Available via license: CC BY-NC-ND 4.0
Content may be subject to copyright.
The overlapping geography of cognitive ability and chronotype
Péter Przemyslaw Ujma ,
1,2
and Emil Ole William Kirkegaard
3
1
Institute of Behavioural Sciences, Semmelweis University, Budapest, Hungary,
2
National Institute of
Clinical Neuroscience, Budapest, Hungary,
3
Ulster Institute for Social Research, Ulster, UK
Abstract: Chronotype and cognitive ability are two human phenotypes with an uneven geographic distribution due to both selective
migration and causal environmental effects. In our study, we aimed to examine the relationship between geographic variables, cognitive
ability and chronotype. We used a large anonymized sample (n=25,700, mostly from the USA, UK, Canada and Australia) of dating
site users to estimate chronotype and cognitive ability from questionnaire responses using item response theory. We matched each user to
geographic coordinates and city size using the reported locations and geographic databases. In line with previous research we found that
male sex, younger age, residence in a more populous locale, higher cognitive ability and more westward position within the same time
zone were associated with later chronotype. Male sex, younger age, residence in a more populous locale, later chronotype and higher lati-
tude were associated with higher cognitive ability, but the effect of population on chronotype and latitude on cognitive ability was only
present in the USA. The relationship between age and chronotype was stronger in males, and the relationship between chronotype and
cognitive ability was stronger in males and in older participants. Population density had an independent association with cognitive ability,
but not chronotype. Our results confirm the uneven geographic distribution of chronotype and cognitive ability. These findings generalize
across countries, but they are moderated by age and sex, suggesting both biological and cultural effects.
Keywords: chronotype; circadian preference; g factor; geography; intelligence; IQ
Correspondence to: Péter Przemyslaw Ujma, Institute of Behavioural Sciences, Semmelweis University, 1089 Budapest, Nagyv
arad
tér 4, Hungary. Email: ujma.peter@med.semmelweis-univ.hu
Received 8 December 2020. Accepted 1 July 2021.
Introduction
Humans live in diverse environments, even those sharing
the same culture and living within the same country. Both
the man-made (e.g., construction type and density, utilities,
available jobs) and natural (e.g., altitude, climate, geo-
graphic position) features of these environments may have
an effect on human psychophysiological characteristics
(Rentfrow & Jokela, 2016). On the other hand, different
human environments may also attract different people
(e.g., urban life may be more attractive for younger individ-
uals), leading to an uneven geographic distribution of the
same psychophysiological characteristics due to selective
migration rather than causal environmental effects (Clark &
Cummins, 2018; Obschonka et al., 2018). In our study, we
investigated how a combination of these environmental
effects and selective migration leads to the uneven geo-
graphic distribution of two specific phenotypes: chronotype
and cognitive ability.
Chronotype refers to the entrainment of the sleep–wake
cycle to a certain phase of the natural diurnal light–dark
cycle. Chronotype is determined by both genetic–biological
and environmental factors. Younger individuals, particu-
larly young males, are characterized by a substantially later
chronotype than both children and older adults, but the sex
difference may disappear or even reverse at a later age
(Duarte et al., 2014; Fischer et al., 2017; Roenneberg
et al., 2004; Tonetti et al., 2008). Genetic factors account
for up to 50% of individual differences in chronotype at
the same age (Inderkum & Tarokh, 2018; Jones
et al., 2019; Koskenvuo et al., 2007), but environmental
determinants are also important. Artificial light in the
PsyCh Journal (2021)
DOI: 10.1002/pchj.477
© 2021 The Authors. PsyCh Journal published by Institute of Psychology, Chinese Academy of Sciences and John Wiley & Sons
Australia, Ltd.
This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use
and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations
are made.
evening delays the sleep–wake cycle (Figueiro et al., 2014;
Porcheret et al., 2018; Shawa et al., 2018), but so does the
variation of the timing of the day–night cycle within
the time zone, with individuals living in the western
(Giuntella & Mazzonna, 2019; Roenneberg et al., 2007;
Sl
adek et al., 2020) and northern (Leocadio-Miguel
et al., 2017; Miguel et al., 2014; Porcheret et al., 2018)
parts of the same time zone having a later chronotype.
Some evidence suggests that living in a larger inhabited
area also contributes to a later chronotype, presumably
because of increased artificial lighting and more opportuni-
ties for nocturnal activity (Roenneberg & Merrow, 2007;
Sl
adek et al., 2020). Individual differences in chronotype –
especially a late chronotype –may result in a misalignment
of the daily rhythm of society and negatively impact health
(Castilhos Beauvalet et al., 2017; Knutson & von
Schantz, 2018; Partonen, 2015; Vetter et al., 2015;
Wittmann et al., 2006).
Cognitive ability in our operationalization refers to a
general knowledge acquisition/retention, problem solving
and pattern recognition capacity which is most frequently
measured by IQ tests. Cognitive ability is a key psychologi-
cal phenotype which is associated with career success,
income and education (Strenze, 2007) as well as social suc-
cess (Hegelund et al., 2018; Hegelund et al., 2019;
Strenze, 2015a, 2015b) and health (Calvin et al., 2011;
Calvin et al., 2017; Gale et al., 2010). The majority of the
variance in adult cognitive ability is accounted for by
genetic factors (Bouchard, 2013; Plomin & Deary, 2015),
but selective migration creates notable geographic differ-
ences, with higher average cognitive ability in urban areas
(Alexopoulos, 1997; Gist & Clark, 1938; Lehmann, 1959;
Teasdale et al., 1988). A recent analysis (Abdellaoui
et al., 2019) of the UK Biobank revealed that genetic vari-
ants associated with higher cognitive ability are associated
with migrating out of geographic areas with a low average
socioeconomic profile (which tend to be rural areas) and
migrating into economically successful areas (which tend
to be large cities).
Chronotype and cognitive ability are themselves corre-
lated. A meta-analysis estimated the correlation between
chronotype and cognitive ability at r=.04 –.08 (Preckel
et al., 2011), with later chronotypes in individuals with
higher cognitive ability. However, the moderating effect of
age was significant, with a higher chronotype–cognitive abil-
ity correlation in older samples. This trend has since been
borne out in individual studies of children and adolescents,
which often found a negative correlation between cognitive
ability and late chronotype (Arbabi et al., 2015; Rahafar
et al., 2017). We recently hypothesized (Ujma et al., 2020)
that social effects introduce the correlation between cognitive
ability and late chronotype, because individuals with higher
cognitive ability tend to work in professions with later or
more flexible work schedules, which permits a late chro-
notype in those inclined to one.
In the current study, we used a large archival dataset and a
novel, item response theory-based statistical approach of phe-
notype estimation to investigate the geographic and demo-
graphic correlates of both cognitive ability and chronotype.
Our aims were twofold. First, we aimed to replicate previous
findings about cognitive ability and chronotype, such as sex
and age effects, in order to demonstrate the validity of our
method. Second, we aimed to take advantage of our large,
geographically informed international dataset to investigate
novel questions: (i) the consistency of the correlates of cogni-
tive ability and chronotype in various countries; and (ii) the
moderating effect of demographic features on the correlation
of the two phenotypes as well as the consistency of their other
correlates across the sexes and throughout young adulthood.
