The correction: Science 1995 268 1553 Quantitative trait locus for reading disability: correction. Cardon et al. state that 4 MZ twins had been incorrectly included with the fraternal twins. The probability of linkage after removal of these MZ twins falls from p = 0.0003 to p = 0.0094 in the unselected twins and to p = 0.037 in the extreme pairs.
The new study: Reading Deficits on Chromosome 6p To appear in American Journal of Human Genetics
Much thanks to Andrew Paterson at the University of Toronto for bringing these to my attention.