Dropouts, spells of time, and credits in postsecondary education: Evidence from longitudinal surveys☆