Delay discounting of real and hypothetical rewards III: Steady-state assessments, forced-choice trials, and all real rewards