Beyond standard quality control protocols administered in the wet lab, genotypic data requires the same sort of pre-analysis cleaning procedures that social scientists would apply to any phenotypic or environmental data. For example, one would want to assess whether willingness to provide a DNA sample is systematically associated with the non-genetic variables of interest (i.e., are the participants who gave DNA different in any way). Likewise, it is important to verify that the values corresponding to each of the alleles reflect viable nucleotide bases; that is, at a locus polymorphic for adenine and guanine, none of the study participants should have a value representing cytosine. In addition to these generic practices, genotypic data should also be screened to determine whether the observed distribution of genotypes differs from that which is expected, given the allelic frequencies observed within a given population – in other words, do the genotypes in your sample deviate from Hardy–Weinberg equilibrium (HWE)? In the simple case of a bi-allelic variant, like a SNP, the expected distribution of genotypes is easily derived via a Punnett square, and HWE
