I've read PhDs that by undertaking different experiments have come to conclusions diametrically opposite each other.

Either there is a massive amount of bias, or small studies can't be relied on.