Methods
A prior study used a scraper script to collect data from the
user profiles on the dating website OKCupid
(Kirkegaard & Bjerrekær, 2016). We re-used the dataset
generated by this study. User profiles were freely accessible
to all web users with a free profile at the time of data col-
lection, and the collected data was anonymized by deleting
user names and any other information which would permit
identification. The analysis of such web-based, pre-existing
data constitutes archival rather than human research
(Bruckman, 2002; Herring, 1996; Kosinski et al., 2015)
and generally does not require permission from an ethical
board (Catanese et al., 2011; Gjoka et al., 2010; Kim &
Escobedo-Land, 2015; Kirkegaard & Lasker, 2020;
Rahman, 2012; Subirats et al., 2018). Our research did not
violate specific guidelines (Kosinski et al., 2015) which
would render closer ethical scrutiny necessary:
•It was reasonable to assume that the data were knowingly
made public by the individuals
•Data was anonymized after collection and no attempts
were made to deanonymize them
2 Geography, intelligence, chronotype
© 2021 The Authors. PsyCh Journal published by Institute of Psychology, Chinese Academy of Sciences and John Wiley & Sons
Australia, Ltd.
•There was no interaction or communication with the
individuals in the sample
•No information that can be attributed to a single individ-
ual, including demographic profiles and samples of text
or other content, is published or used to illustrate the
results of the study.
User profiles contained self-reported data on age, sex, loca-
tion and the responses the user gave to a set of questions
(Table 1) posted on the website, intended to match individ-
uals with similar responses. Responses to a set of these ques-
tions were used to extract a latent cognitive ability factor and
a latent chronotype factor using item response theory (IRT;
DeMars, 2010), implemented by the mirt R-package. (For a
similar approach with the same dataset, see Kirkegaard &
Lasker, 2020). We used the 2PL model, allowing for items
to vary in difficulty and factor loading. IRT is similar to
principal component analysis in that it extracts the common
variance of ordinal responses or correct/incorrect responses,
using response frequencies or difficulty based on the propor-
tion of correct responses to transform responses to approxi-
mate a normal distribution. User-reported locations were
cross-referenced to the freely available SimpleMaps
(US locales, https://simplemaps.com/data/us-cities) or Geo-
Names cities500 (non-US locales, https://download.
geonames.org/export/dump/) database, finding matching
combinations of country, state and city names. Both data-
bases contained information about the longitude, latitude,
population and time zone of the participants’reported loca-
tion, while SimpleMaps data for the USA also had data on
population density. We also calculated a relative longitude
by subtracting from the longitude of each participant the
mean longitude of all participants within the same time zone.
A negative relative longitude thus signifiesarelativeposition
west of the time zone mean and a positive relative longitude
a relative position east of the time zone mean.
In the analytical sample, we included participants who:
(i) reported a location which was successfully matched to a
place name with geographic coordinates; (ii) answered
enough OKCupid questions to permit the estimation of
both chronotype and cognitive ability; and (iii) reported
their sex as either “Male”or “Female”. This resulted in a
final analytical sample of 25,700 individuals, of whom
8,004 were females and 17,696 were males (there were
153 “Other”responses). Nineteen thousand five hundred
and seventy-five participants reported a location in the
USA, 1,648 in the UK, 1,171 in Canada and 516 in
Australia. Two thousand eight hundred and thirty-four indi-
viduals reported a location in another country. Countries
with at least 100 participants were Germany (n=455), the
Netherlands (n=219), Sweden (n=146), France
(n=143), Denmark (n=124), Finland (n=144) and
Brazil (n=100). The mean age of participants was
33.98 years (range, 18–100 years, SD =7.63 years). Age,
chronotype and relative longitude followed an approxi-
mately normal distribution. We used the 10-base logarithm
of city populations in linear models due to the heavily
skewed distribution of this variable. Histograms of all con-
tinuous variables are reported in Figure 1.
We used ordinary least square (OLS) regression models
to investigate the demographic and geographic correlates of
cognitive ability and chronotype. STATISTICA 12 was
used for all statistical analyses.
Results
Main analyses
We first investigated the relationship between age, sex,
geographic variables, cognitive ability and chronotype in
the entire sample. We ran two models, each with age, sex
and geographic variables as independent variables and
either chronotype or cognitive ability as the dependent vari-
able, while the other was added to the independent
variables. Results are reported in Table 2. A full correlation
matrix of all variables is reported in Table 3.
Male sex, younger age, higher cognitive ability and loca-
tion with a larger population or at a more westward
location within the same time zone were associated with
later chronotype. Conversely, male sex, younger age, later
chronotype and location with a larger population or at more
northerly latitude were associated with higher cognitive
ability. All effects were modest, the largest being the effect
of log population on cognitive ability (β=.149) and the
effect of age on chronotype (β=.178).
Country differences in geographic effects
We next investigated whether the main associations
between geographic factors, cognitive ability or chronotype
generalize across countries (Table 4) by adding country
interaction effects to the original models. With chronotype
as the dependent variable, the age*country interaction
effect was significant (F=3.075, p=.0152) but the effect
of age was substantial and negative in all countries. With
PsyCh Journal 3
© 2021 The Authors. PsyCh Journal published by Institute of Psychology, Chinese Academy of Sciences and John Wiley & Sons
Australia, Ltd.
Table 1
The OKCupid questions used to construct cognitive ability and chronotype scores. In the column. The percentage of responses given by participants is
shown in parentheses after the question text, in the order as they appear in the table. The last percentage shows the proportion of missing responses. For
cognitive ability items, the correct response rate has been highlighted in bold. All percentages are relative to a total sample size of 25,700
No. Question text Option 1 Option 2 Option 3 Option 4
Cognitive ability
1 Which is bigger? (1.47/96.92/1.59%) The earth The sun
2 STALE is to STEAL as 89,475 is to…
(77.35/5.1/8.22/3.36/5.96%)
89,457 98,547 89,754 89,547
3 What is next in this series? 1, 4, 10,
19, 31, _ (0.98/1.77/
79.97/11.63/5.63%)
36 48 46 Do not know/do not
care
4 If you turn a left-handed glove inside
out, it fits…(17.66/68.61/13.73%)
On my left hand On my right hand NA NA
5 In the line ‘Wherefore art thou
Romeo?’what does ‘wherefore’
mean?
(56.6/28.47/1.51/7.62/5.79%)
Why Where How Who cares/wtf?
6 How many fortnights are in a year?
(2.51/1.53/35.73/1.84/58.4%)
52 14 26 365
7 Half of all policemen are thieves and
half of all policemen are
murderers. Does it follow logically
that all policemen are criminals?
(59.32/9.26/31.41%)
Yes No
8 Which is longer?
(70.27/9.77/16.82%)
A mile A kilometer I do not know!
