Age, Personal Characteristics, and the Speed of Psychological Time

Abstract

Adults often report the impression that time seems to pass more and more quickly as they get older. The purpose of this study is to identify how individual characteristics relate to this impression of acceleration. To do so, 894 participants aged 15 to 97 completed a questionnaire that surveyed sociodemographic characteristics, impulsivity, anxiety, personality, and relation to time. They also indicated how fast different lapses of time seemed to have passed: yesterday, the past week, the past month, the past year, the past three years, the past five years, and the past 10 years. For each period, except for one year, adolescents found that time passes more slowly than participants from older groups (18–29 years, 30–59 years, and 60 years and over). A composite score for all these periods also indicates that female participants found that time passes more rapidly than males. However, a multiple linear regression analysis reveals that the variables that best predict the impression that time passes faster as we get older are high anxiety, the belief in the phenomenon of temporal compression, as well as conscientiousness and agreeableness personality traits, with other factors explaining little variance. These results add further weight to the impression that time seems to pass more quickly as we age, but also indicate that other variables than age play a critical role in explaining this impression.

Full text

*To whom correspondence should be addressed. E-mail: simon.grondin@psy.ulaval.ca
Age, Personal Characteristics, and the Speed of Psychological
Time
Audrey-Anne Gagnon-Harvey, Jamie McArthur, Émie Tétreault,
Daniel Fortin-Guichard and Simon Grondin*
École de psychologie, Université Laval,  rue des Bibliothèques, Québec, QC, Canada GV A
Received  October ; accepted  December 
Abstract
Adults often report the impression that time seems to pass more and more quickly as they get old-
er. The purpose of this study is to identify how individual characteristics relate to this impression
of acceleration. To do so,  participants aged  to  completed a questionnaire that surveyed
sociodemographic characteristics, impulsivity, anxiety, personality, and relation to time. They also
indicated how fast diferent lapses of time seemed to have passed: yesterday, the past week, the past
month, the past year, the past three years, the past ve years, and the past  years. For each period,
except for one year, adolescents found that time passes more slowly than participants from older
groups (– years, – years, and  years and over). A composite score for all these periods
also indicates that female participants found that time passes more rapidly than males. However, a
multiple linear regression analysis reveals that the variables that best predict the impression that
time passes faster as we get older are high anxiety, the belief in the phenomenon of temporal com-
pression, as well as conscientiousness and agreeableness personality traits, with other factors explain-
ing little variance. These results add further weight to the impression that time seems to pass more
quickly as we age, but also indicate that other variables than age play a critical role in explaining this
impression.
Keywords
Speed of psychological time, time perception, individual diferences, aging
Timing & Time Perception  () –
© Koninklijke Brill NV, Leiden,  DOI: ./-bja


.Introduction
Time and its estimation are central to the way humans interpret and organize
many aspects of their life and environment. While humans possess internal mech-
anisms that account for high levels of timing precision, certain stimulation condi-
tions (like the physical characteristics of stimuli that bound time intervals) and
pathologies (for instance schizophrenia, attention decit, or anxiety) may disturb
this timing capability (Grondin, ). Along with these particular conditions
that may impair timing precision, the literature on time perception also contains
some reports of the common belief that, for humans, time seems to go faster as
they get older. The present paper is a new attempt to verify if this impression holds
true and tries to uncover what personal features may account for it.
There are several hypotheses that might explain why time seems to pass more
rapidly for humans as they get older (Friedman & Jansen, ). One explanation
is referred to as the ratio theory, by which each time segment represents a smaller
portion of our life as we get older; it is not a period per se that counts, but this
period in contrast to the rest of our lives (Lemlich, ). Therefore, as we age,
the years of life ahead of us reduce while the portion of life passed increases, and,
according to the ratio theory, it is the contrast between these two periods that cre-
ates an impression of acceleration. Another hypothesis is related to the number
of memorable events that occur at diferent times in life. While many new events
are experienced during childhood, this number decreases over the years. As we
get older, more things are done on a routine basis and we are exposed to less
novelty. Given that an interval is estimated to be longer if it contains more events
and details that can be recalled (Block, ; Block & Zakay, ; Ma & Cheung,
), it is reasonable to believe that, as we get older, time intervals lled with
less memorable events will lead to an impression that these intervals seem to
pass rapidly. Also, if memory declines with age (Grady & Craik, ; Salthouse,
), then impressions that time passes rapidly should increase with age as we
recall less details. Another hypothesis is related to a phenomenon called forward
telescoping, which is the tendency to underestimate how long ago some events
have occurred. If someone often experiences the impression that an event has oc-
curred much longer ago than it has, this should lead to an overall impression that
time is passing quickly. As we get older, there is more room for such impressions,
considering that the number of much older events increases in our lives and that
the magnitude of the underestimation may also increase. Finally, the impression
that time passes more rapidly as we get older may also depend on the amount of
temporal pressure in daily life. People often underestimate the time required to
complete a task (Roy & Christenfeld, ; Roy et al., ) and, consequently, of-
ten nd themselves in situations where they run out of time. This situation might
cause the impression that time has passed quickly. According to this viewpoint,
A.-A. Gagnon-Harvey et al. / Timing & Time Perception  () –


