Abstract:
Extensive computer simulations were used to check the means and variances under the null hypothesis distributions of some tests which have long been used to test for significant evidence of disease associations with a polymorphic genetic region. Among the eight rival tests, three are new and derived by the present author, and five were taken from literature. The results based on these large simulated data confirmed both the validity of the tests under null hypothesis and the agreement between nominal and actual significance levels. Also the statistical (not genertical) independence of sharing of haplotypes, between more than two affected siblings in a sibship is established. This was a suspected assumption in the application of one of the tests to families having more than two affected children.
Page(s):
79-88
DOI:
DOI not available
Published:
Journal: Scientific khyber, Volume: 8, Issue: 1, Year: 1995