Critical issues in construct validation: Comment on Boyle and Lennon (1994)