it is when people are experiencing the highest amount of temporal pressure that
they are more inclined to nd that a given time period has passed rapidly (Jans-
sen et al., ).
Even though the common belief that time passes more rapidly as we get old-
er is widespread, the empirical demonstration of the link between age and the
speed of time is not that robust. Wittmann and Lehnhof () report data from
- to -year old participants that support this belief for a -year interval, but
they note that age accounts for roughly only % of the variance. Flaherty and
Meer () report data that are complementary to Wittmann and Lehnhof’s
ndings. Their participants had to report their perception of how three periods
of time seemed to have passed: yesterday, the past month, and the past year.
They observed that the impression that time has passed faster is stronger for the
past year item than for the past month item, which is itself stronger than for
the yesterday item. They explained their results by a loss of memory over time.
Even more important is the fact that in their data, there was a signicant non-
linear age efect. The impression that time passes more rapidly was higher in the
middle-aged adult group than in the young adult and elderly groups. This nd-
ing is consistent with a temporal pressure hypothesis, which posits that middle-
age adults are busier and have more to do on a daily basis than young or elderly
adults. This nding is also consistent with the idea that the level of engagement
has an impact on the perception of speed of time (Larson, ; Larson & von
Eye, ).
The overall empirical picture is even more blurred if we look at the data re-
ported by Friedman and Janssen (). They asked their participants (aged from
 to ) how fast the previous week has passed, the previous month, the previ-
ous year, and the past  years. For this series of questions regarding the speed of
time, there were not many age diferences. Indeed, only the question relative to
the past  years revealed a signicant age diference, with older adults perceiving
this interval as having passed faster. Indeed, only the question relative to the past
 years revealed a signicant age diference, with older adults perceiving this
interval as having passed faster. This holds true even when assessed with Japanese
participants (Janssen et al., ), with German-language samples (Wittmann &
Lehnhof ; Wittmann et al., ), and with people from The Netherlands
and New Zealand (Friedman & Janssen, ).
In addition to memory and time pressure, the scientic literature indicates
that there are other characteristics that could be linked to the subjective impres-
sion of the speed of time. First, gender might be an important factor to consider.
Women tend to remember a greater number of past events than men (Pillemer et
al., ). If, as previously noted, the more memories or the more details can be
recalled for a certain period of time, the longer it will seem to have lasted, then the
impression that time passes rapidly should be weaker for women than for men.
A.-A. Gagnon-Harvey et al. / Timing & Time Perception  () –


Gender diferences are even more plausible considering that research has also
shown a gender efect on the estimation of short (Block et al., ) and even very
long (Hancock, ) time intervals (see Grondin, ).
Secondly, anxiety, which can be conceptualized as the fear of something that
is in the future (Craske et al., ; Krauss & Ruiz, ), could also inuence the
subjective speed of time. For instance, when asked to estimate the duration of a
waiting period, participants with higher levels of anxiety reported longer dura-
tions than less anxious participants (Sarason & Stoops, ). Fear and anticipa-
tion, both being conceptually related to anxiety, can also be linked to an increase
in perceived duration (Droit-Volet et al., ; Schif & Thayer, ; Stetson et al.,
; Watts & Sharrock, ). In addition, on a larger time scale, when assessing
speed of time, Lamotte and colleagues () reported a signicant correlation
between level of anxiety (assessed with the STAI questionnaire) and the response
to the question “How do you think that time ies by?”: the higher the anxiety level,
the slower time seems to pass by.
Thirdly, impulsivity, dened as the tendency to respond quickly to stimuli with-
out evaluating possible consequences (Gerbing et al., ), might also inuence
the subjective speed of time. Impulsive people can have a tendency to overes-
timate the duration of a time interval (Berlin et al., ; Wittmann & Paulus,
), and children with attention decit hyperactivity disorder (a disorder in
which impulsivity can be a central trait) may have a poorer overall sense of time
(see Grondin, ).
Taken together, these potential factors, in addition to advancement in age, il-
lustrate that the impression that time passes more or less quickly constitutes a
multifaceted concept worth investigating.
..The Present Study
A rst objective of the present study is to verify whether a person’s age has an im-
pact on their impression that time, over long intervals, passes more or less quick-
ly. To do so, we replicate the approach used by Flaherty and Meer () where
they compared the subjective impression of the speed of time across diferent
age groups. People from four age groups (–; –, –, and  or more)
completed a questionnaire surveying their impression that time passes more or
less quickly, as well as their beliefs towards this impression, anxiety, subjective
memory, occupational level, impulsivity and personality. A second objective is to
document which personal characteristics are involved in the impression that time
passes more or less quickly. Several authors have demonstrated an efect of vari-
ous personal characteristics on temporal estimates, but these studies have mainly
been carried out with short intervals. This investigation will help determine if the
same efects also apply to longer intervals, and, if so, to what extent each of them
predicts the impression that time passes more or less rapidly.
A.-A. Gagnon-Harvey et al. / Timing & Time Perception  () –