9 When birds stand on power lines and
do not get hurt, it’s most likely
because of: (1.78/6.46/
34.77/1.77/55.22%)
Good timing, they
only land
between calls
Body materials that
are insulated
from current
Not touching
anything else at
the same time
They do get hurt,
they just express it
poorly
10 Etymology is…(0.03/1.26/
17.24/1.3/80.18%)
The study of
culinary arts
The study of insects The study of the
origins of words
I do not know
11 If some men are doctors and some
doctors are tall, does it follow that
some men are tall? (39.34/
38.91/21.75%)
Yes No
12 A little grade 10 science. Ideal Gas
Law?
(3.37/0.2/0.05/1.67/94.72%)
PV =nRT G +V=1/T y =mx +b Not sure/wish I could
skip this one
13 If you flipped three pennies, what
would be the odds that they all
came out the same? (10.46/9.56/
10.6/14.37/55.02%)
I admit, I do not
know!
1in3 1in4 1in8
14 Which is the day before the day after
yesterday?
(30.72/4.4/1.05/63.85%)
Yesterday. Today. Tomorrow.
Chronotype
1 On a typical night, what time do
you go to sleep?
(2.58/27.52/28.84/16.28/24.78%)
By 9 PM By 11 PM By 1 AM Later
2 Does your ideal schedule involve
staying up very late at night and
sleeping during the day?
(28.36/43.61/28.02%)
Yes No
3 If you have no obligations, at what
time do you prefer to get up in the
morning?
(2.63/30.02/42.89/7.95/16.51%)
Early bird gets the
worm! I’mup
before the sun!
Pretty early
(6:00ish -
9:30ish AM)
I like to sleep in a
bit. (9:30ish
AM–Noonish)
Morning? Curse that
AM light!
(afternoon or
dark)
4 On weekends/days off do you like to
get out and make the most of the
day or do you prefer to sleep late
and relax?
(13.28/13.91/64.38/8.43%)
Get up and do
something
Sleep late and relax It varies
4 Geography, intelligence, chronotype
© 2021 The Authors. PsyCh Journal published by Institute of Psychology, Chinese Academy of Sciences and John Wiley & Sons
Australia, Ltd.
cognitive ability as the dependent variable, the country*log
population interaction was significant (F=9.63, p< .001),
but the relationship between cognitive ability and log popu-
lation was also substantial and positive in all except
“Other”countries. The country*latitude interaction
approached significance (F=2.294, p=.056) and the
relationship between latitude and cognitive ability was only
significant in the USA and in “Other”countries.
In sum, while some variation in effect sizes was seen,
higher age was unambiguously associated with earlier chro-
notypes in all countries while later chronotype and residence
in a larger city was associated with higher cognitive ability.
While their interaction with the country variable did not reach
significance, the modest association between larger population
and later chronotype and more northerly latitude and higher
cognitive ability was not unambiguous across countries.
Figure 1. Histograms of continuous variables
Table 2
Model results about the variables associated with cognitive ability and chronotype. Zero-order correlations are reported in addition to multiple standardized
regression coefficients for comparison. p-values are reported for the beta coefficients. Beta coefficients for sex refer to males with females for reference.
n=25,700 for all analyses. All regression coefficients are standardized
Dependent variable: cognitive ability Dependent variable: chronotype
βrF p βrF p
Chronotype .051 .062 65.678 <.001 Cognitive ability .050 .062 65.678 <.001
Latitude .030 .024 24.309 <.001 Latitude .005 .002 0.731 .393
Relative longitude .007 .021 1.373 .241 Relative longitude .042 .033 47.215 <.001
Log population .149 .148 576.653 <.001 Log population .020 .035 10.469 .001
Age .040 .044 38.238 <.001 Age .178 .173 788.221 <.001
Male sex .065 103.425 <.001 Male sex .029 20.492 <.001
Table 3
Correlation matrix of all variables used in the analyses
Age Cognitive ability Chronotype Latitude Relative longitude Log population
Age 0.044 0.173 0.008 0.039 0.076
Cognitive ability 0.062 0.024 0.021 0.148
Chronotype 0.002 0.033 0.035
Latitude 0.060 0.061
Relative longitude 0.094
PsyCh Journal 5
© 2021 The Authors. PsyCh Journal published by Institute of Psychology, Chinese Academy of Sciences and John Wiley & Sons
Australia, Ltd.
The effect of relative longitude
We investigated the effect of relative longitude by time
zones instead of countries. We restricted analyses to
selected time zones which were well represented in the
dataset: GMT -8 to GMT -5 (Pacific, Mountain, Central
and Eastern time zones, respectively, of the USA and
Canada), GMT +0 (UK, Portugal, Iceland), GMT +1
(most of continental Europe), GMT +8 and GMT +10
(time zones of parts of Russia and several East Asian coun-
tries and the major Australian population centers, respec-
tively). Twenty-four thousand seven hundred and five
participants (96.1% of the total) resided in these time
zones. We found a negative association between relative
longitude and chronotype –that is, a typically later chro-
notype in those residing in more westward locations –in
all time zones except one (GMT +10), which, however,
only reached significance in the GMT -5 (Eastern Standard
Time), GMT +0 (UK, Portugal, Iceland) and GMT +8
(Australia Western Standard Time) time zones, in part due
to limited sample size (Table 5) and limited longitude vari-
ance (see Discussion). Figure 2 illustrates chronotype dif-
ferences as a function of relative longitude.
Sex differences
We re-ran our main models with the addition of an inter-
action effect for sex. This analysis suggested sex differ-
ences in some of the correlates of both cognitive ability
and chronotype (Table 6). Specifically: (i) the relation-
ship between higher age and earlier chronotype was
stronger in males; (ii) the relationship between higher
age and lower cognitive ability was stronger in females;
(iii) the relationship between later chronotype and higher
cognitive ability was stronger in males; and (iv) the rela-
tionship between higher log population and later chro-
notype was stronger in males (and only reached
significance in this subgroup).
Age, chronotype and cognitive ability
We next investigated whether the relationship between
chronotype and cognitive ability was moderated by age.
We split the sample to age quartiles. The quartile bound-
aries were at 18–29 years, 29–33 years, 33–38 years and
over 38 years, with individuals with ages at the quartile
boundary always assigned to the higher quartile (n=6,073–
7,062 within quartile, unequal samples are due to the gran-
ularity of year-rounded age data). We added interaction
effects (age quartile*continuous predictors) to our original
model with cognitive ability as the dependent variable.
Interaction effects were significant for chronotype and log
population. The effect of age within age quartiles was also
significantly different (Table 7). In older populations, the
relationship between later chronotype and higher cognitive
ability was stronger, but the relationship between more
populous place of residence and higher cognitive ability
was weaker. The negative relationship between age and
cognitive ability also changed from mildly positive to nega-
tive in older age quartiles (Figure 3).
