Alternate‐choice test items: Implications for measuring clinical judgment