This study has found no empirical evidence to support Annett's contention that handedness is maintained as a balanced polymorphism by the intellectual advantage shown by heterozygotes relative to homozygotes (Annett, 1991c).
The conclusion seems inescapable that if the balanced polymorphism hypothesis is true then it will be very difficult to test within university students, since no conceivable study is likely to have sufficient power to distinguish differences in mean IQ of small fractions of an IQ point.
That may be true at a strong level, but is not necessarily true at a weak level, since, as noted earlier, Annett (1991a) has herself reported data on differences in reading ability in a student population, which she has interpreted as providing support for the hypothesis of a balanced polymorphism. Given that that study found significant differences between handedness groups then the present study was a legitimate attempt to find similar effects on other measures of intellectual ability in a group of students of similar intelligence level (and the differences might have been present empirically, irrespective of the truth or otherwise of the specific hypothesis of a balanced polymorphism).
Strictly the hypothesis of a balanced polymorphism requires that the three genotypes differ in their fitness, which must be construed in a strict sense in terms of differential survival of offspring.
The conclusion must therefore be that if Annett's balanced polymorphism model is to have any viability it must assume that the disadvantages of both homozygotes are measured on the same metric, of intellectual ability.
The analyses of the present study leave Annett's hypothesis of a balanced polymorphism for handedness resulting from an intellectual advantage for right-shift heterozygotes, in a difficult, if not impossible, position.
If the empirical results demonstrated by Annett are replicated (and we have been unable to find any such effect in a university-based population) then they will require some form of explanation which does not invoke a balanced polymorphism. A possible hypothesis is that subjects differ in their response to being asked to carry out a skilled task with their non-dominant hand (Beaumont & Kenealy, 1990).
In summary, it is difficult to accept the studies of Annett as providing acceptable evidence for the hypothesis that the intellectual advantage of heterozygotes explains the balanced polymorphism for handedness.