Table 4
Country-wise analysis of the relationship between cognitive ability, chronotype and geographic variables. All multiple regression coefficients are adjusted
for all other predictors (full list: age, sex, chronotype, latitude, longitude deviation, chronotype, cognitive ability). Significant associations are shown in
bold. All regression coefficients are standardized
USA
(n=19,575)
UK
(n=1,615)
Canada
(n=1,171)
Australia
(n=516)
Other
(n=2,823)
βpβpβpβpβp
Age vs. chronotype .182 <.001 .169 <.001 .201 <.001 .173 <.001 .135 <.001
Chronotype vs. cognitive ability .050 <.001 .046 .068 .043 .143 .062 .165 .039 .035
Log population vs. chronotype .021 .003 .021 .451 .020 .513 .057 .202 .022 .272
Log population vs. cognitive ability .151 <.001 .178 <.001 .136 <.001 .125 .005 .003 .897
Latitude vs. cognitive ability .034 <.001 .035 .276 .010 .770 .130 .017 .133 <.001
Table 5
The association between relative longitude and chronotype in selected time
zones. Multiple standardized regression coefficients (β) are corrected for
age, sex, cognitive ability, latitude and log population
Time zone nβp
GMT -8 5,034 .006 .76
GMT -7 1,493 .02 .058
GMT -6 4,941 .019 .194
GMT -5 9,175 .06 <.001
GMT +0 1,762 .051 .038
GMT +1 1,546 .02 .445
GMT +8 288 .158 .024
GMT +10 466 .017 .81
Note: The associations are statistically significant when shown in bold.
6 Geography, intelligence, chronotype
© 2021 The Authors. PsyCh Journal published by Institute of Psychology, Chinese Academy of Sciences and John Wiley & Sons
Australia, Ltd.
Population density in the USA
From the USA, we had access to data on population den-
sity in addition to raw population numbers. We re-ran our
main models on this population in order to investigate a
potential independent effect of population density. We
found that higher population density is independently asso-
ciated with higher cognitive ability (β=.056, p< .001),
but it is not significantly associated with chronotype
(β=.012, p=.163). We added sex and age quartile
interaction effects to further models, but we found that the
interaction of population density never reached significance
with either variable in either model (i.e., whether cognitive
ability or chronotype was used as the dependent variable).
Discussion
In our study, we used a large archival database of dating
site users to estimate geographic effects on chronotype and
Figure 2. Mean chronotype by relative lon-
gitude (rounded to integer degrees). The ori-
entation of the figure is similar to most maps
as negative relative longitudes refer to west-
ward and positive relative longitudes to east-
ward locations. Categories “-10”and “8”
include participants with a relative longitude
or equal to or more extreme than 10 and
8 degrees, respectively. Error bars represent
95% confidence intervals
Table 6
Interaction effects and multiple standardized regression coefficients by sex. Bold values indicate significant interaction effects. All regression coefficients
are standardized
Females Males
Interaction
F
Interaction
pβ
CI
(lower)
CI
(upper) pβ
CI
(lower)
CI
(upper) p
Dependent variable: chronotype
Age .119 0.141 0.097 <.001 .192 0.206 0.177 <.001 15.215 <.001
Latitude .014 0.008 0.035 .221 .000 0.014 0.015 0.981 0.780 0.377
Relative
longitude
.055 0.077 0.033 <.001 .036 0.051 0.022 <.001 1.037 0.308
IQ .020 0.002 0.042 .072 .064 0.049 0.078 <.001 11.058 0.001
Log population .001 0.021 0.023 0.929 .029 0.015 0.044 <.001 4.506 0.034
Dependent variable: cognitive ability
Age .074 0.014 0.008 <.001 .025 0.005 0.001 .001 16.864 <.001
Latitude .031 0.001 0.003 .005 .031 0.001 0.003 <.001 0.362 .548
Relative
longitude
.008 0.003 0.007 .452 .006 0.002 0.005 .424 0.024 .878
Log population .164 0.132 0.172 <.001 .141 0.118 0.146 <.001 2.542 .111
Chronotype .020 0.002 0.042 .072 .065 0.048 0.076 <.001 9.473 .002
PsyCh Journal 7
© 2021 The Authors. PsyCh Journal published by Institute of Psychology, Chinese Academy of Sciences and John Wiley & Sons
Australia, Ltd.
cognitive ability. Since our method of estimating the latter
two phenotypes was novel and relied on non-targeted ques-
tionnaire responses instead of proper psychometric scales,
one goal of our study was to demonstrate the validity of
this approach by replicating robust findings about our phe-
notypes. We successfully replicated previous findings about
the relationship between age, sex and chronotype (Duarte
et al., 2014; Fischer et al., 2017; Paine et al., 2006; Till
Roenneberg et al., 2004; Sl
adek et al., 2020; Tonetti
et al., 2008), relative longitude and chronotype
(Giuntella & Mazzonna, 2019; Roenneberg et al., 2007;
Sl
adek et al., 2020) as well as city population and both
chronotype and cognitive ability (Abdellaoui et al., 2019;
Alexopoulos, 1997; Bass et al., 2008; Gist & Clark, 1938;
Lehmann, 1959; Sl
adek et al., 2020; Taji et al., 2019;
Teasdale et al., 1988). Notably, we also found a modest
positive relationship between cognitive ability and chro-
notype (β=.05), in line with a previous meta-analysis
(Preckel et al., 2011) and a large study (Kanazawa &
Perina, 2009). This demonstrated the validity of our
methods of estimating chronotype and cognitive ability and
allowed further fine-grained global analyses about the pos-
sible moderating effects of geography, sex and age. Impor-
tantly, the validity of our chronotype and cognitive ability
measures complement similar recent research using the
same dataset and the same statistical approach (Figueroa,
2018; Kirkegaard, 2018; Kirkegaard & Lasker, 2020)
suggesting that using IRT on non-targeted questions may
be a generally valid way of measuring psychological
phenotypes.
Most effects –specifically, the association between older
age, more easterly location and earlier chronotype; as well
as later chronotype, higher cognitive ability and larger
place of residence and higher cognitive ability –general-
ized across countries and time zones. The association
between age and chronotype and cognitive ability and place
of residence was especially robust. Universally, earlier
chronotypes in more aged participants are in line with theo-
ries that the shift of diurnal rhythms towards the earlier
hours of the day is a biological consequence of aging (Juda
et al., 2006; Roenneberg et al., 2004). The observation that
individuals from larger cities on average tend to have
Table 7
Interaction effects and multiple standardized regression coefficients by age quartile. The dependent variable is cognitive ability. All regression coefficients
are also adjusted for sex and the main effect of age quartiles. Bold values indicate significant interaction effects. All regression coefficients are
standardized
18–29 years 29–33 years 33–38 years Over 38 years
Interaction FInteraction pβpβpβpβp
Latitude .041 .001 .036 .005 .035 .005 .011 .331 0.746 .525
Relative longitude .018 .151 .001 .935 .003 .796 .008 .482 0.400 .753
Log population .183 <.001 .148 <.001 .134 <.001 .129 <.001 4.431 .004
Age .034 .007 .006 .658 .031 .011 .066 <.001 5.997 <.001
Chronotype .021 .090 .052 <.001 .050 <.001 .079 <.001 4.011 .007
Figure 3. The effect of chronotype, log population and age on cognitive ability differs as a function of age. Bar plots show multiple regression coeffi-
cients of chronotype (Panel A) log population (Panel B) or age (Panel C) as independent variables on cognitive ability as the dependent variable. Regres-
sion coefficients are adjusted for age, sex, latitude, relative longitude log population, and chronotype (except for the variable whose effect is shown). Error
bars indicate 95% confidence intervals
8 Geography, intelligence, chronotype
© 2021 The Authors. PsyCh Journal published by Institute of Psychology, Chinese Academy of Sciences and John Wiley & Sons
Australia, Ltd.
higher cognitive ability is also in line with many previous
reports about similar urban–rural IQ differences
(Alexopoulos, 1997; Gist & Clark, 1938; Lehmann, 1959;
Teasdale et al., 1988). While our cross-sectional design
cannot establish causality about the latter effect, we hypoth-
esize that both selective migration (i.e., the migration of
individuals with higher potential migrating to larger cities;
Abdellaoui et al., 2019) and causal environmental effects
(the availability of better education in larger cities;
Ritchie & Tucker-Drob, 2018) contribute to this trend.
