Prospective and retrospective time estimates as a function of clock duration.

12 independent groups of 18 subjects each estimated the duration of one of six 'empty' intervals (10, 18, 26, 34, 42, and 50 sec.). Subjects were told of the time task either prior to (prospective paradigm) or after (retrospective paradigm) the presentation of the interval. The results are consistent with Hicks' 1992 finding of an interaction between temporal paradigm and interval duration. Whereas the shorter intervals (< 42 sec.) were estimated accurately in both the retrospective and the prospective paradigms, the longer intervals were estimated rather accurately in the prospective paradigm and underestimated in the retrospective paradigm.