Interpreting Nonsignificant Outcomes of Heterogeneity Tests in
Gene Mapping
C. Merette, T. Lehner, J. Ott
AJHG,49, 1381--1384 (1991)
This is a Letter to the Editor.
quote from the first paragraph..
A statistical test leads to either a significant or a nonsignificant
outcome. For example, when one tests whether the recombination fraction
(theta) for two loci is the same in different families, a significant
result is interpreted as representing evidence for genetic heterogeneity.
A nonsignificant test result, however, is more difficult to interpret.
Generally, such a result is due to either (1) absence of the effect
investigated or, if the effect is present, (2) lack of power (sample
size is too small). Often, only the former explanation for a nonsignificant
test result is invoked, but it is, of course, important to consider
how one might be able to distinguish between the two possible reasons
for the failure to find a significant effect.
