Original Research
Application of statistical contrasts to mean refractive state
African Vision and Eye Health | South African Optometrist: Vol 65, No 1 | a250 |
DOI: https://doi.org/10.4102/aveh.v65i1.250
| © 2006 S Abelman
| This work is licensed under CC Attribution 4.0
Submitted: 19 December 2006 | Published: 19 December 2006
Submitted: 19 December 2006 | Published: 19 December 2006
About the author(s)
S Abelman, Centre for Numerical Analysis and Computational Mathematics, School of Computational and Applied Mathematics, University of the Witwatersrand, Private Bag 3, Wits, 2050, South AfricaFull Text:
PDF (509KB)Abstract
Testing for differences between specific groups or combinations of groups is referred to as comparison or contrast testing. Statistical significance of comparisons can be assessed by first forming contrasts and then testing for their significance. A contrast essentially tests wheth-er or not two means are significantly different, where each mean could be a weighted average of two or more means. Gillan investigated whether the instillation of a cycloplegic (Mydriacyl 1% (tropicamide)) into the right eye of a 30-year-old female sub-ject would affect the variability of her distance refractive state as measured by an autorefractor.The purpose of this paper is to introduce mutually orthogonal linear contrasts of sample data to optometric research. By constructing particular contrasts, mean refractive states are compared before instillation, before and after instillation, and after instillation.
Keywords
Multiple comparisons, contrasts, hypothesis testing
Metrics
Total abstract views: 3476Total article views: 2546