In the Eye of the Beholder: The Effect of Rater Variability and Different Rating Scales on QTL Mapping