.Method
..Participants
Originally,  participants were recruited for this study. These participants in-
clude men and women of all ages and from diverse socio-economic backgrounds.
Participants were contacted via the email list of Université Laval and via social
media. Adolescents were recruited from secondary schools in the Quebec region.
Also, in order to recruit more people over the age of , paper questionnaires were
distributed in senior-citizen clubs, community organizations, and independent-
and assisted-living facilities for seniors in the Quebec region. No exclusion criteria
were established, apart from the minimum age of  years.
A total of  participants were excluded from analyses due to missing values
for at least one question. Although this may seem like a large number, it is impor-
tant to note that many participants would begin completing the online question-
naire and abandon after the rst few questions. In other cases, some participants
would omit (or forget) certain answers in a seemingly random way. Since many of
the analyses used in this study rely on the summation of questions, it was neces-
sary to remove these participants as a score of  would considerably skew their
results. Moreover, the Mahalanobis distance was calculated with a critical thresh-
old of .%, χ2crit() = .. This measure allows the identication of rare ob-
servations and indicates that four participants in the study are considered rare.
These participants were removed from the analysis. Also, three participants that
had identied their sex as ‘other’ were removed from the analysis as the frequency
of this category accounts for less than .% of the sample (Tabachnick & Fidell,
).
The nal sample therefore included  participants with a mean age of 
and a median age of . The youngest participant was  years old and the old-
est . Participants were assigned to one of four age groups (see Table ). The age
groups were established based on criteria used in the studies of Flaherty and Meer
().
..Materials
Participants provided their answers to the self-administered questionnaire either
via an online platform named Portail Intégré d’Applications Numériques pour Ordi-
nateur (PIANO), or directly on the paper version. The PIANO platform allowed
the transfer of data into an Excel le, which was then used to compile the paper
questionnaires.
...Sociodemographic Prole
Participants were rst asked to complete a short demographic prole in order to
identify their personal characteristics (sex and age).
A.-A. Gagnon-Harvey et al. / Timing & Time Perception  () –


...Functional and Dysfunctional Impulsivity
The validated French version of Dickman’s Functional and Dysfunctional Impulsiv-
ity Questionnaire (Caci et al., ) was used to measure impulsivity. Of the 
items included in this questionnaire,  measure functional impulsivity ( rep-
resenting a low level of functional impulsivity and  representing a high level of
functional impulsivity), and  measure dysfunctional impulsivity (with scores
of  and  representing low and high dysfunctional impulsivity, respectively).
Participants had to answer either ‘yes’ or ‘no’. The internal consistency coecients
of those scales were, respectively, . and . in the validation study and .
and . in the present study.
...Anxiety
To evaluate anxiety levels, the French version of the Beck Anxiety Inventory (BAI;
α = .) was used (Freeston et al., ). This questionnaire includes  symp-
toms of anxiety, to which the participant answers on a scale from , ‘not at all’, to
, ‘severely’. Each symptom is evaluated according to how it afected them during
the past week. The maximum score being , a score of  to  is considered low
anxiety,  to  is considered moderate anxiety and  and above is considered
severe anxiety.
...Personality
The French version of the Big-Five Inventory (BFI; Plaisant et al., ) was used
to assess personality. This questionnaire is composed of  items and assesses
personalities on ve factors: extraversion, agreeableness, conscientiousness, neu-
roticism, and openness to experience. The internal consistency coecient is satis-
factory for each of the ve factors, ranging from . to .. For each statement,
participants had to answer on a Likert scale ranging from  ‘strongly disagree’ to
 ‘strongly agree’. The questionnaire then provides an independent total score for
each of the ve factors.
Table .
Sociodemographic characteristics of the sample (n=).
Age groups Sociodemographic characteristics
M Age Med Age S.D. nFemale Male
Adolescents (15–17 years) 15.42 15 0.63 146 98 48
Young adults (18–29 years) 23.16 23 2.87 370 277 93
Middle-aged adults (30–59 years) 40.11 38 8.61 233 177 56
Elderly adults (60 years and over) 68.98 66 8.35 145 111 34
M = Mean, Med = Median, S.D. = Standard deviation
A.-A. Gagnon-Harvey et al. / Timing & Time Perception  () –


...Temporal Impression, Occupation (Temporal Pressure), and Subjective
Memory
Participants were administered a general questionnaire comprised of  ques-
tions on the general impressions of speed of time (see Appendix A). These time-
related items were partly based on the ones used by Friedman and Janssen (),
which was inspired by the one created by Wittmann and Lehnhof (). Partici-
pants were asked to report their level of agreement on a seven-point Likert scale
(=“If you are in total disagreement or if this statement seems completely false to
you”, and =“if you are in total agreement or if this statement seems completely
true to you”).
Items , , , , , , and  relate to the impression of the speed of time in
the past. Participants are asked for their impressions using diferent time inter-
vals, ranging from yesterday to the past  years. These items, as well as item 
(which relates to the impression of the speed of time with regard to the present)
are used to measure a critical dependent variable in this study, which is the com-
posite score that adds all items about the speed of time together [a higher score
(maximum of ) means that time seems to pass more quickly].
Items , ,  and  refer to the occupation level (temporal pressure) of the
person, with  indicating a low level and  indicating a high level of occupation,
according to the participant. Items , ,  and  refer to the person’s autobio-
graphical memory and the composite score varies from  to , with  indicating
that the person considers their autobiographical memory to be excellent. Lastly,
items  and  measure participants’ belief in the phenomenon that time passes
more rapidly as we get older. The scores vary from  to ,  indicating that they
believe in it strongly. These questions aim to separate participants’ real impres-
sions from their adherence to these beliefs.
..Procedure
Participants were asked to complete the online questionnaires. They rst gave their
informed consent to participate in the study (the study was approved by the Eth-
ics committee in research with humans of Université Laval, approbation number:
– A/--). A booklet containing paper versions of every question-
naire was provided to elderly participants who were recruited in living facilities and
to anyone who requested it. Participants could answer the questions themselves or
with the help of an assistant if they so wished. This procedure was put in place to
assist participants who are less comfortable using computers. Nevertheless, many
participants aged  years or over asked to complete the questionnaires online. The
answers were then codied by two assistants in order to avoid transcription errors.
These questionnaires were combined so that participants did not have to open
several diferent internet links. All questionnaires were completed in the same
order for every participant. Thus, participants answered a total of  questions,
A.-A. Gagnon-Harvey et al. / Timing & Time Perception  () –


