Journal of the International Neuropsychological Society (2004), 10, 332–341.

Copyright © 2004 INS. Published by Cambridge University Press. Printed in the USA.
DOI: 10.10170S1355617704103019

Assessment of time perception: The effect of aging

MIGUEL COELHO,1 JOAQUIM JOSÉ FERREIRA,1,2 BEATRIZ DIAS,1 CRISTINA SAMPAIO,2

ISABEL PAVÃO MARTINS,1 and ALEXANDRE CASTRO-CALDAS 1
1 Laboratório de Estudos da Linguagem, Centro de Estudos Egas Moniz, Faculdade de Medicina de Lisboa, Hospital Santa Maria,
Av. Egas Muniz, 1649-028 Lisboa, Portugal
2 Instituto de Farmacologia e Terapêutica Geral, Faculdade de Medicina de Lisboa, Hospital Santa Maria, Av. Egas Muniz,

1649-028 Lisboa, Portugal

(Received June 3, 2002; Revised July 28, 2003; Accepted August 26, 2003)

Abstract
Studies concerning time perception lack a validated assessment tool and a consensual “gold-standard” measure.
Moreover, the present evidence suggests modification of timing with aging. This study aimed to develop and
validate a neuropsychological tool to measure time perception and to study temporal perception with aging.
Eighty-six healthy participants, aged 15–90 years old, were asked to verbally estimate and produce empty
intervals signaled by auditory beeps, of 7-, 32-, and 58-s duration. Two tests were used as “gold-standards”:
estimation of the duration of time necessary to draw a clock (“clock time”) and estimation of the duration of
neuropsychological evaluation (“global time”). Results showed a correlation between estimation and production
( p , .01) and a correlation between estimation or production and “global time” ( p , .01). The correlation
between either estimation or production and age ( p , .01), suggested a faster “internal-clock” in the older
participants. However, this finding lost significance when controlled for literacy. The results suggest that these
tests are potentially a useful tool to measure subjective perception of time. They also corroborate the hypothesis
of a change in subjective time perception with aging. It was not possible to conclude if this effect was a specific
result of aging or biased by the interference of literacy. (JINS, 2004, 10, 332–341.)
Keywords: Time perception, Aging, Working memory, Attention

