Content uploaded by Talat Arbabi
Author content
All content in this area was uploaded by Talat Arbabi on May 05, 2016
Content may be subject to copyright.
Chronobiology International, Early Online: 1–9, (2014)
!
Informa Healthcare USA, Inc.
ISSN: 0742-0528 print / 1525-6073 online
DOI: 10.3109/07420528.2014.980508
ORIGINAL ARTICLE
The influence of chronotype and intelligence on academic
achievement in primary school is mediated by conscientiousness,
midpoint of sleep and motivation
Talat Arbabi
1
, Christian Vollmer
2
, Tobias Do
¨
rfler
2
, and Christoph Randler
1
1
Institute of Science, Technology & Geography, University of Education, Heidelberg, Germany and
2
Institute of Psychology,
University of Education, Heidelberg, Germany
Individuals differ in their timing of sleep (bed times, rise times) and in their preference for morning or evening
hours. Previous work focused on the relationship between academic achievement and these variables in
secondary school students. The main aim of the study is to investigate the relationship between chronotype
and academic achievement in 10-year-old children (n ¼ 1125) attending 4th grade of primary school. They filled a
cognitive test (Culture Fair Intelligence Test, CFT 20-R) and questions about rise times and bed times, academic
achievement, conscientiousness and motivation. We used the ‘‘scales for the assessment of learning and performance
motivation’’ (SELLMO; Skalen zur Erfassung der Lern- und Leistungsmotivation for motivation), the short version of
the Five-Factor Personality Inventory Children (FFPI-C) to measure conscientiousness, and the Composite Scale
of Morningness (CSM) to assess morningness–eveningness. Mean CSM score was 37.84 ± 6.66, midpoint of sleep
was 1:36 ± 00:25 and average sleep duration (time in bed) was 10:15 ± 0:48. Morningness orientation was positively
related to intelligence, conscientiousness and learning objectives. Eveningness orientation was related to avoidance
performance objectives and work avoidance. Early midpoint of sleep, conscientiousness and intelligence
were associated with better grades. The multivariate model showed that intelligence was the strongest predictor
of good grades. Conscientiousness, motivation, younger age and an earlier midpoint of sleep were positively related
to good grades. This is the first study in primary school pupils, and it shows that the relationship between evening
orientation and academic achievement is already prevalent at this age even when controlling for important predictors
of achievement.
Keywords:
Academic achievement, children, chronotype, conscientiousness, intelligence, morningness–eveningness,
school performance
INTRODUCTION
One striking fact in school and University students
is that late chronotype (or evening preference) is
linked with poor school or academic performance
(Dı
´
az-Morales & Escribano, 2013; Preckel et al., 2013;
Randler & Frech, 2006; Vollmer et al., 2013), but all of
these studies have been carried out in secondary schools
or in University settings. The present study is the
first to investigate this relationship in primary school
pupils. In addition, the study aims at declaring the incre-
mental validity of chronotype on grade that goes beyond
the well-established predictors of school achievement as
intelligence, conscientiousness, achievement motiv-
ation, age and sex.
Predictors of academic achievement
Chronotype
Chronotype is an individual difference reflecting the
time of day at which individuals are ‘‘at their best’’
(Adan et al., 2012). Some people (morning types, or
sometimes called ‘‘larks’’) prefer morning hours for
intellectual and physical activities. They have no prob-
lems with early rising and soon achieve their maximum
of mental and physical activity and become tired early in
the evening. In contrast, evening types, or ‘‘owls’’ feel
and perform best at late afternoon or in the evening.
They tend to have late sleep schedules, irregular waking
time, bedtime, and sleep time, and are more often
dissatisfied with their sleep (Wittmann et al., 2006).
Evening types have difficulties to get out of bed in
Correspondence: Christoph Randler, Institute of Science, Geography and Technology, University of Education, Im Neuenheimer
Feld 561-2, 69120 Heidelberg, Germany. E-mail: randler@ph-heidelberg.de
Submitted September 2, 2014, Returned for revision October 19, 2014, Accepted October 22, 2014
1
Chronobiol Int Downloaded from informahealthcare.com by 217.224.209.130 on 11/13/14
For personal use only.
the morning, and they need a longer time to have their
senses cleared. However, owls are able to work until
late in the evening and they often achieve their phys-
ical and mental acrophase during late afternoon and
evening hours.
In addition, other variables of the sleep–wake
rhythm have been used in previous work, e.g. sleep
duration (time spent sleeping) and social jetlag, that is,
a misalignment between one owns’ internal clock
(or internal rhythm) and the social rhythm/schedule.
The latter is measured in the absolute difference in
midpoint of sleep between weekdays or days with
obligation and free days, based on the idea that sleep
on free days follows the internal clock (Wittmann et al.,
2006). In school pupils, these variables are moderately
correlated with morningness–eveningness scores.
