Syllabus
Introduction to Applied Quantitative Methods
Winter 2009
Mondays 2:00-3:20pm, Rytirska 31, Room 201
Jan Rovny jrovny@gmail.com
Course Requirements
This course is designed for students whose primary interest lies in social sciences. It thus aims to make statistics and programing understandable to students who tend to have limited knowledge of mathematics, and who may even be skeptical of using mathematics and statistics in social research.
While I do not expect you to be confident users of math, I do expect you to be keen learners. Learning quantitative methods requires continuous practice and repetition. This course steadily builds on previous material which is needed for understanding subsequent lectures. It is thus essential that you come to class, do all your readings on time and complete all homework assignments. Conscientious students who make good progress will be rewarded with good grades.
Readings
I understand that for some of you reading about quantitative methods in
English will be a double challenge. I have thus selected shorter readings that use as little statistical jargon and mathematical symbols as possible.
Neither can, however, be completely avoided. Please be patient with your readings. Know that you might need to read some of them twice in order to understand them. Also, please ask in class about any difficulties you might have. If you are interested, I can organize extra-curricular sessions that will provide additional help to students. I will also be able to give some explanations in Czech if necessary.
I will make all readings available to registered students in electronic format on my website.
Statistical Software Package ?R?
This course will teach you how to program in R. R is considered to be the most sophisticated statistical package and is used by many serious social science methodologists. It has three major advantages: it is extremely flexible, it has very pretty graphics and it is FREE! The software is available for Linux, Mac OS X and Windows. Please download it from here. 2
Evaluation
Attendance and Participation 10% You are expected to be present at all sessions. You will be granted one unexplained absence during the semester.
All other absences must be explained with a legitimate excuse, such as a medical note. I will deduct 1% for every unexcused absence. Besides attending, you are expected to actively participate during lectures.
Midterm Assignment 20% The midterm assignment will be handed out in class on November 23. It will be due on November 25 at 5pm. I will deduct 1% for each day late.
Homework 30% In order to practice your programing skills and review the theoretical material we cover, you will be asked to complete six small homework assignments throughout the semester. Each homework will be worth 5% of your final grade. Each homework is due one week after it is assigned. I will deduct 1% for each day late.
Final Assignment 40% You will have two options regarding the final assignment: 1. Advanced students, who are working on their MA theses and wish to develop a quantitative dimension of their research, may submit a quantitative research paper. This paper will develop testable hypotheses based on the student?s theoretical framework and it will test these hypotheses using appropriate quantitative methods. Students who chose to pursue this option should contact me early in the semester. I will be happy to guide you throughout your writing. 2. Students who are not ready to test their theoretical propositions quantitatively may choose to complete the final assignment, which will be a take home assignment similar to the midterm. It will be given out on the last day of class on January 4. It will be due on January 6 at 5pm. I will deduct 1% for each day late. 3
Academic Honesty and Plagiarism
Plagiarism is the practice of taking someone else?s work or ideas and passing them as your own. As such, it constitutes one of the most serious academic offenses. Consequently, I expect all students to adhere to rules of academic honesty by completing all their assignments individually, and using proper rules of citation when presenting someone else?s ideas.
Course Schedule
The course schedule may adjust according to our progress.
Week 1, October 5th
Introduction to Quantitative Methods
? Structure of the Course and Evaluation
? Scientific Inquiry: Theory and Methods
? Qualitative and Quantitative Methods: Strengths and Weaknesses
? Quantitative Applications: Are Extreme Right Parties Really on the
Right?
Reading: 1. Sir Arthur Conan Doyle ?Sherlock Holmes: The Speckled Band?
Homework 0: What does Sherlock Holmes?s adventure have to do with quantitative methodology?