INTRODUCTION explanation for subjective timing also assumes the exis-

Time perception and measure are essential components of
cognition, behavior, and motor performance, representing
one of the basic mechanisms of cerebral function (Artieda
& Pastor, 1996). Accordingly, temporal processing is an
integral part of many everyday goal-oriented behaviors
(Mangels & Ivry, 2001), making it a crucial tool for plan-
ning future actions.
Temporal perception comprises several subjective phe-
nomena including the judgement of subjective duration (Pop-
pel, 1997). The different aspects of psychological time are
interlinked with various attentional, memory, and other cog-
nitive processes (Binkofski & Block, 1996). However, the
Reprint requests to: Miguel Coelho, Laboratório de Estudos da
Linguagem, Centro de Estudos Egas Moniz, Faculdade de Medicina de
Lisboa, Hospital Santa Maria, Av. Egas Muniz, 1649-028 Lisboa, Portu-
gal. E-mail: migcoelho2002@yahoo.es
332
tence of an “internal-clock” mechanism (Binkofski & Block,
1996).
Classically, time perception has been explained by two
different theoretical models: scalar expectancy theory (SET)
(Gibbon et al., 1984) and attentional-gate theory (AG)
(Block, 1990). Both models postulate the existence of a
pacemaker (the “clock”) that generates neuronal pulses and
the existence of a counter accumulating them. This counter
signals when the number of pulses generated reaches a tar-
get value corresponding to a given interval duration (Fortin
& Breton, 1995). The accumulator serves as a working mem-
ory buffer and attention as the gate to working memory.
Concomitant nontemporal tasks interfering with working
memory and0or competing for attention will reduce the ac-
curacy in temporal perception (Fortin et al., 1993). To our
best knowledge there is no data regarding the effect of lit-
eracy on time perception.
Perception of time
Quantifying timing is difficult because there is no vali-
dated method, despite many tools being available (Bindra
& Waksberg, 1956), and it is still lacking a consensual
“gold-standard” test to measure temporal perception. There
is also some evidence that timing changes with age
(Carrasco et al., 2001; Craik & Hay, 1999; Fraisse, 1963),
but to what extent and in what direction is still a matter of
debate. Since timing is dependent on working memory
and attention, it is hypothesized that it should deteriorate
along neuropathological processes that interfere with those
cognitive functions.
The aims of our study were twofold: to develop and test
a neuropsychological tool to measure time perception and
to study time perception along normal aging.
MATERIAL AND METHODS
Research Participants
The study was performed in healthy volunteers, defined as
participants with no known diseases and not taking drugs
known to interfere directly with the central nervous system.
It was intended to include ten patients per decade of life, with
an age spectrum ranging from 15–90 years old. This sample
was not controlled for literacy due to the pragmatic need to
define normative values for the Portuguese population, which
is characterized by a natural higher literacy in younger indi-
viduals; moreover, we were not aware of a previous clear doc-
umentation of a specific effect of literacy on time perception.
Exclusion criteria were a diagnosis of major depression (MINI
International Neuropsychiatry Interview version 4.4 –
Portuguese version) (Sheehan et al., 1998), a diagnosis of
dementia [Mini Mental State Examination (MMSE) vali-
dated for the Portuguese population] (Folstein et al., 1975;
Guerreiro et al., 1994), the presence of any relevant neuro-
logical or psychiatric disease, and the consumption of drugs
with central nervous system effect (benzodiazepines with an
equivalent dose of diazepam less than 10 mg daily were how-
ever accepted). Participants were also excluded if they abused
alcohol, consumed more than four “espresso coffee” cups a
day, and if they were illiterate or unable to understand the
tests. Participants could only be evaluated if they had had a
good night sleep according to their usual sleep pattern. All
participants gave informed consent.
Experimental Tasks
Temporal estimation and production tests were carried out
with a standard Toshiba laptop personal computer and par-
ticipants wore headphones to listen to the auditory stimuli.
The digits on the Digit Span Forward (DSF) and Digit Span
