Total evidence versus relevant evidence: response to O’Leary et al. (2003)