Time Series Analysis
A 4-Day Remote Seminar Taught by
Jon Pevehouse, Ph.D.
Time series data are incredibly common in the natural and social sciences. Any process that is measured repeatedly over time yields data with properties that must be properly modeled.
This seminar covers statistical methods used to analyze these data-generating processes that occur over time. Many textbooks and courses on time series focus on techniques that merely adjust for temporal variation and dependencies. In contrast, this seminar treats time series properties as phenomena of substantive interest, not simply as a statistical nuisance.
Starting August 3 we are offering this seminar as a 4-day synchronous*, remote workshop for the first time. Each day will consist of a 3-hour live lecture held via the free video-conferencing software Zoom. You are encouraged to join the lecture live, but will have the opportunity to view the recorded session later in the day if you are unable to attend at the scheduled time.
Each lecture session will conclude with a hands-on exercise reviewing the content covered, to be completed on your own. An additional lab session will be held Tuesday and Thursday afternoons, where you can review the exercise results with the instructor and ask any questions.
*We understand that scheduling is difficult during this unpredictable time. If you prefer, you may take all or part of the course asynchronously. The video recordings will be made available within 24 hours of each session and will be accessible for two weeks after the seminar, meaning that you will get all of the class content and discussions even if you cannot participate synchronously.
Closed captioning is available for all live and recorded sessions.
MORE DETAILS ABOUT THE COURSE CONTENT
The course introduces participants to time series methods in the context of applications to various types of data. It begins with a discussion of univariate models and diagnostic tests, including autoregressive moving average (ARMA) models, interrupted time series analysis, autoregressive conditional heteroskedastic (ARCH) models, and tests for stationarity including fractional non-stationarity.
The course then moves to multivariate models including time series regression, reduced form methods (Granger causality and vector autoregression), cointegration, and error correction models. The aim is to provide a working knowledge of important time series diagnostic tests and models.
The seminar will include lab sessions to provide practice implementing the methods, using data provided by the instructor.
All methods will be illustrated using statistical software. Stata will be the primary package, but parallel code will be provided in R. To participate in the hands-on exercises, you are strongly encouraged to have your computer available with Stata or R already installed.
If you’d like to use R for this course but don’t yet have much experience with that package, here are some excellent on-line resources for building your R skills.
WHO SHOULD Register?
You should have a firm grounding in basic probability and statistics, as well as in linear modeling. Experience with Stata or R is also highly desirable.
1. Univariate Models
a. ARMA/ARIMA models
b. Intervention models
c. ARCH models
2. Tests of stationarity
a. Tests for unit root
b. Tests for fractional non-stationarity
3. Time Series Regression Models
a. Basics of time series regression
b. Autoregressive distributed lag models
4. Methods for Stationary Data
a. Granger causality
b. Vector autoregression
c. Innovation accounting
d. Dynamic conditional correlation models
5. Methods for Non-stationary data
b. Vector error correction models
c. Innovation accounting