Reverse (DSR) (Garcia, 1984; Wechsler, 1969) tests were
read by the examiner. Participants sat in a comfortable chair,
in a quiet and empty room without any rhythmic sound
(e.g., wall clock) and they were asked to remove their watch
during the test. All participants were evaluated in the after-
noon due to logistical constraints.
The experimental tasks were performed as follows:
333
Verbal estimation and production of a time
interval
This experiment tested the ability to verbally estimate and
produce the duration of empty time intervals signaled by
auditory beeps. The auditory beeps were of 20-ms duration.
A specific software was developed to meet this aim. The
durations of time intervals were of 7, 32, and 58 s for both
estimation and production tests; every duration was re-
peated three times in a pseudorandom order (not allowing
the repetition of the same number in sequence), so that for
both estimation and production tests nine tasks were per-
formed. Participants were not aware they would make nine
trials and they did not know the test involved only three
possible interval durations.
Estimation test. Participants were told that they would
have to verbally estimate the duration of different intervals;
one auditory beep indicated the onset of the interval while
another indicated its end; the beeps were separated by a
silent interval. Participants were instructed to start internal
(“mental”) counting of seconds after listening to the first
auditory beep and to stop counting when they heard the
second beep. They were specifically instructed not to count
aloud nor to perform any digital counting, or use any body
rhythm to help in the estimation. Then they were asked
“how many seconds elapsed between the two beeps?”. The
test proper was preceded by a test run with an interval du-
ration of 4 s. The outcome measure was the ratio between
the estimated duration and the target (7, 32, or 58 s) one.
For each participant, we calculated the mean of the nine
ratios and this value was entered for the group statistical
analysis.
Production test. Participants were told that they would
have to verbally produce different interval durations. In this
task, an auditory beep indicated the onset of the interval
and its end was indicated by the participants, by telling the
examiner when they thought they had reached the target
duration. Participants were asked to start internal counting
of seconds after they heard the first beep and to tell the
examiner when they thought they had reached the target
duration. The examiner wore a chronometer to know how
many seconds had the participant really produced. The test
proper was preceded by a test run with an interval duration
of 4 s. The outcome measure was the ratio between the
produced duration (the value on the examiner’s chronom-
eter) and the target (7, 32, or 58 s) one. For each partici-
pant, we calculated the mean of the nine ratios and this
value was entered for the group statistical analysis.
“Gold-standards”
We assumed as “gold-standards” two tests we accepted as
ecological, which replicate real-life time measurements. In
both cases, the participants were not aware they would be
asked to retrospectively estimate the time elapsed during a
certain task.
334
Clock time. Participants were asked to draw a clock on
a blank sheet of paper. After finishing the drawing, the ex-
aminer asked the participant how much time had elapsed
between the moment he held the pen and finished the draw-
ing. The examiner used a chronometer to measure the du-
ration of the task. The chronometer was started when the
participant picked up the pen and stopped when the partici-
pant placed it on the paper. The outcome measure was the
ratio between the estimated time and the “real” duration
measured by the examiner.
Global time. At the very end of the neuropsychological
evaluation, the examiner asked the participant how much time
had elapsed since they started the current evaluation. The on-
set of the evaluation had been indicated out loud by the ex-
aminer with the sentence: “we are going to start the tests now”
and the examiner specifically referred to that moment when
asking the participant about the elapsed time. The examiner
wore a chronometer to know how much time the evaluation
had lasted. The chronometer was started when the examiner
said the mentioned sentence and stopped just before the ex-