In line with earlier research (Roenneberg et al., 2007;
Sl
adek et al., 2020), we found that respondents from lower
longitudes with the same time zone had on average later
chronotypes, a finding that generalizes across most individ-
ual time zones. These individuals experience sunset and sun-
rise times at a later clock time as a result of their more
westward location within the time zone which entrains a
later chronotype (Roenneberg et al., 2007). This effect was
consistent across most time zones with enough (n>1,000)
participants, with the notable exception of GMT -8 (Pacific
Standard Time). We note that in this time zone, most
respondents resided in a coastal strip in California, resulting
in little relative longitude variation (SD
GMT-8
=2.34
degrees, for comparison SD
GMT-7
=3.74, SD
GMT-6
=4.13,
SD
GMT-5
=4.12). In contrast to some previous studies
(Leocadio-Miguel et al., 2017; Miguel et al., 2014; Porcheret
et al., 2018), however, we found no relationship between lat-
itude and chronotype despite a broad and representative sam-
pling of latitudes from each country (Mean
USA
=38.18,
SD =5.22,range:18.23–64.84;Mean
Canada
=46.56,
SD =3.25, range: 42.05–55.76; Mean
UK
=52.2,
SD =1.37, range: 50.15–57.48;Mean
Other
=38.13,
SD =23.86,range:45.87 to 78.22).
We found that sex and age moderates some effects. Pre-
vious research (Duarte et al., 2014; Fischer et al., 2017;
Roenneberg et al., 2004; Tonetti et al., 2008) suggests that
young males have later chronotypes at a young age, but this
trend disappears or even reverses by middle age. We repli-
cated this finding and found a stronger negative relation-
ship between age and chronotype in males. We also found
that: (i) the relationship between chronotype and cognitive
ability is stronger in males; and (ii) the relationship
between chronotype and ognitive ability increases with age.
We hypothesize that these effects show a common mecha-
nism and that the relationship between cognitive ability and
chronotype is driven by social effects which are amplified
by higher age and male sex. Specifically, we hypothesize
that individuals with higher cognitive ability preferentially
assume jobs with a later or more flexible work schedule
(such as office-based or entrepreneurial jobs as opposed to
factory-based jobs, agriculture or construction) which sup-
ports or enables a later chronotype. While little research is
available on this topic, in a sample of German Mensa mem-
bers, we found that the later sleep time of gifted partici-
pants was only present during work days and fully
accounted for by later work time (Ujma et al., 2020). The
sorting of higher-ability individuals into high-prestige
(Herrnstein & Murray, 2010; Strenze, 2007) or high-
income (Lang & Kell, 2019) jobs occurs at a relatively late
stage of the career path, resulting in a delayed appearance
of the relationship between chronotype and cognitive abil-
ity. Since males are often overrepresented in high-prestige
jobs, including among those with exceptional cognitive
ability (Lubinski et al., 2014), these trends may be stronger
among males. We emphasize that while due to the size our
dataset these findings are likely to be reliable, our design
cannot unambiguously establish causation and further
research is needed to confirm our hypotheses. This also
concerns the observation that while individuals with higher
cognitive ability cluster in more populous places of resi-
dence at all ages, this tendency is somewhat weaker in
older participants (Table 7). This could be explained by a
tendency for moving into less populated suburban locales
with higher age (Johnson & Winkler, 2015).
Our study was decidedly exploratory with a limited abil-
ity to uncover causal mechanisms. However, while age and
sex differences may result from both social and biological
effects, our assumption was that geographic differences in
the correlates of either cognitive ability or chronotype are
mainly due to social effects and the effects that generalize
across countries are likely to be due to biological effects or
at least the general features of modern societies. These lat-
ter include the correlations between age and chronotype,
relative longitude and chronotype, log population and cog-
nitive ability as well as cognitive ability and chronotype
which we universally observed. Other effects, however,
were only found in certain countries. The association
between log population and chronotype varied widely
across countries, and the relationship between longitude
and cognitive ability was only present in the USA and
among “Other”countries. Concerning the first effect, we
speculate that there may be differences in the extent to
which living in a locale with a higher population introduces
lifestyle changes that contribute to a later chronotype. For
PsyCh Journal 9
© 2021 The Authors. PsyCh Journal published by Institute of Psychology, Chinese Academy of Sciences and John Wiley & Sons
Australia, Ltd.
example, if in a country even small towns are usually
tightly built up, well-illuminated and its residents princi-
pally work in similar jobs as those from large cities, then
we expect a small urban–rural chronotype divide. Con-
cerning the second effect, the presence of better socioeco-
nomic indicators in the northern areas of the USA is
well-documented, sometimes referred to as Moynihan’slaw
of proximity to the Canadian border (Moynihan, 1993),
likely due to historical and cultural effects leading to a gen-
erally higher level of social development in the North.
While some work suggests similar effects even with possi-
bly genetic reasons in other countries (Daniele, 2015;
Kura, 2013; Lynn, 2010, 2012), our findings are more in
line with cultural–geographic explanations. In the UK and
Australia, where the major economic and population cen-
ters are located in the south, higher cognitive ability is
associated with more southern latitudes (which in the latter
case also corresponds to greater distance from the equator,
like in the USA), and in Canada where most of the popula-
tion is concentrated in a thin latitudinal strip there is no
appreciable relationship.
Our work has a number of limitations. First, we did not
use psychometric tools to assess cognitive ability or chro-
notype, but instead relied on questionnaire responses.
While we did not validate these directly against psychomet-
ric tools, we successfully replicated the geographic and
demographic correlates of the phenotypes in question,
which together with previous results from the same dataset
(Figueroa, 2018; Hauser, 2018; Kirkegaard, 2018;
Kirkegaard & Lasker, 2020) demonstrate the validity of our
method. Second, our database was not representative and
the correlation between our variables of interest and likeli-
hood of participation in the database (i.e., the use of
OKCupid to find romantic partners) carried the risk of col-
lider bias (Munafò et al., 2017). For instance, if individuals
with higher cognitive ability marry later and they are there-
fore more likely to be still looking for partners on
OKCupid at a higher age, then we would see a spurious
positive correlation between participant age and cognitive
ability. While we cannot fully eliminate the presence of
such effects, we successfully replicated the findings from
more representative databases about the general relation-
ship between age, sex, chronotype, cognitive ability and
geographic parameters, suggesting at least acceptable repre-
sentativeness. We also note that our database was likely
much more representative in terms of age and education
than a substantial proportion of the psychology literature
which tends to focus on university students. Third,
OKCupid is an English-language dating website, but we
gathered data from users from multiple countries who were
likely not native English speakers. While this is unlikely to
severely bias our findings in the pooled sample as most
respondents were from English-speaking countries, varying
degree of English proficiency may affect the validity or bias
our findings from “Other”countries. Fourth, our database
was well-powered to find small effects and all of the effects
we found were modest in strength. This suggests that the
entire range of chronotype and cognitive ability is well rep-
resented at all ages and geographic locations.