Week 2, October 12
Elements of Quantification
? Types of Variables: Levels of Measurement
? Describing Variables: Measures of Central Tendency, Measures of Variance,
Measures of Association
Reading: 4 1. Saarle ?Measurement? 2. Lewis-Beck ?Data Analysis? pp.1-22
Homework 1: Describing and Summarizing Data
Week 3, October 19
Elements of Statistics
? Probability Distributions, Central Limit Theorem
? Basic Hypothesis Testing (means tests), Types of Errors
Reading: 1. Lewis-Beck ?Data Analysis? pp.30-35; 38-41
Week 4, October 26
Introduction to R
? Working with R
? Data Management, Descriptive Statistics, Tabulations, Correlations,
Graphing
Reading: 1. Introduction to R (manual)
Homework 2: Data Management, Tabulations, Two-way Plots, Box Plots and interpretation using R
Week 5, November 2
Linear Regression 1
? Statistical Inference, Theorizing Relationships, Modeling Relationships
? The Logic of the Linear Regression Model
? Estimation of Linear Regression 5
Reading: 1. Lewis-Beck ?Data Analysis? pp.41-53 2. Gujarati pp.37-52
Week 6, November 9
Linear Regression 2
? Assumptions and Properties of Least Square Estimators
? Decomposition of Sample Variance ? Errors and Model Fit
? Programing and Plotting Simple Regression in R
Reading: 1. Gujarati pp.58-81
Homework 3: Calculate a case of simple regression by hand, program and plot the same simple regression model in R and interpret the results.
Week 7, November 16
Multiple Regression
? Multiple Regression Model, Assumptions and Estimation
? Uncertainty: Standard Errors and Variance-Covariance of Estimators
? Comparing Regression Coefficients, Non-Linear Regression Models
Reading: 1. Lewis-Beck ?Data Analysis? pp.53-60
Homework 4: Estimate Multiple Regression Models in R and interpret the results. 6
Week 8, November 23
Hypothesis Testing and Model Comparison
? T-tests, F-tests
? Model Comparisons: AIC, BIC
Reading: 1. Gujarati pp.119-142 (on Estimation and Hypothesis Testing) 2. Gujarati pp.536-7 (on R2, AIC, BIC)
Midterm Assignment handed out in class, due November 25 at 5pm
Week 9, November 30
Interaction Effects
? Interactions Between Dummy Variables
? Interactions Between Dummy and Continuous Variables
Reading: 1. Lewis-Beck ?Data Analysis? pp.65-67 2. ***optional*** Brambor et al. 2006
Week 10, December 7
Assumption Violations 1
? Multicollinearity: Causes, Consequences, Diagnosis, Remedies
? Heteroscedasticity: Causes, Consequences , Diagnosis, Remedies
Reading: 1. John Fox ?Regression Diagnostics? pp.5-21; pp.49-61 2. ***optional*** Gujarati pp.341- 370; 387-422
Homework 5: Estimate and Interpret Complex Regression Models in R.
Perform hypothesis tests and discuss interaction effects. 7
Week 11, December 14
Assumption Violations 2
? Model Mis-specification: Consequences, Diagnosis, Remedies
? Measurement Error: Random v. Systematic Error, Remedies
? Influential Data Points and Outliers: Plotting, Diagnosis, Remedies
Reading: 1. Gujarati pp.506-529 (on model specification and measurement error) 2. John Fox ?Regression Diagnostics? (Outlying and Influential Data) pp. 21-40
Homework 6: Estimate Regression Models in R with various assumption violations. Diagnose these violations, suggest possible remedies and perform them.
Week 12, January 4
Conclusion: The Limits of Linear Regression
? Role of Data Gathering
? Importance of Theory
? What to do with non-continuous dependent variables: a short excursion into Maximum Likelihood Estimation
Final Assignment handed out in class, due on January 6 at 5pm
Students who select to write a quantitative research paper are exempt from the final assignment. 8
Annotation
Course Description
Recent years have witnessed a surge in digital data that describe various social phenomena such as economic indicators, incidences of armed conflict, voter preferences, indexes of government structures or demographic patterns.
Simultaneously, the increasing computational power of our computers allows more and more sophisticated analyses of quantitative data. Social sciences have thus turned to enrich their traditionally qualitative methods by using these quantitative indicators and statistical programing for developing and testing scientific knowledge. While all social sciences yield fruitful qualitative research programs, quantitative analysis opens further research avenues and supplies important additional evidence about the social world.
Basic understanding of quantitative data analysis is thus an essential skill of a successful social scientist.
This is an introductory course of quantitative methods for social scientists.
It aims to prepare researchers to intelligently apply basic statistical methods for the purposes of empirical research. This course is thus a practical guide to statistical application for social scientific research. However, to become effective users of statistics, students must learn elementary statistical concepts and theory, and be aware of the various assumptions of the methodology. The course will consequently combine simple exposition to statistical theory with practical use of statistical modeling. The course will alternate between lectures and practical lab sessions where students will be encouraged to apply the material while learning to program in the most sophisticated (and free!) statistical software package R.