aminer asked the participant about the elapsed time. The out-
come measure was the ratio between the estimated time and
the duration measured by the examiner.
Digit Span Forward and Digit Span Reverse
These tests were used to measure the attention0short-term
and working-memory capacity of the participants. In the
Digit Span Forward, the examiner told the participant a
maximum of seven sequences of digits, each with an in-
creasing number of digits (up to a maximum of nine digits)
and at the end of each sequence the participants were asked
to repeat it in the same order. The Digit Span Reverse was
identical but the participant had to repeat each sequence of
digits in the reverse order. The outcome was the highest
number of digits the participants could repeat correctly. The
digit span was the last test to be performed at the end of the
evaluation.
Statistical Analysis
Estimation test
The outcome measure was the ratio between the estimated
duration and the target (7, 32, or 58 s) duration. For each
participant, we calculated the mean of the nine ratios.
If the ratio was superior to 1, the participant was said to
overestimate, indicating that the hypothetical “internal-
clock” was going faster. If the ratio was inferior to 1, the
participant was said to underestimate, indicating that the
“internal-clock” was going slower.
Production test
The outcome measure was the ratio between the produced
duration (the value on the examiner’s chronometer) and the
target (7, 32, or 58 s) duration. For each participant, we
M. Coelho et al.
calculated the mean of the nine ratios. If the ratio was su-
perior to 1, the participant was said to overproduce, indi-
cating that the hypothetical “internal-clock” was going
slower. If the ratio was inferior to 1, the participant was
said to underproduce, indicating that the “internal-clock”
was going faster.
Clock time
The outcome measure was the ratio between the estimated
time and the duration measured by the examiner.
Global time
The outcome measure was the ratio between the estimated
time and the duration measured by the examiner.
Digit Span
For each participant, we recorded the highest number of
digits the participant could repeat, both in DSF and DSR.
We studied the relation between estimation and produc-
tion tests results and the two “gold-standards”. A sub-
analysis was also performed, studying the relation between
the different time intervals (7, 32, and 58 s) and both clock
and global time.
The age effect on estimation, production, clock time, glo-
bal time, and digit span were studied. The relation between
the digit span and estimation, production, clock time, and
global time was also investigated.
A comparison of estimation, production, clock time, and
global time was made between three predefined age groups
(group A: 15– 40 years old; group B: 41– 64 years old;
group C: 65–90 years old). These age intervals were chosen
according to what we assumed represents different stages
of aging: young, middle-age, and elderly people, respec-
tively. We studied the effect of education and gender in
estimation, production, clock time, and global time. Educa-
tion (number of years in school) was compared between the
three aforementioned age groups. The effect of literacy on
time perception was analyzed as a covariate. The analysis
of metric properties of estimation and production variables
showed they had a near-normal distribution. Although the
size of the sample allowed for parametric tests, we chose to
make use of nonparametric tests in statistical analysis. Re-
lations between variables were studied using nonparamet-
ric tests for correlations (Spearman test), and comparisons
of variables between the three age groups were made using
nonparametric tests of variance (Kruskal-Wallis test).
RESULTS
Ninety-seven participants were screened to participate in
this study. Seven participants were excluded due to diagno-
sis of dementia according to MMSE (4 participants) and
diagnosis of major depression according to the MINI Inter-
national Neuropsychiatry Interview (3 participants). One
participant was excluded because she was not able to un-
Perception of time
Table 1. Demographics
Age groups (years) 15–20 21–30
N (total 5 58) 10 12
F0M 109 1101
Education (years) mean 11.8 17.3
Note. N 5 number of participants; F 5 female; M 5 male.
derstand the test instructions. Three individuals did not com-
