SECTION 13.1 Summary

• Data collected at regular time intervals form a time series.
• Analysis of time series data should always begin with a time plot.
• A random process is a patternless process of independent observations, which vary constantly about their mean over time. Variation from a random process is sometimes to referred to as an irregular component.
• Common systematic components found in time series include trend, seasonality, and cyclic (autoregressive) behavior.
• The runs test and the autocorrelation function (ACF) are statistical tests of randomness.
• A forecast is a prediction of a future value of the time series.