BIOL 2007- NOTES
ON POPULATION GENETICS TUTORIAL - Single locus selection and drift
NOTE ON QUESTION 2a: Yates' Correction
Yates' correction supposedly "corrects for continuity", in other words,
it is supposed to make the calculated statistic conform more
closely to an actual chi-square distribution, which is a continuous
distribution. However, the correction is very small when you have
lots of data, as here, where the largest proportional correction is
less than 1% (i.e. 74 goes to 74.5). Given that Yates' correction
is overly conservative anyway, many people just don't use it any more.
See: p.736 in Sokal, RR; Rohlf, FJ (1981): Biometry. 2nd ed.
Freeman, San Francisco.
Back to: Tutorial