At the primary school level (up to the age of 10 years),
most children are morning oriented, i.e. active in the
morning, even at the weekend (Randler & Truc, 2014).
Adolescents shift from morningness to eveningness
around the age of puberty (12–14 years) which has
been reported in many studies (Carskadon et al., 1993;
Dı
´
az-Morales & Sorroche, 2008; Laberge et al., 2001;
Randler, 2008). This change is associated with pubertal
development (Carskadon et al., 1993). Young people
turn back towards morningness at the end of adoles-
cence, which occurs around the age of 19.5 in women
and at 21 years in men (Roenneberg et al., 2004; Tonetti
et al., 2008). Some studies have reported differences
among gender. On average, boys and men are later
chronotypes than girls and women (Adan et al., 2012).
The relationship between chronotype and academic
performance in adults and adolescents has been
examined in some studies showing that eveningness
and academic achievement are strongly and inversely
related, whereas morningness and performance are
positively related. These patterns hold for both second-
ary school children (Be¸soluk, 2011; Giannotti & Cortesi,
2002; Randler & Frech, 2009; Vollmer et al., 2013) and
University students (Be¸soluk et al., 2011; Randler &
Frech, 2006) but have not yet been investigated in
primary school pupils. Further, no study yet covered all
the confounding variables as predictors of academic
performance in addition to chronotype. We hypothesize
that chronotype is a predictor of school achievement
in primary school pupils with morningness being
positively related to better achievement.
Intelligence
Individual differences in cognitive ability are a good
single predictor of academic performance (Deary et al.,
2007; Mayes et al., 2009). The relationship between
measures of intelligence and measures of school
achievement is usually around 0.30–0.50 (Gustafsson &
Undheim, 1996; Rindermann & Neubauer, 2004; Spinath
et al., 2006).
In their meta-analysis, Preckel et al. (2011) investi-
gated the relationship between cognitive ability and
chronotype and reported a mean effect size of 0.08
between eveningness and cognitive ability and of 0.04
between morningness and cognitive ability, suggesting
that evening types are more intelligent. In detail,
some studies showed that evening types were more
intelligent (mental speed, working memory; Roberts &
Kyllonen, 1999), or that eveningness scored higher on
verbal abilities (Killgore & Killgore, 2007). Morning-
oriented students scored lower on inductive reasoning
than evening-oriented students (Dı
´
az-Morales &
Escribano, 2013). In total, Preckel et al. (2011) reported
seven positive and four negative correlations between
eveningness and cognitive ability. However, the effect
size is rather low and the fail-safe number, the number
of non-significant, potentially unpublished or missing,
studies that are needed to draw the result (effect size)
to zero was n ¼ 7, suggesting that further studies are
needed to assess this relationship.
There was a weak difference between the sexes in
intelligence but a larger variance in males (Deary, 2003;
Hedges & Nowell, 1995). Girls generally perform better
at school than boys (e.g. Burusic et al., 2012; Demie,
2001; Duckworth et al., 2006; Gibb et al., 2008; Leeson
et al., 2008; Steinmayr & Spinath, 2008), especially in
languages while boys perform better in mathematics
(e.g. Jacobs et al., 2002; Spinath et al., 2008).
It is still unclear what interplay between the vari-
ables chronotype, intelligence and school achievement
occurs at primary school age. We hypothesize a correl-
ation between chronotype and intelligence with even-
ingness being positively related to intelligence and
assume intelligence to be a strong predictor of academic
achievement in primary school.
Conscientiousness
Conscientiousness predicts academic outcomes
among school students (Bratko et al., 2006; Heaven
et al., 2002; Spinath et al., 2010; Steinmayr & Spinath,
2008; Wolfe & Johnson, 1995), undergraduates
(Busato et al., 1998; Diseth, 2003; Furnham et al., 2002;
Lounsbury et al., 2002) and postgraduates (Rothstein
et al., 1994). Conscientiousness was confirmed as the
strongest Big Five predictor of academic performance,
faring better in some samples than intelligence (cor-
rected r ¼ 0.22, meta-analysis by Poropat, 2009).
Moreover, in a young age group (between 3 and 12
years of age), boys were rated less conscientious than
girls based on parents’ statements (De Fruyt et al., 1998).
Concerning chronotype, numerous studies reported
that morning people are more conscientious (review:
Adan et al., 2012). We proclaim a mediation hypothesis:
It can be presumed that conscientiousness is the
mediating variable between chronotype and school
achievement.
Motivation
Studying the construct of intrinsic motivation in young
children is important, because academic intrinsic
2 T. Arbabi et al.
Chronobiology International
Chronobiol Int Downloaded from informahealthcare.com by 217.224.209.130 on 11/13/14
For personal use only.
motivation in the early elementary years will impact on
initial and future school achievement (Gottfried, 1990).
