Introduction to Time Series Analysis

Identification

• Time series analysis refers to a set of statistical techniques for analyzing time series data. A time series is a set of observations taken at different points in time. Examples include daily stock prices, weekly sales figures, the monthly Consumer Price Index or the quarterly gross national product.

Benefits

• Time series analysis helps explain changes in variables over time by examining the underlying context of the data. For this reason it can help explain past events, as well as forecast future conditions. Time series analysis is especially popular for economic analysis and forecasting

Types

• Time series analysis can involve multiple regression or forecasting techniques to analyze changes over time and make predictions. Multiple regression is a statistical technique that examines the response of a dependent variable to a series of independent, or predictor, variables. A regression using time series data, for example, could examine presidential approval ratings as a function of economic variables such as the annual federal budget deficit, the monthly unemployment rate and the monthly consumer price index. Forecasting techniques extend regression by using complex statistical models to predict future conditions based on past events.

  Time series analysis also can be a research design in which a researcher uses statistical methods to analyze whether a change in time series data is related to a variable of interest. This is sometimes referred to as an interrupted time series study. For example, a market researcher could use this type of analysis to examine sales figures over time and whether a new marketing campaign, introduced at a particular point in time, helped increase sales.

Function

• Time series analysis has a wide range of uses. Marketing departments can use these techniques to help plan sales strategies. Financial professionals use time series analysis to forecast interest rates and other investment conditions. Economists use time series techniques to forecast future economic conditions. Political scientists use these techniques to analyze poll data, such as presidential approval ratings.

Considerations

• Because of its complexity, time series analysis requires use of a computer with statistical software. Popular statistical software programs include SAS and SPSS. In addition, there are specialized programs designed for time series, such as RATS (Regression Analysis of Time Series).

  Spreadsheet programs such as Excel are not designed for complex statistical procedures, such as time series analysis.