The effects of temporal aggregation on tests of linearity of a time series

Time-series aggregates are often used in performing tests for departure from linearity. In this paper, we study the effects of temporal aggregation on testing for linearity, basing our analysis on both time- and frequency-domain tests. The results show that temporal aggregation weakens nonlinearity...

Full description

Saved in:
Bibliographic Details
Published in:Computational statistics & data analysis 2000-07, Vol.34 (1), p.91-103
Main Authors: Teles, Paulo, Wei, William W.S
Format: Article
Language:English
Subjects:
Arranged autoregression
Bispectral test
Exact sciences and technology
Inference from stochastic processes
time series analysis
Linear inference, regression
Mathematics
Noncentrality parameter
Nonlinear time-series
Power loss
Probability and statistics
Sciences and techniques of general use
Statistics
Time-series aggregates
Volterra expansion
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Time-series aggregates are often used in performing tests for departure from linearity. In this paper, we study the effects of temporal aggregation on testing for linearity, basing our analysis on both time- and frequency-domain tests. The results show that temporal aggregation weakens nonlinearity and reduces the power of the tests. The impact is severe. The use of aggregate data greatly hampers the detection of the nonlinear nature of the process.
ISSN:0167-9473
1872-7352
DOI:10.1016/S0167-9473(99)00072-9