plete the evaluation due to tiredness, and their results were
not included in the analysis. The final sample was of 86
participants and the data of these participants were the ba-
sis for the analysis. Their age ranged from 15–90 years old.
There were 58 female and 28 male participants. There were
roughly ten participants per decade of life (Table 1).
Overall, the verbal estimation and production tests were
easy to explain and perform.
Concerning the metric properties of estimation and pro-
duction, the median value for estimation was 1.081 (Min
.2128 and Max 6.3988) and the median value for produc-
tion was .8224 (Min .3382 and Max 1.788), with a statisti-
cal significant negative correlation between estimation and
production ( p , .01; r 5 2.5509).
A positive correlation was found between estimation and
global time ( p , .01; r 5 .0712), and a negative correlation
between production and global time ( p , .01; r 5 2.1591).
A correlation of estimation or production and clock time
was not demonstrated, as well as a correlation between clock
and global time. The different time intervals (7, 32, and
58 s), in both estimation and production tests, were all cor-
related with global time but not with clock time, and the
power of correlation was similar between them.
335
31– 40 41–50 51– 60 61–70 71–80 81–90
10 10 11 10 14 9
802 703 1001 505 905 702
14 12.2 8.8 7.1 7.0 7.5
A positive correlation was found between estimation and
age, suggesting that older participants had faster “internal-
clocks” ( p , .01; r 5 .2978) (Figure 1a). Conversely, a
negative correlation was found between production and age,
also indicating an association between older participants
and faster “internal-clocks” ( p , .01; r 5 2.2731) (Fig-
ure 1b). We could not find a correlation between age and
either clock time or global time. All previous analysis lost
statistical significance when controlled for literacy. Age was
negatively correlated with education ( p , .01; r 5 2.6).
The results showed a negative correlation between educa-
tion and estimation ( p , .01; r 5 2.3951) and global time
( p , .05; r 5 2.2521), and a positive correlation between
education and production ( p , .01; r 5 .3612), suggesting
that participants with lower level of education had faster
“internal-clocks”.
We compared the difference in performance between three
predefined age groups (group A: 15– 40 years old, mean
26.84, SD (standard deviation) 7.21; group B: 41– 64 years
old, mean 50.72, SD 4.93; group C: 65–90 years old, mean
75.84, SD 7.18; 95% confidence interval). When estimation
was compared between the age groups there was a trend
toward overestimation with older age, suggesting accelera-
tion of “internal-clocks” with age (Figure 2a). The same
Fig. 1a. Correlation of age with time es-
timation. Age was positively correlated
[ p , .01 (Spearman test r 5 .29)] with
the estimation of time, suggesting an ac-
celeration of the “internal-clock” with
aging.
336
analysis about production showed a trend toward underpro-
duction with age, also suggesting acceleration of “internal-
clock” with age (Figure 2b).
The performance on estimation, production, clock time,
and global time tasks was not significantly different be-
tween female and male participants.
As expected from the literature, we found a negative cor-
relation between DSF0DSR and age ( p , .01; r 5 2.4715,
for DSF; p , .01; r 5 2.4658, for DSR) (Figure 3).
The results showed a negative correlation between DSF0
DSR and estimation ( p , .01; r 5 2.3398, for DSF; p ,
M. Coelho et al.
Fig. 1b. Correlation of age with time pro-
duction. Age was negatively correlated
[ p , .01 (Spearman test r 5 2.27)] with
the production of time, suggesting an ac-
celeration of the “internal-clock” with
aging.
.05; r 5 2.2434 for DSR) (Figure 4a). There was a positive
correlation between production and DSF0 DSR ( p , .01;
r 5 .3286, for DSF; p , .01; r 5 .3902, for DSR) (Figure 4b).
DISCUSSION
The results of this study showed an internal consistency
between the performance on estimation and production tests,
and found a correlation between those tasks and one of the
chosen “gold-standards” tests, suggesting that our assess-
ment tool is a potentially good measure of time duration.
Fig. 2a. Comparison of mean values of
time estimation between three age groups.
There was a trend towards overestima-
tion of time with aging.
Perception of time 337

