Today, water-quality and other scientific data can be measured by automatic recorders or remotely by satellite only seconds apart from one another. Agencies have begun to store, present, and analyze these "real-time data". Data recorded this closely together usually violate the independence assumption of standard statistical procedures – one observation is partly or predominantly a replicate of the previous one. The consequence is that statistical tests such as trend analysis and regression provide invalid results when used on data stored every 1, 5, or 15 minutes apart.

Our Time Series Methods course focuses on performing trend tests and building regression models for data measured frequently in time.

Topics include:
What is serial correlation? How to test for it?
Effective sample size for correlated data
Effects of serial correlation on hypothesis tests
Comparisons to standards for correlated data
Building regression models using time series methods
Trend analysis using time series methods
Autoregressive and Moving Average Time Series models
Forecasting WQ variables – how good is my forecast?
Bootstrap methods for time series

Next class:
Sign up for our newsletter or follow us on Twitter (@PracticalStats) to receive a notice of upcoming offerings.

Course Outline

Download the Course Brochure