which took approximately  to  minutes. As a token of appreciation for their
time, participants were entered in a random draw that could allow them to win
one of nine $ gift-cards.
..Statistical Analyses
All analyses were conducted on the IBM SPSS version  software (IBM Corp.,
Armonk, NY, USA). A  (Sex) ×  (Age groups) factorial ANOVA was used to verify
whether the impression that time passes more quickly (composite score; summa-
tion of the eight temporal items) difers as a function of age group (adolescents,
young adults, middle-aged adults, and elderly adults) and sex (male or female).
Partial eta squared (η2p) values are presented as efect size. Also, post-hoc multiple
comparisons analyses were carried out [Tukey’s multiple comparisons test; as it
has been shown to have superior statistical power when used with large sample
sizes as compared to Bonferroni (Howell, )]. This analysis was also conduct-
ed separately on each of the individual items regarding the impression that time
passes more or less quickly. When homogeneity of variance was not respected,
Kruskal–Wallis H tests were used instead of ANOVAs.
A multiple linear regression (enter method) was conducted to verify if a combi-
nation of the studied personal characteristics is linked to the subjective speed of
time. This analysis also identied the contribution of each variable.
.Results
Unless otherwise specied, the score relating to the speed of time is the composite
score. Table  presents the mean score in each age group, as well as mean scores
for all of the investigated individual characteristics.
..Efect of Age and Sex on the Subjective Speed of Time
Figure  illustrates the mean composite score relating to the speed of time for
each age group and as a function of sex. The  (Sex)×  (Age groups) factorial
 indicates no interaction between age and sex (F,=., p=.,
η2p= .). Note that homogeneity of variances is not respected (Levene’s test,
p=.), suggesting the need for a correction. However, both main efects were
signicant, age: F, = ., p < ., η2p = .; sex: F, = .,
p=., η2p=.. Post-hoc Tukey multiple comparisons revealed that elderly
adults’ scores regarding speed of time are signicantly diferent from those of the
three other age groups, with adolescents feeling like time has passed the slow-
est (p <.), followed by young adults (p <.) and middle-aged adults
(p=.). Adolescents’ scores are also signicantly diferent from the young and
the middle-aged adults’ scores (p<.). However, there is no signicant difer-
ence between young and middle-aged adults (p=.). Finally, female partici-
pants nd that time passes more rapidly, compared to male participants.
A.-A. Gagnon-Harvey et al. / Timing & Time Perception  () –


Kruskal–Wallis tests comparing age groups were run on each of the individual
time items (each temporal interval), as it was not possible to assume the equality
of variance in the groups, as is required for running an ANOVA. Results indicate
a signicant diference for all intervals (all ps<.). Table  presents the mean
score (out of ) and standard deviation for each item and summarizes where
the diferences between groups are according to post-hoc pairwise comparisons.
Adolescents generally stand out from the other three groups, contrary to what
Table .
Mean scores for each variable according to age groups.
Variables Age groups
Adolescents Young adults Middle-aged adults Elderly
Composite time score 41.01 45.27 46.46 48.99
Belief about speed of time 10.00 11.00 12.00 11.92
Occupation level 15.30 18.19 18.68 19.30
Subjective memory 19.62 20.32 20.12 20.27
Anxiety 12.72 8.62 7.57 6.71
Functional impulsivity 5.68 6.28 6.72 6.08
Dysfunctional impulsivity 5.05 2.62 2.62 1.90
Openness to experience 3.42 3.70 3.83 3.76
Conscientiousness 3.32 3.83 3.95 4.20
Extraversion 3.24 3.21 3.16 3.07
Agreeableness 3.54 3.92 3.97 4.27
Neuroticism 3.08 2.87 2.79 2.30
The scores for the BFI factors are composed of the mean scores of the items from each scale and not from the summation of
scores.
0
10
20
30
40
50
60
Adolescents Young adultsAdults Elderly
Composite time score
Male Female
Figure . Mean composite scores for temporal impression as a function of age group and sex. A
higher score indicates a higher impression that time passes rapidly. Bars represent the standard error
of the means.
A.-A. Gagnon-Harvey et al. / Timing & Time Perception  () –