Fig. 2b. Comparison of mean values of

time production between three age groups.
There was a trend towards underproduc-
tion of time with aging.

The results also showed a change of estimation and produc-
tion of time intervals with aging, suggesting an association
of aging with a faster “internal-clock”.
The extensive literature on time perception often differ
on the methods used to measure temporal perception. How-
ever, none of these methods has yet been properly validated
in a normal population, casting doubt on the theoretical
models based on their results. Also, as recognized by Allan
(1979) and McConchie and Rutschmann (1971), no single
method can claim a consistent superiority over the others
and the correlation between them is either poor or absent.
The aim of this study was to develop a neuropsycholog-
ical tool to measure subjective duration and, secondly, to
study temporal perception along normal aging. Firstly, we
tried to overcome the lack of a “gold-standard” test, that is,
a test that reliably measures what we want it to measure–
subjective time perception. Thus, the first step was to iden-
tify a real-life measure against which our tests could be

Fig. 3. Correlation of Digit Span Forward with age [ p , .01 (Spearman test r 5 2.47), left figure] and Digit Span
Reverse with age [ p , .01 (Spearman test r 5 2.46), right figure] showing a decrease performance with aging.
338 M. Coelho et al.
Perception of time 339

compared. The “gold-standard”, which is in any case rela-
tive, should be as feasible and as close as possible a mea-
sure of temporal perception of everyday life, that is to say,
a pragmatic measure of the processing of temporal infor-
mation by the brain in real-life conditions. We assumed as a
“gold-standard” for subjective duration the “clock time”
test and the “global time” test. In both tests, the participants
performed tasks without knowing they would be asked to
estimate the time elapsed during their performance. They
made retrospective evaluations, not paying attention to the
passage of time, instead allocating attention towards the
tasks themselves. We believe they inferred duration from
contents of memory, memory for events, number of transi-
tions between events, and expectation for duration of events
(Block & Zakay, 1997; Ornstein, 1969). Thus, retrospec-
tive timing seems to depend mostly on long-term memory
and contextual events (Mangels & Ivry, 2001; Zakay, 1990).
On the other hand, during the tests of verbal estimation and
production, participants knew they should actively monitor
the passage of time. Their prospective measurement was
mostly dependent on attention, mostly attention to the vari-
able “time”, and working memory (Mangels & Ivry, 2001;
Zakay, 1990).
The evidence that retrospective and prospective timing
depend upon different cognitive processes may explain the
weak association we found of estimation and production
with “global time” and the lacking association of estima-
tion and production with “clock time”. Besides, an ecolog-
ical measure of time perception should have some utility
for the participant, acting as a motivational drive. While the
“global time” measure was surely being subjectively as-
sessed for its duration, the “clock time” measure had little
usefulness or meaning for the participants of the experi-
ment. This may explain its lack of correlation with more
conscious measures of time perception. Possibly, this dif-
ference in usefulness between the “global time” and the
“clock time” measures also explains the lack of association
between each other. An alternative explanation could be
their difference in magnitude duration (the “global time” in
the range of 20–25 min while the “clock time” in the range
of a few seconds). In future experiments, we plan to com-
pare the estimation and production measures with ecologi-
cal “gold-standards” that require prospective timing and
more meaningful time intervals, eventually increasing the
strength of their statistical association; moreover, if more
than one “gold-standard” is used, they should be of same
magnitude and duration. Interestingly, when comparing es-
timation, production, “global time”, and “clock time” be-
tween the three age groups, we found the same trend towards
acceleration of the “internal-clock” in older participants,
suggesting that prospective and retrospective timing may
share common cognitive mechanisms.
We found an association of estimation with production,
which gives internal consistency to our results, and allows
speculating about a common “internal-clock” subserving
estimation and production (Ivry & Hazeltine, 1995; Keele
et al., 1985).
This study showed an age effect on Digit Span, and a
change of estimation and production with aging when com-
pared between the three predefined age groups. This trend
was towards an acceleration of “internal-clock” in older
participants. A smaller study performed by Carrasco and
coworkers (2001) found similar results when comparing
time estimation between young and elderly healthy partici-
pants. The study consisted of 25 adults, and the mean age of
the young (26.15 years) and the elderly groups (79.1 years)
were similar to that of groups A and C in our study. The
methodology and the time-interval duration (10 s) were very
similar to ours. Surprisingly, both results are in contrast
with a previous study that found an association of older age
with slower “internal-clock” (Craik & Hay, 1999). In that
study, the interval durations were of the same range (30, 60,
and 120 s) as ours, but the paradigms of estimation and
production differed slightly, namely, participants per-
formed the estimation and production tests while perform-
ing other tasks at different levels of complexity, and this
discrepancy in methodology can make comparison diffi-
cult. Data concerning age and short time intervals is still
scanty and inconclusive though (Nichelli, 1993). A study
from Fraisse (1963) showed an association of aging with
acceleration of “internal-clock” regarding large units of time
such as days or years; however, people estimates of inter-
vals in range of hours may be related to sleep–wake cycles
(Artieda & Pastor, 1996).
The effect of education on cognition is well established,
although we are not aware of any clear documentation of a
specific effect of literacy on time perception. We found a
correlation of education with estimation, production and
“global time”, suggesting faster “internal-clocks” in partici-
pants with lower education. However, as it was anticipated
from the characteristics of the Portuguese population and
as our sample was not purposefully controlled for literacy,
older participants were less literate. The correlation be-
tween either estimation or production and aging lost signif-

