether a subsample is different from the population mean.

Goals:

1. Formulate a hypothesis about the difference between a population and a sample.
2. Collect the data.
3. Perform the calculations in either Quattro Pro or NCSS
4. Interpret a t-test output and write up a 1 page description of the results of the hypothesis test. See the example of the t-test.

Assignment #12 Two-sample t-test

Develop a hypothesis about the difference between two groups. Collect the data and perform the appropriate comparison between them.
Goals:

1. Formulate a hypothesis about the difference between two sample means.
2. Collect the data.
3. Properly interpret the F-test for different sample variances.
4. Produce and print output using NCSS.
5. Interpret a t-test output and write up a 1 page description of the results of the hypothesis test. See the example of the t-test.

Assignment #13 Statistical Exercises

Complete Problems  Ch 9, pp 291-293 #1, #2, #3, #4 of Basic Social Statistics.  You may use Quattro Pro as your calculator.

Assignment #14 Regression, Correlation, & Multiple Regression

Using a data set which I will supply, test a number of interrelated hypotheses.
Goals:

1. Formulate a set of hypothesis about the relationships between several independent variables and a dependent variable.
2. Test a bivariate hypothesis.
3. Examine a bivariate scatterplot.
4. Test a multiple regression model.
5. Produce and print output using NCSS.
   Interpret the regression output for all analyses and write up a 1 page description of the results of the hypothesis tests. See the example of the regression analysis.