is shown with the composite score. Also, for longer time intervals (ve years, 
years), elderly adults were signicantly diferent from the other age groups.
..Relation Between Other Personal Characteristics and the Subjective Speed of
Time
A multiple linear regression (enter method) was run including several predictors:
age group (young adults, middle-aged adults, and elderly adults; adolescents being
the reference group), sex, occupation level, the belief that the speed of time in-
creases as we get older, functional impulsivity, dysfunctional impulsivity, each of
the ve personality traits from the Big-Five Inventory (openness to experience, con-
scientiousness, extraversion, agreeableness, and neuroticism), subjective memory,
and anxiety level (moderate and severe, the low category being used as reference).
The main assumptions for multiple regression were tested. Firstly, the Breusch–
Pagan test is signicant, χ2=., p<., indicating that homoscedastic-
ity is not respected (i.e., the variance of the error term difers depending on the dif-
ferent levels of predictors). In this condition, White’s variance–covariance matrix
() was used. Upon examination of the residuals, it was determined that the
linearity and normality hypotheses were respected. The overall test indicates that,
taken together, the personal characteristics signicantly predict the impression
that time passes more or less quickly, F,=., p<., R2=.. The
statistical results of the regression are detailed in Table .
The independent variables that contribute the most to this explanation of vari-
ance are, in order of predictive value: the belief in temporal compression, a severe
anxiety level, and a high score for conscientiousness and agreeableness. It should
be mentioned that the prediction level of perceived occupation level (p=.)
and belonging to the elderly adult group of  years or over (p=.) is mar-
ginally signicant. The previously mentioned signicant correlation between the
impression that time passes more or less quickly and dysfunctional impulsivity
Table .
Mean scores on each temporal item and diferences between age groups.
Variables Groups
Adolescents Young adults Middle-aged adults Elderly
Yesterday 4.78a (1.89) 5.18ab (1.72) 5.39b (1.68) 5.77c (1.70)
The past week 5.32a (1.79) 5.88b (1.48) 5.82b (1.59) 6.19b (1.31)
The past month 5.47a (1.80) 6.24b (1.27) 6.08b (1.45) 6.25b (1.32)
The past year 5.79a (1.50) 6.11a (1.33) 6.00a (1.53) 6.15a (1.35)
The past three
years
5.29a (1.59) 5.85b (1.44) 5.87bc (1.49) 6.17c (1.35)
The past ve years 5.23a (1.53) 5.81b (1.39) 5.99bc (1.44) 6.31c (1.24)
The past ten years 4.03a (1.77) 4.43a (1.93) 5.39b (1.66) 5.97c (1.44)
Diferent letters (a, b, or c) in superscript indicate a signicant diference, after Bonferroni correction, as compared to other
groups, whereas the same letter indicates no diference. Note that for the ‘Past year’ measure, the global Kruskal–Wallis testwas
signicant (p<0.04), but none of the post-hoc comparisons were signicant after correction.
A.-A. Gagnon-Harvey et al. / Timing & Time Perception  () –


and neuroticism, as well as the efect of sex on this impression, are therefore no
longer signicant when accounting for other predictors.
.Discussion
This study had two main objectives. First, it aimed to verify whether a person’s age
is related to the impression that time passes more or less rapidly. Second, the study
aimed to document which personal characteristics are involved in this impression.
..Age
The results of the present study partially replicate those typically found in the
literature: in general, older adults nd that time passes more quickly than young
Table .
Standard linear regression for subjective speed of time as a function of age, sex, occupation level,
memory, belief in temporal compression, anxiety level, functional and dysfunctional impulsivity
(Imp), and personality traits (BFI).
Variables B β S.E. B t p p. cor.
Intercept 15.89  3.41 4.66 < 0.01 
Sex 0.30 0.01 0.59 0.51 0.61 0.000
Impulsivity      
Functional 0.14 0.05 0.10 1.39 0.17 0.001
Dysfunctional 0.10 0.03 0.09 1.09 0.27 0.001
BFI      
Openness to exp. 0.14 0.01 0.38 0.36 0.72 0.000
Conscientiousness 1.14 0.09 0.43 2.64 0.01 0.004
Extraversion 0.25 0.02 0.36 0.68 0.49 0.000
Agreeableness 1.04 0.07 0.51 2.05 0.04 0.003
Neuroticism 0.33 0.03 0.33 1.01 0.31 0.001
Occupation 0.15 0.06 0.08 1.87 0.06 0.003
Subjective memory 0.01 0.01 0.07 0.19 0.85 0.000
Belief in the phenomenon 1.75 0.54 0.12 14.62 < 0.01 0.269
Age      
Adolescents (reference) – – – – – 
Young 0.18 0.01 0.76 0.24 0.81 0.000
Middle-aged 0.86 0.04 0.89 0.97 0.33 0.001
Elderly 1.80 0.07 1.00 1.81 0.07 0.002
Anxiety      
Low (reference) – – – – – 
Moderate 0.26 0.01 0.93 0.28 0.78 0.000
Severe 8.20 0.11 1.99 4.12 < 0.01 0.011
Bold indicates signicant predictors.
A.-A. Gagnon-Harvey et al. / Timing & Time Perception  () –


adults (Flaherty & Meer, ; Friedman & Janssen, ; Wittmann & Lehnhof,
). As in Friedman and Janssen’s study, the efect of age showed that elderly
adults report time as passing more rapidly only for the statement relative to a pe-
riod of  years and not for other periods. However, unlike the results of Flaherty
and Meer, the present study showed no signs that would indicate that middle-
aged adults perceive the speed of time as faster than elderly adults do.
This study also reveals that adolescents had the lowest composite score. There-
fore, they found that time passes more slowly than participants from other age
groups do. Indeed, this nding holds true for each temporal item except for the
past year measure. This result may seem inconsistent with the conclusions of Droit-
Volet and Wearden (), who reported no diference between young and elder-
ly people with regards to judgements about the passage of time. However, some
methodological aspects were diferent. In Droit-Volet and Wearden’s study, young
participants were students with an average age of around  years, and in the pres-
ent study they were teenagers (median age ); life experiences in these two age
groups are likely very diferent. Moreover, in thgeir study, Droit-Volet and Wearden
asked participants to make a judgement about the passage of time without having
a specic amount of time passed as reference (e.g., “At the moment, how does time
pass for you compared to the time of the clock”). In the present study, all judge-
ments about the passage of time referred to a specic amount of time (e.g., the past
ten years). Using a specic range of time might involve a diferent mechanism as
this range could be contrasted with the duration of the rest of our life. The ratio of
this range with the rest of our life is smaller for elderly adults than for younger peo-
ple, which might cause the impression that time passes more rapidly for the elderly.
Even though the overall results point in the direction of a positive and linear
age efect on the speed of time, any interpretation of this efect requires caution.
Despite the increases of the composite scores with age, the linear regression mod-
el makes it possible to understand that, in fact, the speed of time is not only a
matter of age; it also depends on a series of other personal characteristics, which
include the belief in the phenomenon, anxiety, and personality.
..Personal Characteristics
Several factors seem to contribute to the understanding of what inuences the im-
pression that the speed of time changes with ageing. The personal characteristic
that best explains this impression seems to lie in believing in the phenomenon.
However, this study does not allow us to determine the direction of the relation.
On the one hand, it could be that participants who believe in the phenomenon
have a bias induced by this belief, similarly to the one explained by Lamotte et
al. (), where a potential attentional bias can afect temporal judgements
(see also Lamotte et al., ). On the other hand, it could be that their belief is
A.-A. Gagnon-Harvey et al. / Timing & Time Perception  () –


