Nonlinear Regression with Autocorrelated Errors

“…The Yule-Walker method (Gallant and Goebel 1976; also known as the two-step full transform method) was used to account for autoregressive errors if the Durbin-Watson statistic (Johnston 1972) identified their presence. In all cases where significant trends were detected, trend residuals were analysed visually to determine whether bias was present or whether they were randomly distributed.…”

Section: Statistical Analysesmentioning

confidence: 99%

Effective Long-term Ecological Monitoring Using Spatially and Temporally Nested Sampling

Hewitt

Thrush

2007

Environ Monit Assess

View full text Add to dashboard Cite

Effective environmental management requires documentation of ecosystem status and changes to that status. Without long-term data, short-term natural variability can mask chronic and/or cumulative impacts, often until critical levels are reached. However, a trade-off generally occurs between sampling in space and time. This study analyses a spatially and temporally nested long-term (12 years) monitoring programme conducted on benthic macrofauna in a large harbour. Sampling was carried out at six sites for 5.5 years, after which only two sites were sampled for the next 5 years. After this period, all six sites were sampled for another 2 years. While ecology is frequently thought of being highly variable, this design was able to detect trends, and cycles, in abundance, with only around 10% of species at each site exhibiting unpredictable temporal variability. Sites exhibiting similar trends in the abundance of a species over the 12.5-year period were generally spatially contiguous, and the spatial scale of change could be assessed. Continuous sampling at two sites identified whether changes in unsampled sites were related to long-term cycles. Moreover, this sampling provided a long-term background of temporal fluctuations against which to assess the ecological significance of observed changes.

show abstract

“…This latter requirement imposes a curvilinear, in this case a square, delayed effect of a regressor on the model. The polynomial restriction on the regression model is based on initial plots of exchange rate and export data that show nonlinear trends in high-technology trade (Gallant and Goebel, 1976;Harvey, 198 1).…”

Section: The Modelmentioning

confidence: 99%

Real exchange rate effects on technology-intensive exports: A three-country multiproduct study

Afriyie

Kundu

1994

The International Trade Journal

View full text Add to dashboard Cite

Kofl Afriyie and Sumit Kundu7% study explores the possible effects of changes in real exchange rates on exports of six high-technology products from Germany, Japan, and the United States over a 24yearperiod. Findings in tbis article indicate significant county and product dzyferences in exchange rate effects on exports of high-technology products. These observations suggest that using aggregate balance-oftrade data to determine the influence of exchange rates on a nation's export perforrnunce can mislead policymukers about the nature of foreign trade and national competitiveness in an increasingly globalized industrial economy.

show abstract

Nonlinear Regression with Autocorrelated Errors

Cited by 49 publications

References 0 publications

Differences between near-surface equivalent temperature and temperature trends for the Eastern United StatesEquivalent temperature as an alternative measure of heat content

Differences between near-surface equivalent temperature and temperature trends for the Eastern United StatesEquivalent temperature as an alternative measure of heat content

Effective Long-term Ecological Monitoring Using Spatially and Temporally Nested Sampling

Real exchange rate effects on technology-intensive exports: A three-country multiproduct study

Contact Info

Product

Resources

About