Share
Time Series Analysis
Regression Techniques
Volume:
9
Other Titles in:
Quantitative/Statistical Research
Quantitative/Statistical Research
January 1990 | 96 pages | SAGE Publications, Inc
The great advantage of time series regression analysis is that it can both explain the past and predict the future behaviour of variables. This volume explores the regression (or structural equation) approach to the analysis of time series data. It also introduces the Box-Jenkins time series method in an attempt to partially bridge the gap between the two approaches.
Introduction
Time Series Regression Analysis
A Ratio Goal Hypothesis
The Error Term
Time Series Regression Model
Nonautoregression Assumption
Consequences of Violating the Nonautoregression Assumption
Conventional Tests for Autocorrelation
An Alternative Method of Estimation
EGLS Estimation (First-Order Autocorrelation)
Small Sample Properties
The Ratio Goal Hypothesis Reconsidered
Extension to Multiple Regression
Conclusion
Alternative Time-Dependent Processes
Alternative Processes
Testing for Higher Order Processes
Process Identification
Estimation
Example
Example
Conclusion
Time Series Regression Analysis
Distributed Lag Models
Lagged Endogenous Variables
Testing for Autocorrelation in Models with Lagged Endogenous Variables
Estimation
EGLA Estimation
Example
A Revised Ratio Goal Model
Interpreting Distributed Lag Models
Conclusion
Forecasting
Forecast Error
Forecast Generation
Modifying the Forecast Equation
Forecast Evaluation
Example
Conclusion
Summary
