1951
DOI: 10.1093/biomet/38.1-2.159
|View full text |Cite
|
Sign up to set email alerts
|

Testing for Serial Correlation in Least Squares Regression. Ii

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

6
507
1
18

Year Published

1993
1993
2017
2017

Publication Types

Select...
8
1

Relationship

0
9

Authors

Journals

citations
Cited by 2,235 publications
(532 citation statements)
references
References 25 publications
6
507
1
18
Order By: Relevance
“…This method is designed to find a spline representation for noisy data without allowing the spline to follow every noise feature. This is achieved by starting with an interpolating spline (one knot at every observed data point) and iteratively removing knots to optimize the Durbin-Watson statistic (Durbin & Watson 1950, 1951, which tests for serial correlation in the least squares regression. When applying SNPy we defined a separate spline curve for the F160W, F125W, and F105W bands.…”
Section: Time Delay Measurements With Flexible Light Curve Modelsmentioning
confidence: 99%
“…This method is designed to find a spline representation for noisy data without allowing the spline to follow every noise feature. This is achieved by starting with an interpolating spline (one knot at every observed data point) and iteratively removing knots to optimize the Durbin-Watson statistic (Durbin & Watson 1950, 1951, which tests for serial correlation in the least squares regression. When applying SNPy we defined a separate spline curve for the F160W, F125W, and F105W bands.…”
Section: Time Delay Measurements With Flexible Light Curve Modelsmentioning
confidence: 99%
“…Intensity decay curves were fitted as a sum of exponential terms: also includes statistical and plotting subroutine packages (OConnor & Phillips, 1984). The goodness of the fit of a given set of observed data and the chosen function was evaluated by the reduced x2 ratio, the weighted residuals (Lampert et al, 1983), the autocorrelation function of the weighted residuals (Grinvald & Steinberg, 1974), the runs test (Hamburg, 1989, and the Durbin-Watson parameters (Durbin & Watson, 1950, 1951. A fit was considered acceptable when plots of the weighted residuals and the autocorrelation function showed random deviation about zero with a x2 value not more than 1.5.…”
Section: 'Vv Crlvhmentioning
confidence: 99%
“…88,89 Violation of the premises related to regression analysis were analyzed: normality by the Ryan and Joiner test, 90 homoscedasticity by the modified Levene test 91 and the Brown and Watson test 92 and independence of the regression residues by the Durbin and Watson test. 93 An F-test was conducted to verify the fit to the linear model by means of the significance evaluation of the regression and deviation from linearity. 94 …”
Section: Validation Of the Methodsmentioning
confidence: 99%
“…The results of the Levene test were not significant (p > 0.05) in all curves examined. The independence of the regression residues was evidenced by the Durbin-Watson test 93 since the distribution of the points did not show positive or negative tendencies. Therefore, the use of the OLSM was adequate for estimation of the regression parameters.…”
Section: Limit Of Detection Limit Of Quantification and Linearitymentioning
confidence: 99%