The use of different types of partial temporal information is shown to
affect dating accuracy and the distribution of errors in event dating
. Several different types of partial temporal information are discusse
d, but three are highlighted by the data. Specifically, subjects' dati
ng error patterns suggest that they (1) use different types of within-
week information, (2) use recalled event sequences, and (3) use bounda
ry landmarks to report the dates of events. In general, these data sug
gest that although precise temporal information is sometimes represent
ed in the memory trace for an event, more often the date-related infor
mation is inferred from other aspects of memory.