Conclusions
In sum, our findings from a large archival database demon-
strate that chronotype and cognitive ability can be esti-
mated from non-targeted questionnaire data with
reasonable predictive validity. In our international dataset,
we found that these phenotypes follow an uneven and par-
tially overlapping geographic distribution. Both later chro-
notype and high cognitive ability is more common among
those residing in larger (and potentially denser) locales,
while chronotype is also influenced by relative longitude
within the time zone. These findings generalize across
countries, but they are moderated by age and sex. We
hypothesize that both biological and social effects contrib-
ute to the relationship between cognitive ability, chronotype
as well as geographic and demographic variables.
Acknowledgments
The authors declare that the current study was not finan-
cially supported by any institution or organization.
Conflict of Interest
The authors declare no conflict of interest.
References
Abdellaoui, A., Hugh-Jones, D., Yengo, L., Kemper, K. E.,
Nivard, M. G., Veul, L., Holtz, Y., Zietsch, B. P.,
Frayling, T. M., Wray, N. R., Yang, J., Verweij, K. J. H., &
Visscher, P. M. (2019). Genetic correlates of social stratifica-
tion in Great Britain. Nature Human Behaviour,3(12), 1332–
1342. https://doi.org/10.1038/s41562-019-0757-5
10 Geography, intelligence, chronotype
© 2021 The Authors. PsyCh Journal published by Institute of Psychology, Chinese Academy of Sciences and John Wiley & Sons
Australia, Ltd.
Alexopoulos, D. S. (1997). Urban vs rural residence and IQ. Psy-
chological Reports,80(3), 851–860. https://doi.org/10.2466/
pr0.1997.80.3.851
Arbabi, T., Vollmer, C., Dorfler, T., & Randler, C. (2015). The
influence of chronotype and intelligence on academic achieve-
ment in primary school is mediated by conscientiousness, mid-
point of sleep and motivation. Chronobiology International,
32(3), 349–357 https://doi.org/10.3109/07420528.2014.980508
Bass, L., K}
o, N., Kuncz, K. E., L
anyiné, E. A., Mész
aros, A.,
Mlink
o, R., Nagyné, I., & R
ozsa, S. (2008). Tapasztalatok a
WISC–IV gyermek-intelligenciateszt magyarorszagi stand-
ardizalasarol. Educatio T
arsadalmi Szolg
altat
o Kht.
Bouchard, T. J. (2013). The Wilson effect: The increase in herita-
bility of IQ with age. Twin Research and Human Genetics,
16(5), 923–930. https://doi.org/10.1017/thg.2013.54
Bruckman, A. (2002). Studying the amateur artist: A perspective
on disguising data collected in human subjects research on the
Internet. Ethics and Information Technology,4(3), 217–231.
https://doi.org/10.1023/A:1021316409277
Calvin, C. M., Batty, G. D., Der, G., Brett, C. E., Taylor, A.,
Pattie, A., Čuki
c, I., & Deary, I. J. (2017). Childhood intelli-
gence in relation to major causes of death in 68 year follow-up:
Prospective population study. BMJ,357, j2708. https://doi.org/
10.1136/bmj.j2708.
Calvin, C. M., Deary, I. J., Fenton, C., Roberts, B. A., Der, G.,
Leckenby, N., & Batty, G. D. (2011). Intelligence in youth and
all-cause-mortality: Systematic review with meta-analysis.
International Journal of Epidemiology,40(3), 626–644. https://
doi.org/10.1093/ije/dyq190
Castilhos Beauvalet, J., Luísa Quiles, C., Oliveira, M.,
Ilgenfritz, C., Hidalgo, M., & Comiran Tonon, A. (2017).
Social jetlag in health and behavioral research: A systematic
review. ChronoPhysiology and Therapy,7,19–31. https://doi.
org/10.2147/cpt.s108750
Catanese, S. A., Meo, P., Ferrara, E., Fiumara, G., & Provetti, A.
(2011). Crawling Facebook for social network analysis pur-
poses. Paper presented at the Proceedings of the International
Conference on Web Intelligence, Mining and Semantics,
Sogndal, Norway.
Clark, G., & Cummins, N. (2018). The Big Sort: Selective Migra-
tion and the Decline of Northern England, 1780-2018 (CEPR
Discussion Paper No. DP13023).
Daniele, V. (2015). Two Italies? Genes, intelligence and the Ital-
ian north–south economic divide. Intelligence,49,44–56.
https://doi.org/10.1016/j.intell.2014.12.004
DeMars, C. (2010). Item response theory. UK: Oxford University
Press.
Duarte, L., Menna-Barreto, L., Miguel, M., Louzada, F.,
Araújo, J., Alam, M., Areas, R., & Pedrazzoli, M. (2014).
Chronotype ontogeny related to gender. Brazilian Journal of
Medical and Biological Research,47(4), 316–320. https://doi.
org/10.1590/1414-431X20143001
Figueiro, M. G., Plitnick, B., & Rea, M. S. (2014). The effects of
chronotype, sleep schedule and light/dark pattern exposures on
circadian phase. Sleep Medicine,15(12), 1554–1564. https://
doi.org/10.1016/j.sleep.2014.07.009
Figueroa, S. (2018). Love, lust, and loss: Attachment style and
relationship preferences of OKCupid users. (Thesis). Florida
Atlantic University.
Fischer, D., Lombardi, D. A., Marucci-Wellman, H., &
Roenneberg, T. (2017). Chronotypes in the US –Influence of
age and sex. PLoS One,12(6), e0178782. https://doi.org/10.
1371/journal.pone.0178782
Gale, C. R., Batty, G. D., Tynelius, P., Deary, I. J., &
Rasmussen, F. (2010). Intelligence in early adulthood
and subsequent hospitalization for mental disorders.
Epidemiology,21,70–77. https://doi.org/10.1097/EDE.
0b013e3181c17da8
Gist, N. P., & Clark, C. D. (1938). Intelligence as a selective fac-
tor in rural-urban migrations. American Journal of Sociology,
44(1), 36–58.
Giuntella, O., & Mazzonna, F. (2019). Sunset time and the eco-
nomic effects of social jetlag: Evidence from US time zone
borders. Journal of Health Economics,65, 210–226. https://
doi.org/10.1016/j.jhealeco.2019.03.007
Gjoka, M., Kurant, M., Butts, C. T., & Markopoulou, A. (2010,
14-19 March 2010). Walking in Facebook: A case study of
unbiased sampling of OSNs. Paper presented at the 2010 Pro-
ceedings IEEE INFOCOM.
