Participants were asked to view a film, and at the same time, to perform a
series of interfering actions at the end of several successive period durat
ions (d). Any delays in executing the required actions were measured, but t
he main focus was the distribution of time checks over each of the periods.
Two small groups, whose mean ages were 22 and 66 years, took part in the s
tudy. The overall time-monitoring strategy was found to conform to the foll
owing relation: c = 2 exp (t-i)/s, where c was the rank of a time check, an
d t the total time elapsed thus far. According to this function, after some
initial waiting time (i), the checks occurred at gradually narrower time i
ntervals, each check taking place after a constant proportion s/(d-i) of th
e time left. While most participants in both groups followed this rule, som
e of them exhibited stronger or weaker forms of the strategy, i.e., forms w
here the rate of time checking increased either faster or more slowly than
the standard function predicts. The weaker forms tended to result in poorer
prospective recall performance, but no major differences were found betwee
n the age groups.