Angular transformation – another effect of different sample sizes
Angular transformation has been widely used to analyze proportions or percentages in ecology. A new type of artifact can arise by applying angular transformation to proportions or percentages. Proportions with the same mean or expected value, but with different sample sizes (denominator of proportions), have different means or expected values after this transformation. Overdispersion causes a larger deviation of the expected value after the transformation. Angular transformation should be avoided unless the sample size (denominator of proportions) is the same for all the proportions analyzed.
Key wordsangular transformation bias sample size statistical analyses transformation
Unable to display preview. Download preview PDF.
- Crawley M. J. (1993) GLIM for Ecologists. Blackwell Science, Oxford.Google Scholar
- McCullagh P. & Nelder J. A. (1989) Generalized Linear Models, 2nd edn. Chapman & Hall, London.Google Scholar
- Snedecor G. W. & Cochran W. G. (1967) Statistical Methods, 6th edn. Iowa State University, Ames.Google Scholar
- Sokal R. R. & Rohlf F. J. (1995) Biometry, 3rd edn. Freeman, New York.Google Scholar
- Zar J. H. (1999) Biostatistical Analysis, 4th edn. Prentice Hall, Upper Saddle River, New Jersey.Google Scholar