1987
DOI: 10.2307/2095357
|View full text |Cite
|
Sign up to set email alerts
|

Estimation with Cross-National Data: Robust and Nonparametric Methods

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2

Citation Types

1
29
0

Year Published

1988
1988
2018
2018

Publication Types

Select...
6
2

Relationship

1
7

Authors

Journals

citations
Cited by 73 publications
(30 citation statements)
references
References 32 publications
1
29
0
Order By: Relevance
“…We also computed estimates of the log-polynomial model using least median of squares, least trimmed squares, and a robust regression that uses Huber estimates followed by biweighted least squares. These techniques are robust with regard to outliers in residuals and in carriers (37,38). The robust estimates yield substantively identical results to the more familiar, ordinary least squares reported here.…”
supporting
confidence: 65%
See 1 more Smart Citation
“…We also computed estimates of the log-polynomial model using least median of squares, least trimmed squares, and a robust regression that uses Huber estimates followed by biweighted least squares. These techniques are robust with regard to outliers in residuals and in carriers (37,38). The robust estimates yield substantively identical results to the more familiar, ordinary least squares reported here.…”
supporting
confidence: 65%
“…Deleting the USSR from the analysis did not significantly change the results. ‡ ‡ The results reported are for ordinary least squares, with SE based on 500 replications of case-based bootstrap resampling, which are preferable to normal theory SE in this context (37,38). To reduce colinearity among polynomial terms, we centered the population and gdp variables by using deviations from the mean of the logs instead of raw values for the linear in log terms and in generating the polynomial terms.…”
mentioning
confidence: 99%
“…Table I presents the means, standard deviations, and zero-order correlations of study variables, while Table II provides the regression results. We used robust regression to test our hypotheses because OLS regression has been shown to perform poorly when used with cross-national data (Dietz, Frey, & Kalof, 1987). A number of the variables in our model suffer from skewness, which persisted after log transformations.…”
Section: Control Variablesmentioning
confidence: 99%
“…Another measure that has been shown to be associated with increased environmental pressure or degradation is urbanization, which is measured as a percentage of total population living in urban areas. While urbanization is important for understanding consumption demands [68] it is particularly important for understanding water usage. As Longo and York [12] (p. 76) explain, "Demographic changes, particularly population shifts to urban areas, are often associated with shifts in water use patterns.…”
Section: Independent Variablesmentioning
confidence: 99%