You definitely cannot assume any of those bias assumptions from this article, for sure, but the point is you can't demonstrate that they are not. Maybe the actual experiment controlled for this, but if so the article here didn't mention any of that.