Fig. 4a. Correlation of time estimation with Digit Span Forward [ p , .01 (Spearman test r 5 2.34), left figure] and
time estimation with Digit Span Reverse [ p , .05 (Spearman test r 5 2.24), right figure] showing that overestimation
of time is associated with a decrease performance in tests of attention and short-term memory0working memory.

Fig. 4b. Correlation of time production with digit span forward [ p , .01 (Spearman test r 5 .39), left figure] and time
production with Digit Span Reverse [ p , .01 (Spearman test r 5 .33), right figure] showing that underproduction of
time is associated with a decrease performance in tests of attention and short-term memory0working memory.
340
icance when education was analyzed as a co-variate. The
absence of a statistical difference of years of education be-
tween the three predefined age groups argues against a pri-
mary effect of education on our results, suggesting that the
association of older participants with faster “internal-
clocks” derives mainly from aging. Nevertheless, it is not
possible to firmly conclude if the association of older par-
ticipants with faster “internal-clocks” is a specific result of
aging or biased by the interference of literacy. As the de-
sign of the present study was not intended to look for an
effect of education on timing, future studies must be con-
ducted specifically to address this interesting question.
Our results demonstrated an association between estima-
tion and production and both DSF and DSR, while there was
no association between DSF and DSR, and either both “clock
time” or “global time”. This adds further evidence regarding
the role of attention0short-term and working memory in pro-
spective timing but not in retrospective timing (Block, 1992;
Block & Zakay, 1997; Brown, 1997; Mangels & Ivry, 2001).
Nevertheless, the association of estimation and production
with Digit Span may result from an effect of age on both tasks.
The performance was similar between female and male
participants in all tasks, though a disproportionate number
of females were present in our sample.
Participants were explicitly asked to perform the estima-
tion and production tests by internal counting of seconds
and the intention was to have them producing their idea of
1 s. The strategy of internal counting has been considered
to increase the accuracy of estimation (Davis, 1962; Fraisse,
1984; Wearden, 1991). Counting was also aimed at giving a
homogeneous activity for all participants during estimation
and production tasks (Rubia et al., 1997).
Physiopathologic speculations may be too premature, be-
cause replication of our results is still needed. Nevertheless,
and according to present theoretical models (Block, 1990;
Gibbon et al., 1984), the observed deficit in time estimation
could derive from a change in the speed of the “internal-
clock”, which has been associated with several conditions
(Binkoski, 1996; Carrasco et al., 2000; Meck, 1996; Pastor
et al., 1992). Alternatively, a deficit in attention and working
memory associated with aging could disrupt the counter sys-
tem that accumulates the pulses generated by the “internal-
clock”. Studies demonstrating deficits in time perception in
Alzheimer and Parkinson disease (Carrasco et al., 2000; Pas-
tor et al., 1992; Riesen & Schnider, 2001) suggest the possi-
ble involvement of neurotransmitters such as acetylcholine
and dopamine in time perception.
Our results showed an acceleration of “internal-clock”
with aging, and also suggested a role for attention and short-
term0working memory on prospective timing. The assess-
ment tool presented was easy to perform and was related to
one ecological measure of time perception. Nevertheless,
and due to its moderate association with the “gold-standards”
and possible bias of results due to literacy, future well-
designed studies are needed to replicate the present find-
ings, to analyze its interobserver and intrasubject validity,
and to confirm its potential usefulness. If the present find-
M. Coelho et al.
ings are confirmed, they must be explained in the light of
current theoretical models for time perception.
ACKNOWLEDGMENTS
This study was supported by Grant 59098 from Fundação Bial,
Portugal.
REFERENCES
Allan, L.G. (1979). The perception of time. Perception and Psy-
chophysics, 26, 340–354.
Artieda, J. & Pastor, M.A. (1996). Time, internal clocks and move-
ment. In G.E. Stelmach & P.A. Vroon (Eds.), Advances in psy-
chology, Vol. 115 (pp. 1–27). Amsterdam: Elsevier.
Bindra, D. & Waksberg, H. (1956). Methods and terminology in the
studies of time estimation. Psychological Bulletin, 63, 155–159.
Binkofski, F. & Block, R.A. (1996). Accelerated time experience
after left frontal cortex lesion. Neurocase, 2, 485– 493.
Block, R.A. (1990). Models of psychological time. In R.A. Block
(Ed.), Cognitive models of psychological time (pp. 1–35). Hills-
dale, New Jersey: Lawrence Erlbaum Associates.
Block, R.A. (1992). Prospective and retrospective duration judg-
ment: The role of information processing and memory. In F.