Furthermore, differences between boys and girls con-
cerning motivational variables like beliefs or interests
can be found (Meece et al., 2006), with a clear
interplay between interests and grades in primary
school (von Maurice et al., 2014).
The predictive validity of achievement motivation
for academic performance has been demonstrated in
several studies (Hejazi et al., 2009; Steinmayr & Spinath,
2009). It has been shown that, above and beyond
intelligence, motivation explains variance in academic
achievement (Gose et al., 1980; Schicke & Fagan, 1994;
Spinath et al., 2006). Concerning chronotype, there are
few studies that show a relationship between motivation
and chronotype. Roeser et al. (2013) showed that the
relationship between chronotype and academic
achievement is mediated by learning motivation. Their
study moved the field forward, but they included less
predictors of achievement than our present one. We
expect a positive correlation between motivation and
grades (Steinmayr & Spinath, 2009) and a mediation of
the relationship between chronotype and achievement
by motivation (Roeser et al., 2013).
The current study
The main aim of the current study was to investigate
the relationship between chronotype and academic
achievement during the 4th grade of primary school.
To our knowledge, there are no studies that investigated
the relationship between chronotype and academic
performance of primary school children, and thus, this
is a neglected age group. As the change towards
eveningness occurs mainly at the ages of 12–14 years,
we hypothesize that the correlation between chronotype
and achievement might be lower compared to older
age groups. First, because the number of evening types
in primary school pupils is lower compared to second-
ary school pupils, but there are already evening
types present in primary school. The morningness–
eveningness scores are normally distributed, so the
scores are generally shifted to morningness in primary
school. Second, the internal sleep–wake cycle of primary
school pupils better fits the social and school schedules,
suggesting a smaller misalignment between their own
internal clock and the social clock, and therefore, a
weaker correlation between achievement and chrono-
type. This could also be viewed as a better person–
environment-fit. In addition, we simultaneously mod-
eled the interplay of many of the above-mentioned
variables that have also been found to influence school
achievement.
METHODS
Participants and data collection
The sample consisted of 1125 students, 536 girls and 584
boys (five sex unspecified), aged 10.22 years (SD ¼ 0.47,
n ¼ 1117) from 48 primary schools (4th grade). On
average schools start at 8:00 in the morning. The
study took place between 15.04.2013 and 02.07.2013
and from 8:00 to 12:15 in southwest Germany (Rhine-
Neckar-Region). Mean testing time was 9:57 ± 1:03,
which is situated right in the middle of the school day.
The cognitive test and the self-administered question-
naire were completed by the students during normal
classroom settings in the presence of a researcher and
their teacher. The cognitive test was carried out in all
classes by the same researcher. Participation was
voluntary and anonymous with written consent of
parents and school administration. The present study
complied with the tenets of the Declaration of Helsinki
and the international ethical standards of chronobiolo-
gical research (Portaluppi et al., 2010).
Measures
Chronotype and sleep variables
The Composite Scale of Morningness (CSM; Smith et al.,
1989) consists of 13 questions in a Likert-type format
regarding the time individuals get up and go to bed,
preferred times for physical and mental activity, and
subjective alertness. Five of the elements of the scale
refer to different times of day. The score is obtained by
adding the items and ranges from 13 (extreme evening-
ness) to 55 (extreme morningness). The CSM score is
unaffected by the time of the day one fills in the
questionnaire. Cronbach’s alpha was 0.78.
Additionally, we asked for habitual rise time and
bed time on weekdays and on the weekend (Giannotti &
Cortesi, 2002; Russo et al., 2007). These variables are
considered as a proxy of sleep length because they focus
on total time in bed (including sleep onset latency and
bed time after awakening). From these, we calculated a
single phase-reference point, the corrected mid-sleep
point (MSFsc; Roenneberg et al., 2004). The self-report
MSFsc has been used in adults, adolescents and
children as young as 10 years of age (Roenneberg
et al., 2003). Social jetlag was calculated using the
method from Wittmann et al. (2006).
Intelligence
We used the Culture Fair Intelligence Test (CFT 20-R)
as a measure of cognitive ability. The CFT 20-R is a
German adaptation of the Culture Fair Intelligence
Test (Weib, 2008). The paper-and-pencil test assesses
fluid intelligence with four types of figural tasks: Series
(15 items), Classifications (15 items), Matrices (15 items)
and Topological Reasoning (11 items). Tasks were
presented in a multiple-choice format. Each subtest is
timed and the items increase in difficulty.
The purpose of a culture-fair intelligence test is
to minimize any social or cultural advantages, or
disadvantages, that a person may have due to their
upbringing. The CFT seems unaffected by time of day,
so a synchrony effect is not expected. To our knowledge,
this has not been tested. In our study population, testing
Chronotype, intelligence and academic achievement in primary school 3
!
