Weaknesses in reports of "fertility" for horses and other species.

Apparent fertility of a male or group of females is considered frequently by veterinarians or animal scientists. Unfortunately, concepts of experimental design and statistics impacting validity and interpretation of values for average pregnancy rate frequently are ignored. The magnitude of this problem was documented by examination of published papers; 51 of 67 (76%) were considered flawed for one or more reasons. The discussion considers why: (a) conclusions from most published fertility studies reporting no significant difference due to treatment(s) are suspect, because too few males and/or females were used; (b) the experimental unit in an IVF study should be a droplet of co-incubating gametes rather than an ovum; (c) apparent fertility of a male is profoundly influenced by the range in actual fertility of the females with which he was mated, and thus might shift over a two- to three-fold range; and (d) scientists should refrain from conduct of studies destined to be inconclusive, and should be candid in reporting each fertility trial. It was emphasized that no fertility data were better than an imprecise average value for a given male or a conclusion based on an inadequate research study or incomplete report of what was done.