UNIVERSITY OF COLORADO                                                                 Fall 2006

DEPARTMENT OF SOCIOLOGY

                                                             

SOCY 4061 SOCIAL STATISTICS

TR 2:00-3:15 PM   Ketchum 119, Ketchum 33

 

Instructor:  Jane Menken                                                                     menken@colorado.edu

103 IBS #1 (1416 Broadway)                                                                             303 492 8148

Ketchum 210 (shared faculty office)                                                                 303 492 4155

Office Hours: 3:30-4:30 TR in Ketchum and by appointment in IBS

Instructor:  Jill Williams                                                                    jill.williams@colorado.edu

105A IBS #3 (1424 Broadway)                                                                              303 492 5253

 

Objectives

 

This course is intended to introduce students to quantitative analysis.  If you already

Miller McPherson, Lynn Smith-Lovin, and Matthew E. Brashears. 2006. Social isolation in America: Changes in Core Discussion Networks over Two Decades.  American Sociological Review: 71(3) 353-375.

then you do not need to take this course. 

 

If you do not yet possess these skills, the objective of this course is to provide you with the opportunity to gain them and be confident that you can both understand quantitative research studies and carry out your own analyses.  The course is intended to prepare students for additional courses on multivariate quantitative analysis and research methods.   The approach throughout will be experiential - we use a recent study of social isolation that has received a great deal of media attention and will replicate most of the published analyses and add some of our own.  The approach to learning statistical theory is visual, using CU Professor Gary McClelland’s unique on-line text, Seeing Statistics, and the approach to learning statistical analysis is through doing analyses using STATA and CU graduate Larry Hamilton’s book, Statistics with STATA.

 

The course begins with our reading the focal article to understand what the authors say they did.  We then turn to an introduction to surveys and, in particular, the General Social Survey (GSS) 1985 and 2004 – the two surveys used in the McPherson et al. article.  We look at the survey questions and their correspondence to the questions the article addresses.  We begin to use STATA first to manage and then to describe the 1985 GSS data – in words, tables, and graphs.  The next portion of the course emphasizes statistical theory – how can we use samples to learn about the population


group of interest?  When we use a sample and calculate a mean, why do we choose to use the mean?  How sure are we that our sample mean is close to the true mean of the population sampled?  Can we estimate how far away from the true value a sample value may be?  If we use two samples – GSS 1985 and GSS 2004 – to look at how confidante networks have changed over time, how do we know whether the difference is real or could have happened simply by taking two samples from a population in which no change over time has taken place?  We will also ask questions about relationships between individual characteristics – for example, do women have more confidantes than men?  Is the number of confidantes greater for young people compared to older people?  Does it vary by education? 

 

Focal article and selected media responses:

Miller McPherson, Lynn Smith-Lovin, and Matthew E. Brashears. 2006. Social isolation in America: Changes in Core Discussion Networks over Two Decades.  American Sociological Review: 71(3) 353-375.

Social Isolation: Americans Have Fewer Close Confidantes.  NPR All Things Considered, June 24, 2006 · Debbie Elliott speaks with sociology professor Lynn Smith-Lovin of Duke University about a new survey documenting what seems to be Americans' growing social isolation. Back in 1985, respondents reported, on average, that they had three people in their lives who were close confidantes. They now report having two people with whom they can discuss important personal topics.  http://www.npr.org/templates/story/story.php?storyId=5509381  

Pat Burson.  Study: Fewer find close friends.  Newsday, July 29, 2006.  Reprinted in Daily Camera

Ellen Goodman.  The demise of friendship.  Syndicated column.  Daily Camera, June 30, 2006.

Ann Hulbert.  The Way We Live Now: Confident Crisis.  New York Times Magazine, July 16, 2006.

Hilary Macgregor.  Study finds Americans need friends.   Los Angeles Times, July 1, 2006.  Reprinted in Daily Camera.

 

Required Texts:   

For basic statistical theory and concepts:

Gary McClelland. 1999.  Seeing Statistics.  Duxbury Press.  www.seeingstatistics.com/ 

Access free through any CU website.  If you’re accessing the web from outside of CU, you need to have VPN dialer on your computer.

For applied statistics and use of STATA:

Lawrence C. Hamilton. 2006.  Statistics with STATA.  Belmont CA: Duxbury, an imprint of Thomson Brooks/Cole.  You may use earlier versions of this book.

 

Recommended Texts:  If you’d like to refer to more standard statistical texts – any edition published in the past five years is fine.

Frederick J. Gravetter and Larry B. Wallnau. 2005. Essentials of Statistics for The Behavioral Sciences. Wadsworth Group/Thompson Learning.

James T. McClave and Terry Sincich.  Statistics. Ninth Edition, 2003 or Tenth Edition, 2006. New York: Prentice Hall.

Larry Gonick & Woollcott Smith.  The Cartoon Guide to Statistics.  2005.  New York: Collins Reference, an imprint of HarperCollins Publishers.  This book is inexpensive, accurate, and a lot more fun to read than most stat texts!

 

Online:

Philip B. Stark.  SticiGui: Statistics Tools for Internet and Classroom Instruction
with a Graphical User Interface

http://www.stat.berkeley.edu/users/stark/SticiGui/index.htm

David Lane. HyperStat Online Statistics Textbook, Rice Virtual Lab in Statistics

http://onlinestatbook.com/rvls/

South Africa Distance Learning Project: The Analysis of South African Household Survey Data

http://saproject.psc.isr.umich.edu/

 

Assignments

 

There will be weekly assignments and three short quizzes.