Informa Healthcare USA, Inc.
Chronobiol Int Downloaded from informahealthcare.com by 217.224.209.130 on 11/13/14
For personal use only.
time correlated weakly positively with CFT scores
(r ¼ 0.095, p50.001). We consider this effect negligible
(below 1% of variance explained), and there was no
synchrony effect because time of testing did not correl-
ate with CSM scores (r ¼ 0.010) nor with midpoint of
sleep (r ¼ 0.017). Cronbach’s alpha of the total CFT (56
items) was 0.74.
Conscientiousness
The short version of the Five-Factor Personality
Inventory-Children (FFPI-C; McGhee et al., 2007) was
used for the measurement of conscientiousness.
The scale consists of 15 bipolar pairs of sentences on a
5-point Likert-scale. Cronbach’s alpha in the present
study was 0.73.
Motivation
Achievement motivation was measured by SELLMO,
the ‘‘Skalen zur Erfassung der Lern- und
Leistungsmotivation’’ (scales for the assessment of
learning and performance motivation; Spinath et al.,
2002). It contains of 4 scales and 31 items and is suitable
for use in primary school (Swoboda, 2010). The response
scale was a 5-point Likert scale ranging from ‘‘not true
at all’’ (1) to ‘‘exactly true’’ (5). Cronbach’s alpha
was 0.68 for ‘‘learning objectives’’, 0.75 for ‘‘approach
performance objectives’’, 0.81 for ‘‘avoidance perform-
ance objectives’’ and 0.80 for ‘‘work avoidance’’. The
‘‘learning objectives’’ scale describes the goal of wanting
to expand one’s own abilities. The ‘‘approach perform-
ance objectives’’ scale describes the goal of wanting to
demonstrate one’s skills in front of others, a property
associated with somewhat short-term learning success,
but without ensuring adequate long-term learning
success. The ‘‘avoidance performance objectives’’
scale describes the tendency to try to hide low skills
or inability/ignorance due to previous negative experi-
ences; a property associated with poor short and long-
term benefits. The ‘‘work avoidance’’ behavior is not
learning or performance motivated, i.e. the motivation
to invest as little effort as possible. This attitude has a
particularly negative effect on interest and intrinsic
motivation.
Academic achievement
School performance was measured by self-reported
grades. Students reported their half year grades
(February 2013) in Mathematics, German, English and
Nature & Culture (a combined elementary school sub-
ject including fine arts, music, biology and culture) on a
21-point scale from 1.0 ¼ fail, 1.25, 1.5, [...] 5.5, 5.75 to
6.0 ¼ outstanding and thus, higher scores in grades
indicate higher achievement. Thus, self-reported
grades do not reflect grades from a single test but
represent accumulations of attainments of a whole
school term. Further, school grades are real measure-
ments that have an influence on career decisions.
Research suggests that self-reported school grades can
be assumed to be valid, since they do not seem to be
subject to systematic bias (Dickha
¨
user & Plenter, 2005).
Statistical analyses
SPSS 21 and AMOS 21 (both IBM, Somers, NY) were
used for statistical calculation. t-Tests and Pearson’s
correlations were used to analyze the bivariate relation-
ships between all variables under study. Structural
equation modeling (SEM) was used to explore associ-
ations between variables in context. Gender differences
were included in a group analysis to investigate gender
as a moderator variable. Specification search in AMOS
21 was used with associations between variables with
50.20 specified as optional for best model fit. Missing
values were substituted with estimates in the multivari-
ate analyses.
RESULTS
Descriptive statistics for the sleep–wake and chronotype
variables are depicted in Table 1.
Bivariate analyses
Boys and girls did not differ in chronotype and intelli-
gence (Table 2). There were significant differences
between the sexes in academic performance but the
direction was subject-specific: girls did better in
Languages and Nature & Culture, but boys had better
scores in Mathematics. Overall, there were no gender
differences in grades. There were significant gender
differences in midpoint of sleep with girls sleeping later
and in social jetlag with girls sharing more social jetlag.
Girls scored higher on conscientiousness and learning
objectives than boys. Boys scored higher on avoidance
performance objectives. There were no gender differ-
ences in work avoidance (Table 2).
There was no significant correlation of chronotype
as measured by the CSM and age, but midpoint of
sleep was significantly related to age with a later
midpoint at an older age (Table 3). Evening-oriented
individuals shared higher social jetlag. Morningness
orientation was positively related with higher intelli-
gence, higher scores in conscientiousness and learning
objectives. Eveningness orientation was related to higher
avoidance performance objectives and higher work
avoidance. Age correlated negatively with intelligence.