actually caused by their personal experience of having the impression that time
passes faster as they get older.
The second most predictive variable of the impression that the speed of time
changes as we get older is anxiety level. Participants with a high level of anxiety,
as measured by the French version of BAI, had the impression that time passes
more slowly. This nding is consistent with studies that show that people strug-
gling with anxiety overestimate durations (Bar-Haim et al., ) and that people
that present anxiety in some situations, when compared to situations where they
do not experience it, estimate durations as longer than they really are (Sarason &
Stoops, ; Watts & Sharrock, ).
Likewise, a conscientious or agreeable personality also contributes to the ex-
planation of the subjective impression relative to the speed of time. It is still not
completely clear why these characteristics can be identied as predictors, but an
explanation could be linked to time perspective. Conscientious people tend to be
good planners; they like control and are well aware of due dates (Costa & McCrae,
; Goldberg, ), which leads them to usually score higher on the future-
oriented dimension of a time perspective inventory (e.g., Kairys & Liniauskaite,
; Zimbardo & Boyd, ). Future-oriented people tend to sacrice a lot of
personal time (e.g., the time allocated toward fun, hobbies, or friends) in order
to complete work-related or project-related goals (Zimbardo, ). This type of
personal time contributes to building good memories and helping people nd a
balance in their level of occupancy. One might even argue that creating poorer
memories and experiencing temporal pressure, thusly, can give a person the feel-
ing that time goes by way too quickly. This is consistent with the ndings of Wit-
tman and colleagues (), where a general perception of time passing faster
was also associated with a more pronounced future perspective. An agreeable, or
pleasant, personality is also a predictor in this model and a possible explanation
for this might lie in the multiple functions that pleasant people tend to endorse.
Pleasant people tend to put others before themselves (Rolland, ) in order to
be a good parent, friend, employee, and so on. This kind of pressure to make sure
everything and everyone is ne might make pleasant people feel like they lack
the time to do everything they want to (Beaujot, ; Zukewich, ) and, ul-
timately, it makes them feel, as they endorse new roles, that time goes faster than
it used to. This particular feature of the agreeable personality trait is consistent
with the hypothesis of Larson and von Eye (), which stipulates that the level
of engagement – here the diferent functions they tend to endorse – afects the
speed of time.
Although temporal pressure has been reported to be a core element in deter-
mining the subjective speed of time (Flaherty & Meer, , Winkler et al., )
(see Note ), the present study only partially supported that statement. The idea
behind temporal pressure is that the more stressed out by time people are, the
A.-A. Gagnon-Harvey et al. / Timing & Time Perception  () –


faster time seems to go by. In the present study, there is a correlation between
occupation level and subjective speed of time, but the linear regression model
shows that when considering other factors, occupation levels fail at explaining a
signicant portion of the variance of the subjective speed of time. Note that Witt-
mann and colleagues () reported that time pressure is a small, but signicant
predictor of the perceived passage of time. In the present study, a pooled score
of all durations was used, whereas in Wittmann et al. (), this predictor was
signicant only for the -year duration.
Although not a factor that contributes to explaining variance, according to the
regression analysis, the ANOVA indicates that there is a signicant efect of sex on
the composite results. Female participants reported overall higher scores (com-
posite index) relative to the speed of time: they nd that time passes more rapidly
than males do. This nding is surprising if we consider that in the literature, it is
reported that females have a better autobiographical memory. Moreover, in the
present study, there was no efect of sex on the following two factors: temporal
pressure and occupation level.
..Limits of the Study and Future Directions
The present study sufers from two nonnegligible limits. One is the fact that the
ndings are essentially correlational and that they do not allow to further test
which one of the theoretical explanations previously reported remains the most
robust. Another limit to this investigation is that it relies on a questionnaire that,
although it contains several items used in previous studies, is not validated.
We now know that the belief in the phenomenon of temporal compression,
the degree of anxiety, as well as conscientious and agreeable personality traits are
important determinants of the impression that time passes faster as we get older.
In addition to the development of a validated questionnaire, future work should
strive to determine which of these factors, in addition to some others like tempo-
ral perspective or temporal personality, mediates the link between age and the
subjective impression about the speed of the passage of time.
.Conclusion
This study indicates that time seems to pass faster as we get older. However, the
perception of long time intervals depends on multiple personal factors other than
age. A single variable cannot explain the variations in the perceived speed of the
passage of time. Moreover, the portion of variance explained by each of the sig-
nicant factors is often quite small, which leaves room for many additional hy-
potheses. In combination, all studied factors account for % of the variance of
the subjective impression that time accelerates with age. This suggests that other
important personal characteristics will prove to be predictors of this impression,
thus leaving much room for future work in this eld of research.
A.-A. Gagnon-Harvey et al. / Timing & Time Perception  () –