Hauser, K. E. (2018). Development from cupid to evolution:
Empirical support for life history and ecobiodevelopmental
frameworks. (Bachelor of Arts in Liberal Arts and Sciences).
Florida Atlantic University.
Hegelund, E., Flensborg-Madsen, T., Dammeyer, J., &
Mortensen, E. (2018). Low IQ as a predictor of unsuccessful
educational and occupational achievement: A register-based
study of 1,098,742 men in Denmark 1968–2016. Intelligence,
71,46
–53. https://doi.org/10.1016/j.intell.2018.10.002
Hegelund, E., Flensborg-Madsen, T., Dammeyer, J.,
Mortensen, L., & Mortensen, E. (2019). The influence of famil-
ial factors on the association between IQ and educational and
occupational achievement: A sibling approach. Personality
and Individual Differences,149, 100–107. https://doi.org/10.
1016/j.paid.2019.05.045
Herring, S. (1996). Linguistic and critical analysis of computer-
mediated communication: Some ethical and scholarly consider-
ations. The Information Society,12(2), 153–168. https://doi.
org/10.1080/911232343
Herrnstein, R. J., & Murray, C. (2010). The bell curve: Intelli-
gence and class structure in American life. New York: Free
Press.
Inderkum, A. P., & Tarokh, L. (2018). High heritability of adoles-
cent sleep–wake behavior on free, but not school days: A long-
term twin study. Sleep,41(3), 1–9. https://doi.org/10.1093/
sleep/zsy004
Johnson, K. M., & Winkler, R. L. (2015). Migration signatures
across the decades: Net migration by age in U.S. counties,
1950-2010. Demographic Research,32, 1065–1080. https://
doi.org/10.4054/DemRes.2015.32.38
Jones, S. E., Lane, J. M., Wood, A. R., van Hees, V. T.,
Tyrrell, J., Beaumont, R. N., Jeffries, A. R., Dashti, H. S.,
Hillsdon, M., Ruth, K. S., Tuke, M. A., Yaghootkar, H.,
Sharp, S. A., Jie, Y., Thompson, W. D., Harrison, J. W.,
Dawes, A., Byrne, E. M., Tiemeier, H., …Weedon, M. N.
(2019). Genome-wide association analyses of chronotype in
697,828 individuals provides insights into circadian rhythms.
Nature Communications,10(1), 343. https://doi.org/10.1038/
s41467-018-08259-7
PsyCh Journal 11
© 2021 The Authors. PsyCh Journal published by Institute of Psychology, Chinese Academy of Sciences and John Wiley & Sons
Australia, Ltd.
Juda, M., Münch, M., Wirz-Justice, A., Merrow, M., &
Roenneberg, T. (2006). The biological clock and sleep in the
elderly. Zeitschrift für Gerontopsychologie &-psychiatrie,
19(1), 45–51. http://doi.org/10.1024/1011-6877.19.1.45
Kanazawa, S., & Perina, K. (2009). Why night owls are more
intelligent. Personality and Individual Differences,47(7), 685–
690. https://doi.org/10.1016/j.paid.2009.05.021
Kim, A. Y., & Escobedo-Land, A. (2015). OkCupid data for
introductory statistics and data science courses. Journal of Sta-
tistics Education,23(2), 1–15. https://doi.org/10.1080/
10691898.2015.11889737
Kirkegaard, E. (2018). Self-reported criminal and anti-social
behavior on a dating site: The importance of cognitive ability.
Open Differential Psychology. https://doi.org/10.26775/ODP.
2018.01.10
Kirkegaard, E., & Bjerrekær, J. D. (2016). The OKCupid dataset:
A very large public dataset of dating site users. Open Differen-
tial Psychology. https://doi.org/10.26775/ODP.2016.11.03
Kirkegaard, E., & Lasker, J. (2020). Intelligence and religiosity
among dating site users. Psych,2(1), 25–33. https://doi.org/10.
3390/psych2010003
Knutson, K. L., & von Schantz, M. (2018). Associations between
chronotype, morbidity and mortality in the UK Biobank cohort.
Chronobiology International,35(8), 1045–1053. https://doi.
org/10.1080/07420528.2018.1454458
Kosinski, M., Matz, S. C., Gosling, S. D., Popov, V., &
Stillwell, D. (2015). Facebook as a research tool for the social
sciences: Opportunities, challenges, ethical considerations, and
practical guidelines. American Psychologist,70(6), 543–556.
https://doi.org/10.1037/a0039210
Koskenvuo, M., Hublin, C., Partinen, M., HeikkilÄ, K., &
Kaprio, J. (2007). Heritability of diurnal type: A nationwide
study of 8753 adult twin pairs. Journal of Sleep Research,
16(2), 156–162. https://doi.org/10.1111/j.1365-2869.2007.
00580.x
Kura, K. (2013). Japanese north–south gradient in IQ predicts dif-
ferences in stature, skin color, income, and homicide rate. Intel-
ligence,41, 512–516. https://doi.org/10.1016/j.intell.2013.
07.001
Lang, J. W. B., & Kell, H. J. (2019). General mental ability and
specific abilities: Their relative importance for extrinsic career
success. Journal of Applied Psychology,105, 1047–1061.
https://doi.org/10.1037/apl0000472
Lehmann, I. J. (1959). Rural-urban differences in intelligence. The
Journal of Educational Research,53,62
–68. https://doi.org/10.
1080/00220671.1959.10882612
Leocadio-Miguel, M., Louzada, F., Duarte, L., Areas, R.,
Alam, M., Freire, M., Fontenele-Araujo, J., Menna-
Barreto, L., & Pedrazzoli, M. (2017). Latitudinal cline of chro-
notype. Scientific Reports,7(1), 5437. https://doi.org/10.1038/
s41598-017-05797-w
Lubinski, D., Benbow, C., & Kell, H. J. (2014). Life paths and
accomplishments of mathematically precocious males
and females four decades later. Psychological Science,25(12),
2217–2232. https://doi.org/10.1177/0956797614551371
Lynn, R. (2010). In Italy, North–South differences in IQ predict
differences in income, education, infant mortality, stature, and
literacy. Intelligence,38(1), 93–100. https://doi.org/10.1016/j.
intell.2009.07.004
Lynn, R. (2012). North-south differences in Spain in IQ, educa-
tional attainment, per capita income, literacy, life expectancy
and employment. Mankind Quarterly,52(3/4), 265–291. https://
doi.org/10.46469/mq.2012.52.3.1.
Miguel, M., Oliveira, V. C., Pereira, D., & Pedrazzoli, M. (2014).
Detecting chronotype differences associated to latitude: A com-
parison between Horne—Ostberg and Munich Chronotype
questionnaires. Annals of Human Biology,41(2), 105–108.
https://doi.org/10.3109/03014460.2013.832795
Moynihan, D. (1993). Defining deviancy down. The American
Scholar,62(1), 17–30.