TABLE 1. Descriptive sleep–wake variables of the sample.
Mean SD
Chronotype (CSM) 37:84 6:66
Midpoint of sleep before school days 01:36 0:25
Midpoint of sleep at the weekend 03:23 0:59
Midpoint of sleep (MSFsc) 02:58 0:57
Sleep duration before school days (time in bed) 10:10 0:47
Sleep duration on weekends (time in bed) 10:27 1:32
Average sleep duration (time in bed) 10:15 0:48
Social jetlag 1:46 0:54
4 T. Arbabi et al.
Chronobiology International
Chronobiol Int Downloaded from informahealthcare.com by 217.224.209.130 on 11/13/14
For personal use only.
Students with later midpoint of sleep scored lower
in intelligence. Higher intelligence scores were posi-
tively related to higher conscientiousness. Age was not
significantly correlated with conscientiousness.
Conscientiousness was positively related to better
grades. Older students reported higher approach per-
formance objectives, higher avoidance performance
objectives, and higher work avoidance. There were no
age differences in learning objectives.
Age had significant influence on grades. Grades
decreased with increasing age and younger children
were earlier chronotypes than older children as mea-
sured by the midpoint of sleep. Both conscientious-
ness and intelligence were associated with good grades.
Good grades, in turn, correlated with the four aspects
of motivation. Furthermore, better grades were related
to an early midpoint of sleep, lower social jetlag and
younger age.
Multivariate analyses
Goodness of fit statistics of the SEM revealed that
the overall model (M1) and the unconstrained gen-
der group analysis (M2) fitted best (Table 4). The
specification search of the SEM removed two facets of
motivation, resulting in a single factor of motivation
labeled ‘‘negative motivation’’ (avoidance performance
objectives and work avoidance).
Higher intelligence was the strongest predictor of
good grades. Moreover, conscientiousness, motivation,
younger age and an earlier midpoint of sleep were
positively related to good grades. Although earlier
CSM score was associated with good grades in bivari-
ate analysis, CSM scores did not directly contribute
to differences in grades in the SEM. However, chron-
otype contributed to grades mediated by midpoint of
sleep and conscientiousness. Whereas intelligence con-
tributed on a direct path to grades, intelligence also
contributed indirectly by motivation, conscientiousness
and midpoint of sleep (Figure 1).
DISCUSSION
The strength of the study is that it controls for many
co-variates and predictors of academic achievement
to unveil the effects of chronotype on academic
achievement. Nevertheless, an effect of chronotype on
achievement remained significant. Considering all these
variables, higher intelligence contributed the most to
good grades. This goes in line with most studies and is a
well-known fact. The negative association between
age and intelligence can be explained by the German
school system where gifted children are sent to school
earlier (sometimes around the age of 5 years, whereas
less skilled pupils start schooling around the age of
7 years), or they skip a grade and thus are younger in our
population, the 4th grade of primary school. Older and
less conscientious children with lower motivation and a
later sleep rhythm perform worse at school.
In line with previous work (Dı
´
az-Morales & Sorroche,
2008), we found that earlier chronotype was associated
with earlier midpoint of sleep and less social jetlag,
showing that individual circadian preferences are mani-
fest, and, thus late chronotype can be detrimental to
early school schedules already in pre-adolescent chil-
dren. The mean CSM score was 37.84, and thus much
more shifted towards morningness compared, e.g. to
adolescents (e.g. about 30–32 at age of 15–17 years;
Randler, 2011). Similarly, midpoint of sleep was
very early at 1:36 compared to an average of 4:28 in
about 14-year-old adolescents (Vollmer et al., 2012).
This indicates that primary school pupils in grade 4 are
more morning oriented compared to adolescents.
Generally, the influence of chronotype on academic
achievement is lower compared to studies based on
TABLE 2. Means, standard deviations of study variables and gender differences (t-test).