Acknowledgements
This study was part of the  mémoire doctoral of AAGH and was supported
by a research grant from the Social Sciences and Humanities Research Council of
Canada to SG. Portions of this work were presented at the st Annual Meeting
of the Société québécoise pour la recherche en psychologie held in March . We
would like to thank Yves Lacouture and Philippe Landreville for their comments
on this project, Jean Vézina for the access to PIANO, and two anonymous review-
ers for their comments and suggestions on an earlier version of this manuscript.
General Questionnaire
Note
. Note that in Winkler et al. (), the number of routines in daily life was,
with time pressure, the most important factor for explaining the speed of
time in the participants’ current life situations and previous life periods: the
more routines, the faster the speed of time.
References
Bar-Haim, Y., Kerem, A., Lamy, D., & Zakay, D. (). When time slows down: The inuence of threat
on time perception in anxiety. Cogn. Emot., , –. doi: ./.
Beaujot, R. (). Les deux transitions démographiques du Québec, –. Cah. Qué. Demogr.,
, –. https://doi.org/./ar.
Berlin H. A., Rolls E. T., & Kischka U. (). Impulsivity, time perception, emotion, and reinforce-
ment sensitivity in patients with orbitofrontal cortex lesions. Brain, –. doi: ./
brain/awh.
Block, R. A. (). Temporal judgments and contextual change. J. Exp. Psychol. Learn. Mem. Cogn.,
, –. doi: .//-....
Block, R. A., & Zakay, D. (). Prospective and retrospective duration judgments: A meta-analytic
review. Psychon. Bull. Rev., , –. doi: ./BF.
Block, R. A., Hancock, P. A., & Zakay, D. (). Sex diferences in duration judgments: A meta-
analytic review. Mem. Cogn., , –. doi: ./BF.
Caci, H., Nadalet, L., Baylé, F. J., Robert, P., & Boyer, P. (). Functional and dysfunctional im-
pulsivity: contribution to the construct validity. Acta Psychiatr. Scand., , –. doi:
./j.-...x.
Costa, P. T., & McCrae, R. R. (). The ve-factor model of personality and its relevance to person-
ality disorders. J. Personal. Disord., , –. doi: ./pedi.....
Craske, M. G., Vansteenwegen, D., & Hermans, D. (). Introduction: Etiological factors of fears
and phobias. In M. G. Craske, D. Hermans, & D. Vansteenwegen (Eds), Fear and learning: From
basic processes to clinical implications (pp. –). Washington, DC, USA: American Psychological
Association.
A.-A. Gagnon-Harvey et al. / Timing & Time Perception  () –


Droit-Volet, S., & Wearden, J. (). Passage of time judgments are not duration judgments: evi-
dence from a study using experience sampling methodology. Front. Psychol., , . doi: ./
fpsyg...
Droit-Volet, S., Mermillod, M., Cocenas-Silva, R., & Gil, S. (). The efect of expectancy of a threat-
ening event on time perception in human adults. Emotion, , –. https://doi.org/./
a.
Flaherty, M. G., & Meer, M. D. (). How time ies: Age, memory, and temporal compression.
Sociol. Q. , –. https://doi.org/./j.–..tb.x.
Freeston, M. H., Ladouceur, R., Thibodeau, N., & Gagnon, F. (). L’inventaire d’anxiété de Beck.
Propriétés psychométriques d’une traduction française. Encephale, , –.
Friedman, W. J., & Janssen, S. M. J. (). Aging and the speed of time. Acta Psychol. (Amst.), ,
–. DOI: ./j.actpsy....
Gerbing, D. W., Ahadi, S. A., & Patton, J. H. (). Toward a conceptualization of impulsivity: com-
ponents across the behavioral and self-report domains. Multivar. Behav. Res., , –. doi:
./smbr_.
Goldberg, L. R. (). An alternative “description of personality”: the big-five factor structure.
J. Pers. Soc. Psychol., , –. doi: ./-....
Grady, C. L., & Craik, F. I. (). Changes in memory processing with age. Curr. Opin. Neurobiol., ,
–. doi: ./S-()-.
Grondin, S. (). The perception of time — your questions answered. New York, NY, USA: Routledge.
https://doi.org/./.
Hancock, P. A. (). The efect of age and sex on the perception of time in life. Am. J. Psychol., ,
–. https://doi.org/./amerjpsyc....
Howell, D. C. ().Statistical methods for psychology. Belmont, CA, USA: Cengage Learning.
Janssen, S. M. J., Naka, M., & Friedman, W. J. (). Why does life appear to speed up as people get
older? Time Soc., , –. https://doi.org/./X.
Kairys, A., & Liniauskaite, A. (). Time perspective and personality. In M. Stolarski, N. Fieulaine,
& W.van Beek (Eds), Time perspective theory; review, research and application (pp. –). Cham,
Switzerland: Springer International Publishing. https://doi.org/./----_.
Krauss, H. H., & Ruiz, R. A. (). Anxiety and temporal perspective. J. Clin. Psychol., , –.
https://doi.org/./–():<::AID-JCLP>..CO;-N.
Lamotte, M., Izaute, M., & Droit-Volet, S. (). Awareness of time distortions and its relation with
time judgment: A metacognitive approach. Consc. Cogn., , –. https://doi.org/./j.
concog....
Lamotte, M., Chakroun, N., Droit-Volet, S., & Izaute, M. (). Metacognitive question-
naire on time: feeling of the passage of time. Timing Time Percept., , –. https://doi.
org/./-.
Larson, E. (). The time of our lives: The experience of temporality in occupation. Can. J. Occup.
Ther., , –. https://doi.org/./.
Larson, E., & von Eye, A. (). Predicting the perceived ow of time from qualities of activ-
ity and depth of engagement. Ecological Psychology, , –. https://doi.org/./
seco_.
Lemlich, R. (). Subjective acceleration of time with aging. Percept. Mot. Skills, , –.
https://doi.org/./pms.....
A.-A. Gagnon-Harvey et al. / Timing & Time Perception  () –