Munafò, M. R., Tilling, K., Taylor, A. E., Evans, D. M., & Davey
Smith, G. (2017). Collider scope: When selection bias can sub-
stantially influence observed associations. International Journal
of Epidemiology,47(1), 226–235. https://doi.org/10.1093/ije/
dyx206
Obschonka, M., Stuetzer, M., Rentfrow, P. J., Shaw-Taylor, L.,
Satchell, M., Silbereisen, R. K., Potter, J., & Gosling, S. D.
(2018). In the shadow of coal: How large-scale industries con-
tributed to present-day regional differences in personality and
well-being. Journal of Personality and Social Psychology,
115(5), 903–927. https://doi.org/10.1037/pspp0000175
Paine, S., Gander, P. H., & Travier, N. (2006). The epidemiology
of morningness/eveningness: Influence of age, gender, ethnic-
ity, and socioeconomic factors in adults (30-49 years). Journal
of Biological Rhythms,21(1), 68–76. https://doi.org/10.1177/
0748730405283154
Partonen, T. (2015). Chronotype and health outcomes. Current
Sleep Medicine Reports,1(4), 205–211. https://doi.org/10.
1007/s40675-015-0022-z
Plomin, R., & Deary, I. J. (2015). Genetics and intelligence differ-
ences: Five special findings. Molecular Psychiatry,20(1), 98–
108. https://doi.org/10.1038/mp.2014.105
Porcheret, K., Wald, L., Fritschi, L., Gerkema, M., Gordijn, M.,
Merrrow, M., Rajaratnam, S. M. W., Rock, D., Sletten, T. L.,
Warman, G., Wulff, K., Roenneberg, T., & Foster, R. G.
(2018). Chronotype and environmental light exposure in a stu-
dent population. Chronobiology International,35(10), 1365–
1374. https://doi.org/10.1080/07420528.2018.1482556
Preckel, F., Lipnevich, A. A., Schneider, S., & Roberts, R. D.
(2011). Chronotype, cognitive abilities, and academic achieve-
ment: A meta-analytic investigation. Learning and Individual
Differences,21(5), 483–492. https://doi.org/10.1016/j.lindif.
2011.07.003
Rahafar, A., Randler, C., Vollmer, C., & Kasaeian, A. (2017).
Prediction of school achievement through a multi-factorial
approach –The unique role of chronotype. Learning and Indi-
vidual Differences,55,69–74. https://doi.org/10.1016/j.lindif.
2017.03.008
Rahman, M. M. (2012). Mining social data to extract intellectual
knowledge. International Journal of Intelligent Systems and
Applications,4(10), 15–24. http://dx.doi.org/10.5815/ijisa.
2012.10.02.
Rentfrow, P. J., & Jokela, M. (2016). Geographical psychology:
The spatial organization of psychological phenomena. Current
Directions in Psychological Science,25(6), 393–398. https://
doi.org/10.1177/0963721416658446
Ritchie, S. J., & Tucker-Drob, E. M. (2018). How much does edu-
cation improve intelligence? A meta-analysis. Psychological
12 Geography, intelligence, chronotype
© 2021 The Authors. PsyCh Journal published by Institute of Psychology, Chinese Academy of Sciences and John Wiley & Sons
Australia, Ltd.
Science,29(8), 1358–1369. https://doi.org/10.1177/
0956797618774253
Roenneberg, T., Kuehnle, T., Pramstaller, P. P., Ricken, J.,
Havel, M., Guth, A., & Merrow, M. (2004). A marker for the
end of adolescence. Current Biology,14(24), R1038–R1039.
https://doi.org/10.1016/j.cub.2004.11.039
Roenneberg, T., Kumar, C. J., & Merrow, M. (2007). The human
circadian clock entrains to sun time. Current Biology,17, R44–
R45. https://doi.org/10.1016/j.cub.2006.12.011
Roenneberg, T., & Merrow, M. (2007). Entrainment of the human
circadian clock. Cold Spring Harbor Symposia on Quantative
Biology,72, 293–299. https://doi.org/10.1101/sqb.2007.72.043
Shawa, N., Rae, D. E., & Roden, L. C. (2018). Impact of seasons
on an individual’s chronotype: Current perspectives. Nature
and Science of Sleep,10, 345–354. https://doi.org/10.2147/
NSS.S158596
Sl
adek, M., Kudrn
acˇov
a Röschov
a, M., Ad
amkov
a, V.,
Hamplov
a, D., & Sumov
a, A. (2020). Chronotype assessment via
a large scale socio-demographic survey favours yearlong standard
time over daylight saving time in Central Europe. Scientific
Reports,10(1), 1419–1419. https://doi.org/10.1038/s41598-020-
58413-9
Strenze, T. (2007). Intelligence and socioeconomic success: A
meta-analytic review of longitudinal research. Intelligence,
35(5), 401–426. https://doi.org/10.1016/j.intell.2006.09.004
Strenze, T. (2015a). Intelligence and socioeconomic success: A
study of correlations, causes and consequences. (Doctoral dis-
sertation). Tartu University.
Strenze, T. (2015b). Intelligence and success. In S. Goldstein, D.
Princiotta, & J. A. Naglieri (Eds.), Handbook of intelligence
(pp. 405–413). New York: Springer.
Subirats, L., Reguera, N., Bañ
on, A., G
omez-Zúñiga, B.,
Minguill
on, J., & Armayones, M. (2018). Mining Facebook
data of people with rare diseases: A content-based and temporal
analysis. International Journal of Environmental Research and
Public Health,15(9), 1877. https://doi.org/10.3390/
ijerph15091877
Taji, W., Mandell, B., & Liu, J. (2019). China’s urban-rural child-
hood cognitive divide: Evidence from a longitudinal cohort
study after a 6-year follow up. Intelligence,73,1–7. https://doi.
org/10.1016/j.intell.2019.01.002
Teasdale, T. W., Owen, D. R., & SØRensen, T. I. A. (1988).
Regional differences in intelligence and educational level
in Denmark. British Journal of Educational Psychology,58(3),
307–314. https://doi.org/10.1111/j.2044-8279.1988.tb00906.x
Tonetti, L., Fabbri, M., & Natale, V. (2008). Sex difference in
sleep-time preference and sleep need: A cross-sectional survey
among Italian pre-adolescents, adolescents, and adults. Chrono-
biology International,25(5), 745–759. https://doi.org/10.1080/
07420520802394191
Ujma, P. P., Baudson, T. G., B
odizs, R., & Dresler, M. (2020).
The relationship between chronotype and intelligence: The
importance of work timing. Scientific Reports,10(1), 7105.
https://doi.org/10.1038/s41598-020-62917-9
Vetter, C., Fischer, D., Matera, J. L., & Roenneberg, T. (2015).
Aligning work and circadian time in shift workers improves
sleep and reduces circadian disruption. Current Biology,25(7),
907–911. https://doi.org/10.1016/j.cub.2015.01.064
Wittmann, M., Dinich, J., Merrow, M., & Roenneberg, T. (2006).
Social jetlag: Misalignment of biological and social time. Chro-
nobiology International,23(1–2), 497–509. https://doi.org/10.
1080/07420520500545979
PsyCh Journal 13
© 2021 The Authors. PsyCh Journal published by Institute of Psychology, Chinese Academy of Sciences and John Wiley & Sons
Australia, Ltd.