Girls Boys t-Test Total
Mean SD n Mean SD nt p Mean SD n
Chronotype (CSM) 37.99 6.64 536 37.65 6.65 581 0.862 0.389 37.81 6.65 1117
Midpoint of sleep (MSFsc) 3:02 am 53 min 526 2:54 am 60 min 566 2.510 0.012 2:58 am 57 min 1092
Social jetlag 1 h 51 min 52 min 526 1 h 41 min 56 min 566 3.094 0.002 1 h 46 min 55 min 1092
Intelligence (CFT total) 0.55 0.10 536 0.55 0.11 584 1.048 0.295 0.55 0.10 1120
Conscientiousness 3.79 0.51 534 3.68 0.50 582 3.766 50.001 3.73 0.51 1116
Motivation: Learning objectives 4.24 0.55 536 4.17 0.59 580 2.065 0.039 4.20 0.57 1116
Motivation: Approach performance
objectives
3.37 0.79 535 3.45 0.79 580 1.744 0.081 3.42 0.79 1115
Motivation: Avoidance
performance objectives
2.73 0.95 535 2.86 0.90 580 2.438 0.015 2.80 0.92 1115
Motivation: Work avoidance 2.62 0.89 535 2.70 0.95 580 1.425 0.154 2.66 0.92 1115
Average grades 4.98 0.50 499 4.92 0.54 531 1.937 0.053 4.95 0.52 1030
Grades, Mathematics 4.75 0.79 510 4.90 0.74 541 3.242 0.001 4.83 0.77 1051
Grades, German 4.87 0.64 504 4.72 0.78 534 3.332 0.001 4.79 0.72 1038
Grades, Science & Culture 5.10 0.54 488 4.94 0.61 522 4.310 50.001 5.02 0.58 1010
Grades, English 5.24 0.56 505 5.12 0.66 537 3.063 0.002 5.18 0.61 1042
CSM, Composite Scale of Morningness; CFT, Cultural Fair Test. Significant differences are depicted in bold.
Please note that a conservative approach would focus on significant variables only (p50.01).
Chronotype, intelligence and academic achievement in primary school 5
!
Informa Healthcare USA, Inc.
Chronobiol Int Downloaded from informahealthcare.com by 217.224.209.130 on 11/13/14
For personal use only.
secondary school pupils and University students in
Germany. For example, Vollmer et al. (2013) reported a
correlation coefficient between achievement and morn-
ingness–eveningness of 0.227 and Randler & Frech
(2009) of 0.182 in secondary school pupils, while
Randler & Frech (2006) reported a coefficient of 0.230
in University students. This might have several reasons:
one might lie in the nature of our study where we
have controlled for many other important predictors
of school achievement thus decreasing the effect size
of the bivariate relationship. Another reason might lie in
the developmental aspects: young people have their
strong transition to eveningness around the age of 12–14
years, thus their delayed sleep pattern is less obvious in
primary school. The sleep–wake schedule of primary
pupils does not differ so much between weekdays and
weekends, so their internal biological rhythm better fits
the school schedules. However, this was one of the
reasons why we carried out the study: we wanted to test
whether the association between chronotype and
achievement is already prevalent in this neglected age
group with a smaller difference in misalignment.
Nevertheless, there are evening-type pupils already in
primary school although their proportion is lower
compared to adolescent samples.
Another possibility might lie in the school start times.
School start times in primary school might be a bit later
(albeit only up to half an hour), but taken together with
the fact that pupils at this age are more morning
oriented and may go to school a bit later, these factors
may be responsible for the lower correlation between
morningness–eveningness and achievement in primary
pupils.
In contrast to previous studies (Killgore et al., 2007;
Roberts & Kyllonen, 1999), morning orientation was
positively related with higher intelligence in bivariate
analyses. This could be based on the samples, because
we assessed school children in primary school (thus a
different age) and further, a more representative sample,
since Roberts & Kyllonen (1999) were based on army
people and Killgore & Killgore (2007) relied on a small
sample size. However, the results are not contradictory
because Preckel et al. (2011) reported four studies with a
positive correlation between cognitive ability and
morningness, and their main effect size is low so that
the result should be treated with caution. Unfortunately,
exactly these four studies with a positive relation-
ship between morningness and intelligence are unpub-
lished (for details, see Preckel et al., 2011), so we here
conclude that the relationship between cognitive ability
or intelligence on the one hand and chronotype on the
other is far from being resolved. The few studies focus
on different populations (pupils, students and army
people) and use different measures of cognitive ability.
Morning orientation was related to proactive behav-
ior such as a higher conscientiousness and higher
learning objectives, which is important for school
performance, while late midpoint of sleep was related
TABLE 3. Correlations of study variables. Please note that a conservative approach would focus on highly significant variables only (p 50.001, ***).
123456789101112
1 Average grades 1.000 ***0.135 0.042 ***0.217 ***0.198 ***0.266 ***0.370 ***0.225 ***0.114 ***0.283 ***0.281 ***0.360
2 Chronotype (CSM) 1.000 *0.063 ***0.327 ***0.306 0.010 *0.061 ***0.149 0.037 **0.096 ***0.125 ***0.355
3 Average sleep length 1.000 0.045 **0.083 ***0.140 0.036 0.030 0.034 0.057 0.040 0.027
4 Midpoint of Sleep (MSFsc) 1.000 ***0.833 ***0.139 *0.140 *0.062 **0.098 ***0.117 ***0.106 0.056
5 Social jetlag 1.000 ***0.129 *0.152 *0.072 **0.098 ***0.121 ***0.121 0.054
6 Age (in months) 1.000 0.122 **0.029 **0.091 ***0.106 ***0.142 0.042
7 Intelligence (CFT total) 1.000 **0.103 *0.073 ***0.136 ***0.148 ***0.131
8 Motivation: Learning objectives 1.000 ***0.262 0.003 ***0.156 ***0.332
9 Motivation: Approach performance objectives 1.000 ***0.610 ***0.420 ***0.127
10 Motivation: Avoidance performance objectives 1.000 ***0.635 ***0.113
11 Motivation: Work avoidance 1.000 ***0.179
12 Conscientiousness 1.000
Pearson’s correlation coefficients, *Significance50.05. **Significance50.01. ***Significance50.001.