Macar & V. Pouthas (Eds.), Time, action and cognition: To-
wards bridging the gap (pp. 141–152). Dordrecht, Nether-
lands: Kluwer Academic.
Block, R.A. & Zakay, D. (1997). Prospective and retrospective
duration judgments: A meta-analytic review. Psychonomic Bul-
letin and Review, 4, 184–197.
Brown, S.W. (1997). Attentional resources in timing: Interference
affects in concurrent temporal and nontemporal working mem-
ory tasks. Perception and Psychophysics, 59, 1118–1140.
Carrasco, M.C., Guillem, M.J., & Redolat, R. (2000). Estimation
of short temporal intervals in Alzheimer’s disease. Experimen-
tal Aging Research, 26, 139–151.
Carrasco, M.C., Bernal, M.C., & Redolat, R. (2001). Time esti-
mation and aging: A comparison between young and elderly
adults. International Journal of Aging and Human Develop-
ment, 52, 91–101.
Craik, F.I.M. & Hay, J.F. (1999). Aging and judgments of dura-
tion: Effects of task complexity and method of estimation. Per-
ception and Psychophysics, 61, 549–560.
Davis, R. (1962). L’estimation d’intervalles temporels en comp-
tant plus ou moins rapidement. Anné Psychologique, 62, 29– 44.
Folstein, M.F., Folstein, S.E., & McHugh, P.R. (1975). Mini-
Mental State. A practical method for grading the cognitive state
of patients for the clinician. Journal of Psychiatry Research,
12, 189–198.
Fortin, C. & Breton, R. (1995). Temporal interval production and
processing in working memory. Perception and Psychophys-
ics, 57, 203–215.
Fortin, C., Rousseau, R., Bourque, P., & Kirouac, E. (1993). Time
estimation and concurrent nontemporal processing: Specific
interference from short-term memory demands. Perception and
Psychophysics, 53, 536–548.
Fraisse, P. (1963). The psychology of time. New York: Harper.
Fraisse, P. (1984). Perception and estimation of time. Annual Re-
view of Psychology, 35, 1–36.
Garcia, C. (1984). A Doença de Alzheimer. Problemas de diagnós-
tico clínico. Dissertação de Doutoramento. Lisboa: Faculdade
de Medicina de Lisboa.
Perception of time
Gibbon, J., Church, R.M., & Meck, W. (1984). Scalar timing in
memory. In J. Gibbon & L. Allan (Eds.), Annals of the New
York Academy of Sciences, Vol. 423: Timing and time percep-
tion (pp. 52–77). New York: New York Academy of Sciences.
Guerreiro, M., Silva, A.P., Botelho, M.A., Leitão, O., Castro-
Caldas, A., & Garcia, C. (1994). Adaptação à População Por-
tuguesa da tradução do “Mini Mental State Examination”
(MMSE). Revista Portuguesa de Neurologia, 1, 9–10.
Ivry, R. & Hazeltine, R.E. (1995). Perception and production of
temporal intervals across a range of duration: Evidence for a
common timing mechanism. Journal of Experimental Psychol-
ogy: Human Perception and Performance, 21, 3–18.
Keele, S.W., Pokorny, R.A., Corcos, D.M., & Ivry, R. (1985). Do
perception and motor production share common timing mech-
anisms: A correlational analysis. Acta Psychologica, 60,
173–191.
Mangels, J.A. & Ivry, R.B. (2001). Time perception. In B. Rapp
(Ed.), Handbook of cognitive neuropsychology: What deficits
recall about the human mind, Chapter 19 (pp. 467– 493). Hove:
Psychological Press.
McConchie, R.D. & Rutschmann, J. (1971). Human time estima-
tion: On difference between methods. Perceptual and Motor
Skills, 32, 319–336.
Meck, W.H. (1996). Neuropharmacology of timing and time per-
ception. Cognitive Brain Research, 3, 227–242.
Nichelli, P. (1993). The neuropsychology of human temporal in-
formation processing. In F. Boller & J. Grafman (Eds.), Hand-
book of neuropsychology, Vol. 8 (pp. 339–371). Amsterdam:
Elsevier.
341
Ornstein, R.E. (1969). On the experience of time. Harmond-
sworth, England: Penguin Books.
Pastor, M.A., Artieda, J., Jahanshahi, M., & Obeso, J. (1992). Time
estimation and reproduction is abnormal in Parkinson’s dis-
ease. Brain, 115, 211–225.
Poppel, E. (1997). A hierarchical model of temporal perception.
Trends in Cognitive Sciences, 1, 56– 61.
Riesen, J.M. & Schnider, A. (2001). Time estimation in Parkin-
son’s disease: Normal long duration estimation despite im-
paired short duration discrimination. Journal of Neurology, 248,
27–35.
Rubia, K., Schuri, U., Cramon, D.Y.V., & Poppel, E. (1997). Time
estimation as a neuronal network property: A lesion study. Neu-
roReport, 8, 1273–1276.
Sheehan, D.V., Lecrubier, Y., Sheehan, K.H., Amorim, P., Janavs,
J., Weiller, E., Hergueta, T., Baker, R., & Dunbar, G.C. (1998).
The Mini-International Neuropsychiatric Interview (M.I.N.I):
The development and validation of a structured diagnostic psy-
chiatric interview for DSM-IV and ICD-10. Journal of Clini-
cal Psychiatry, 59, 22–33; quiz 34–57. Adapted to Portuguese
(Portugal) by T. Guterres.
Wearden, J.H. (1991). Do humans possess an interval clock with
scalar timing properties? Learning and Motivation, 22, 59–83.
Wechsler, D.A. (1969). Manuel de I’èchelle clinique de mémoire.
Paris, France: Centre de Psychologie Appliquée.
Zakay, D. (1990). The evasive art of subjective time measure-
ment. In R.A. Block (Ed.), Cognitive models of psychological
time (pp. 59–84). Hillsdale, New Jersey: Lawrence Erlbaum
Associates.