Method variance as an artifact in self-reported affect and perceptions at work: Myth or significant problem?