6 T. Arbabi et al.
Chronobiology International
Chronobiol Int Downloaded from informahealthcare.com by 217.224.209.130 on 11/13/14
For personal use only.
to higher avoidance performance objectives and higher
work avoidance. This suggests that the ‘‘morning per-
sonality’’ indeed is already prevalent in young children
at the primary level, and, that these factors are respon-
sible for academic achievement in primary school.
Further studies might assess trajectories of personality
and chronotype in combination with progress of the
schooling in a prospective study.
We found a correlation between age and chronotype
only in the midpoint of sleep measure but not with the
CSM scores. Further, we found differences in sleep
timing and social jetlag between boys and girls. Both
aspects taken together suggest that it might be a
possible effect of developmental age, thus further
studies might consider developmental age and compare
it with chronological age. Second, girls being more
evening oriented may be a result of the fact that girls are
on average more advanced in their development.
In addition, we found an unexpected time of testing
effect on CFT scores (cognitive ability). This result does
Age
Chronotype (CSM)
Grades
R
2
= .500
(.499/.514)
Midpoint of sleep (MSFsc)
R
2
= .146 (.154/.144)
Conscienousness
R
2
= .143 (.131/.151)
Negave
movaon
R
2
= .101
(.070/.125)
Intelligence
R
2
= .030
(.035/.025)
Avoidance performance objecves
Work avoidance
Nature & Culture
Mathemacs
German
English
CFT Series
CFT Classificaons
CFT Matrices
CFT Condions
β = –.191 (–.135/–.231)
β = –.353 (–.364/–.339)
β = .293 (.232/.350)
β = –.172 (–.188/–.159)
β = .115 (.100/.140)
β = –.161 (–.121/–.188)
β
= .107 (n.s./.102)
β = .351 (.339/.360)
β = –.149 (–.122/–.166)
β = .140 (.128/.146)
β
= .083 (.098/n.s.)
β = –.145 (–.153/–.147)
β
= –.108 (n.s./–.135)
β = .428 (.446/.424)
β
= –.189
(–.286/–.096)
β
= .766
(.807/.699)
β = .828 (.799/.903)
β = .615 (.621/.625)
β = .457 (.443/.464)
β = .328 (.260/.380)
β = .641 (.559/.688)
β = .694 (.752/.635)
β = .686 (.713/.717)
β = .620
(.586/.628)
β = .715 (.690/.730)
β = .097 (.088/.093)
FIGURE 1. Influence chronotype and intelligence on grades with conscientiousness, midpoint of sleep and motivation as mediators
and gender as moderator variable, structural equation model. Note: Significant regression coefficients ( ) from the overall model
were included: overall model (M1), in brackets: moderator variable (girls/boys; unconstrained model M2). Age in months; chronotype
(CSM, Composite Scale of Morningness) from 13 ¼ extreme evening type to 55 ¼ extreme morning type; midpoint of sleep (MSFsc) in clock
times; intelligence: Cultural Fair Test (CFT) with higher values indicating higher intelligence, conscientiousness (FFPI-C) with higher values
indicating higher conscientiousness; negative motivation (2 facets from SELLMO) with lower values indicating higher motivation; grades
from 1 ¼ fail to 6 ¼ outstanding.
TABLE 4. Goodness of fit statistics of the structural equation model.
Overall model
2
2
/df RMSEA CFI
M1 125.388 1.929 0.029 0.979
Multiple group (boys/girls) comparison
M2: Unconstrained 204.471 1.573 0.023 0.974
M3: Invariance of measurement weights 227.350 1.636 0.024 0.970
M4: Invariance of measurement intercepts 299.839 1.986 0.030 0.949
M5: Invariance of structural weights 336.117 2.037 0.030 0.941
M6: Invariance of structural intercepts 340.664 2.040 0.030 0.940
M7: Invariance of structural residuals 355.776 2.068 0.031 0.937
M8: Invariance of measurement residuals 411.484 2.236 0.033 0.922
2
¼ Chi-square; df ¼ degrees of freedom; RMSEA ¼ root mean square error of approximation; CFI ¼ comparative
fit index. Parameters are constrained to be equal for both groups (boys/girls). M1 and M2 do not differ
significantly. Other models (M3–M8) fitted significantly worse than M1 and M2.