Ma, A. K. C., & Cheung, J. Y. K. (). Subjective acceleration of time: a stochastic approach. Appl.
Math. Sci., , –. http://dx.doi.org/./ams...
Pillemer, D., Wink, P., DiDonato, T., & Sanborn, R. (). Gender diferences in autobiographical
memory styles of older adults. Memory, , –. doi.org/./.
Plaisant, O., Courtois, R., Réveillère, C., Mendelsohn, G. A., & John, O. P. (). Validation par anal-
yse factorielle du Big Five Inventory français (BFI-Fr). Analyse convergente avec le NEO-PI-R.
Ann. Med. Psychol. (Paris), Rev. Psychiatr., , –. doi: ./j.amp....
Rolland, J.-P. (). L’évaluation de la personnalité: le modèle en cinq facteurs. Sprimont, Belgique:
Editions Mardaga.
Roy, M. M., & Christenfeld, N. J. S. (). Efect of the task length on remembered and predicted
duration. Psychon. Bull. Rev., , –. DOI: ./pbr....
Roy, M. M., Christenfeld, N. J. S., & McKenzie, C. R. M. (). Underestimating the duration of fu-
ture events: Memory incorrectly used or memory bias? Psychol. Bull., , –. https://doi.
org/./-....
Salthouse, T. A. (). Aging and measures of processing speed. Biol. Psychol., , –. doi:
./S-()-.
Sarason, I. G., & Stoops, R. (). Test anxiety and the passage of time. J. Consult. Clin. Psychol., ,
–. https://doi.org/./-X....
Schif, W., & Thayer, S. (). Cognitive and afective factors in temporal experience: anticipated or
experienced pleasant and unpleasant sensory events. Percept. Mot. Skills, , –. https://
doi.org/./pms.....
Stetson, C., Fiesta, M. P., & Eagleman, D. M. (). Does time really slow down during a frightening
event? PLoS ONE, , e. https://doi.org/./journal.pone..
Tabachnick, B. G., & Fidell, L. S. (). Using Multivariate Statistics (th ed.). Northridge, CA, USA:
California State University.
Watts, F. N., & Sharrock, R. (). Fear and time estimation. Percept. Mot. Skills, , –.
https://doi.org/./pms.....
White, H. (). A heteroskedasticity-consistent covariance matrix estimator and a direct test for
heteroskedasticity. Econometrica, , –. https://doi.org/./.
Winkler, I., Fischer, K., Kliesow, K., Rudolph, T., Thiel, C., & Sedlmeier, P. (). Has it really been
that long? Why time seems to speed up with age. Timing Time Percept., , –. https://doi.
org/./-.
Wittmann, M., & Lehnhof, S. (). Age efects in perception of time. Psychol. Rep., , –.
doi: ./PR...-.
Wittmann, M. & Paulus, M. P. (). Decision making, impulsivity and time perception. Trends
Cogn. Sci., , –. https://doi.org/./j.tics....
Wittmann, M., Rudolph, T., Linares Gutierrez, D., & Winkler, I. (). Time perspective and emo-
tion regulation as predictors of age-related subjective passage of time. Int. J. Environ. Res. Public
Health, , –. doi: ./ijerph.
Zimbardo, P. G. (). The Psychology of Time [Video]. Tedtalk. Available from https://www.ted.
com/talks/philip_zimbardo_the_psychology_of_time.
Zimbardo, P. G., & Boyd, J. N. (). Putting time in perspective: A valid, reliable individual-difer-
ences metric. J. Pers. Soc. Psychol., , –. doi: ./-....
Zukewich, N. (). Work, parenthood and the experience of time scarcity. Ottawa, ON, Canada: Sta-
tistics Canada, Housing, Family and Social Statistics Division.
A.-A. Gagnon-Harvey et al. / Timing & Time Perception  () –


Appendix A
General Questionnaire
Time-related impressions:
. “The past ve years have seemed to pass very quickly.”
. “The past month has seemed to pass very quickly.”
. “The past year has seemed to pass very quickly.”
. “The past week has seemed to pass very quickly.”
. ”The past three years have seemed to pass very quickly.”
. “Yesterday has seemed to pass very quickly.”
. “The past ten years have seemed to pass very quickly.”
. “Time generally seems to pass very quickly.”
Belief that time passes more rapidly with age:
. “As I get older, time seems to pass more and more quickly.”
. “A day seems to pass more quickly now than when I was ten years old.”
Occupation-related items:
. “I was very bored yesterday.”
. “In the past several years, my life has been fairly routine.”
. “I was very busy this past year.”
. “I often feel like I do not have enough time to accomplish everything I had
planned to in a day.”
Memory-related items:
. “In general, I have a very good memory of past events.”
. “I often notice that events happened much longer ago than I thought.”
. “I remember many important events in my life.”
. “I tend to remember many events that happened in my life, even those that
were not especially remarkable.”
Numbers indicate the order of questions.
A.-A. Gagnon-Harvey et al. / Timing & Time Perception  () –