 

The course highlights four components of learning statistics:  terminology, computation, application, and interpretation of results.  Although most important, the application component depends on mastery of terminology and computations.  Each component requires a different type of assignment. 

 

First, it helps to view statistics as a language with its own terms and concepts.  Students must master the terms and concepts just as they must learn nouns, verbs, and grammar of a foreign language.  Seeing Statistics defines and explain relevant terms and concepts.  I will also review this material but expect students to read the assigned material on their own before class.  Then, we can devote more class time to discussion of issues of application and interpretation.  A series of three quizzes covers the definitions, terms, and concepts of the readings.

 

Second, to understand the concepts and techniques, students need to do computations themselves once or twice.  Abstract equations and ideas take on concrete meaning when one substitutes numbers into formulas and calculates statistics either by hand or using a computer, but not the automatic statistical features of a package like STATA.  Otherwise, statistical packages on computers can do the calculations most efficiently.  I will assign several problem sets using material in Seeing Statistics or using the GSS1985 and/or the GSS2004 for you to complete and turn in for grading.  You may pair with another student to complete these assignments.

 

Third, the ultimate goal of learning statistical techniques is to apply them to real research problems.  With that in mind, we will use the McPherson et al. article extensively.  We will review in class the relevant substantive issues they raise and the ways in which the statistical techniques they use to address the substantive issues.  I will assign a series of short weekly papers in the second half of the course based on analysis of the GSS1985 and/or the GSS2004.  The papers involve using the techniques studied in class and interpreting your own statistical results.  As part of the papers, I may also assign a set of questions that involve the interpretation of the statistical results in a table from a published article.

 

The papers should be clearly written, as if for a professional audience.  One needs considerable practice to write clear, organized, and theoretically meaningful prose when describing statistical results.  Make every attempt to rewrite, revise, edit, and (perhaps most importantly) organize your papers until they read smoothly, proceed logically, and highlight the substantive meaning of the statistical results.

 

You’ll use STATA for Windows on the machines in the Ketchum Labs for the paper assignments.  STATA, a set of pre-written computer programs, performs the step-by-step calculations needed to obtain nearly any desired statistic.  It also includes procedures to organize, access, graph, and print a set of data.  Users need only to select the desired procedures and identify the variables to perform them on.  You will learn to write short STATA DO files (a series of commands) to carry out your analyses.  I assume little or no experience with STATA and an important part of the course is your introduction to data management and statistical analysis using this program.

 

Grading

 

There will be an assignment just about every week: three short quizzes covering the terms and concepts of the readings, six problem sets using material from Seeing Statistics or the GSS and four short papers (at most 3 pages) based on interpretation of your computer output from STATA and statistics presented in a table from a published article.  Each quiz constitutes 10% of the grade, each problem set constitutes 5% of the total grade, and each short paper constitutes 10% of the total grade.

 

Schedule

 

The schedule below lists the topics, readings, and assignment for each week (I will accommodate students whose disability requires special arrangements or whose religious obligations conflict with any assignments).  Although I hope we can stick roughly to this schedule, some adjustment during the semester may prove necessary, especially since this is a completely new approach to teaching this course.

 

 

Week Date               Topic                                                 Reading       

  1        Aug 29           Introduction                          

Aug 31           Discussion                                        McPherson et al.

media responses

  2        Sep 5              Description of GSS                          Handout from ICPSR

                                                                                                Hamilton Ch 1

Introduction to Data Management and Description using STATA (with a short introduction to Seeing Statistics)

            Sep 7              Getting onto STATA                        Hamilton Ch 2

  3        Sep 12           Data and Comparisons                  McClelland Chs 1-2                       

            Sep 14           Describing the Center                     McClelland Ch 3

  4        Sep 19           Finding data on the web                Handouts

            Sep 21           Describing data                                Hamilton Ch 2

  5        Sep 26           Describing the Spread                    McClelland Ch 4

            Sep 28           Seeing Data Again                          McClelland Ch 5

Underpinnings of Statistics

6          Oct 3               Probability                                         McClelland Ch 6

            Oct 5               Normal distribution                          McClelland Ch 7

  7        Oct 10             Inference and confidence              McClelland Ch 8

            Oct 12                continued

  8        Oct 17             One-Sample Comparisons                        McClelland Ch 9

            Oct 19                continued

  9        Oct 24             Two-Sample Comparisons                        McClelland Ch 10

            Oct 26                continued

10        Oct 31             Categorical Data Comparisons     Handouts

            Nov 2                 continued

11        Nov 7              Correlation and Regression          McClelland Ch 12

            Nov 9                 continued

Data Analysis and Interpretation of Results

12        Nov 14           Back to the data                               Hamilton Ch 4

            Nov 16              continued

 

Fall Break and Thanksgiving

 

13        Nov 28           Graphing Data in STATA               Hamilton Ch 3

            Nov 30              continued

14        Dec 5              Summary Statistics and Tables    Hamilton Ch 4

            Dec 7                 continued

15        Dec 12           Linear Regression Analysis          Hamilton Ch 6

            Dec 14           Wrapup                                            

16        Dec 19           Last paper due