As someone who does research in this area, broadly speaking (behavioral individual differences and genetics), I can't help but wonder how much rater effects are playing a role, or washing out findings. Everything it seems is being studied through the lens of human raters, which presumably would add variance and also might not map onto actual dog behavioral differences very well (on the other hand, if the interest is in human experiences of dog behavior, maybe it is on target).<p>I've been reading through the paper though and there's a lot there so maybe I just missed something.