Notes: Rounding Effects in Simple Linear Regression
Author: Bower, D. R.
Source: Forest Science, Volume 16, Number 3, 1 September 1970 , pp. 301-303(3)
Publisher: Society of American Foresters
Abstract:In simple linear regressions relating tree diameters at the beginning and end of 15- and 20-year periods, an increase in the measurement interval for the independent variable from tenths of inches to whole inches caused an 8.7-percent negative bias in the estimated slope, an average loss of 6.2 percent in r², and an increase in mean squared residuals of 29.2 percent. Bias and loss in precision occurred in both large and small samples. Accuracy of measurement of independent variables should be consistent for data used to derive a regression and for new data to which the regression equation is to be applied. Forest Sci. 16:301-303.
Document Type: News
Affiliations: Biometrician, Weyerhaeuser Research Center, Centralia, Wash.
Publication date: 1970-09-01
- Forest Science is a peer-reviewed journal publishing fundamental and applied research that explores all aspects of natural and social sciences as they apply to the function and management of the forested ecosystems of the world. Topics include silviculture, forest management, biometrics, economics, entomology & pathology, fire & fuels management, forest ecology, genetics & tree improvement, geospatial technologies, harvesting & utilization, landscape ecology, operations research, forest policy, physiology, recreation, social sciences, soils & hydrology, and wildlife management.
- Membership Information
- ingentaconnect is not responsible for the content or availability of external websites