Chronotype, intelligence and academic achievement in primary school 7
!
Informa Healthcare USA, Inc.
Chronobiol Int Downloaded from informahealthcare.com by 217.224.209.130 on 11/13/14
For personal use only.
not affect the study because of its small size (below 1%
of variance explained) and should be therefore unim-
portant for survey studies but might become important
for individual diagnostics.
Limitations
We did not assess all variables that were related to
school achievement, e.g. the need for cognition was
not assessed although it might have an influence on
grades because there was no instrument applicable for
primary school – the instrument by Preckel et al. (2013)
for 5th and 6th graders was published after the study
was carried out and should be taken into account in
future work.
CONCLUSIONS AND IMPLICATIONS
Concerning the grades, intelligence, conscientiousness
and motivation were important predictors. The results
further show that these important predictors have to
be taken into account when assessing the relationship
between chronotype and academic achievement.
Nevertheless, chronotype was an important predictor
of school achievement even when controlling for many
confounding variables. In addition, the relationship
between academic achievement and chronotype was
weaker in primary school students, probably because
they are not yet in their transition to evening types,
which occurs around the age of 12–14 years (Adan et al.,
2012). The internal sleep–wake cycle of the primary
school pupils, therefore, better fits the social and school
schedules, suggesting a smaller misalignment between
their own internal clock and the social clock, and
therefore, a weaker correlation between achievement
and chronotype. One implication of the study could be
to reduce the misalignment of adolescents (and hence
improve their person–environment-fit), which are pre-
dominantly evening types, and to start school later in
adolescents to better fit the internal clocks of the
evening types. An implication for primary school
pupils would be to carefully check school start times
and time for travelling to school (which is different
among the many schools) to avoid early getting up
times. Further, as the CFT was weakly related to testing
time, we suggest to write examinations in primary
school pupils later during the day, e.g. at 10:00 o’clock,
and not in the first lesson.
ACKNOWLEDGEMENTS
We are grateful to all pupils, parents, teachers and
principals for supporting our study.
DECLARATION OF INTEREST
The authors report no conflicts of interest. The authors
alone are responsible for the content and writing of this
paper.
REFERENCES
Adan A, Archer SN, Hidalgo MP, et al. (2012). Circadian typology:
A comprehensive review. Chronobiol Int. 29:1153–75.
Be¸soluk ¸S. (2011). Morningness–eveningness preferences and
university entrance examination scores of high school students.
Pers Indiv Differ. 50:248–52.
Be¸soluk ¸S,O
¨
nder
_
I, Deveci
_
I. (2011). Morningness–eveningness
preferences and academic achievement of university students.
Chronobiol Int. 28:118–25.
Bratko D, Chamorro-Premuzic T, Saks Z. (2006). Personality and
school performance: Incremental validity of self- and peer-
ratings over intelligence. Pers Indiv Differ. 41:131–42.
Burusic J, Babarovic T, Seric M. (2012). Differences in elementary
school achievement between girls and boys: Does the teacher
gender play a role? Eur J Psychol Educ. 27:523–38.
Busato VV, Prins FJ, Elshout JJ, Hamaker C. (1998). The relation
between learning styles, the Big Five personality traits and
achievement motivation in higher education. Pers Indiv Differ.
26:129–40.
Carskadon MA, Vieira C, Acebo C. (1993). Association between
puberty and delayed phase preference. Sleep. 16:258–62.
De Fruyt F, van Hiel A, Buyst V. (1998). Parental personal-
ity descriptors of boys and girls. In: Kohnstamm GA,
Halverson Jr, CF, Mervielde I, Havill VL, eds. Parental descrip-
tions of child personality: Developmental antecedents of the
Big Five? Hillsdale, NJ: Erlbaum, pp. 155–67.
Deary I. (2003). Population sex differences in IQ at age 11: The
Scottish mental survey 1932. Intelligence. 31:533–42.
Deary IJ, Strand S, Smith P, Fernandes C. (2007). Intelligence and
educational achievement. Intelligence. 35:13–21.
Demie F. (2001). Ethnic and gender differences in educational
achievement and implications for school improvement strate-
gies. Educ Res. 43:91–106.
Dı
´
az-Morales JF, Escribano C. (2013). Predicting school achieve-
ment: The role of inductive reasoning, sleep length and
morningness–eveningness. Pers Indiv Differ. 55:106–11.
Dı
´
az-Morales JF, Sorroche MG. (2008). Morningness–eveningness
in adolescents. Span J Psychol. 11:201–6.
Dickha
¨
user O, Plenter I. (2005). ‘‘Letztes Halbjahr stand ich zwei’’.
Zur Akkuratheit selbst berichteter Noten [On the accuracy of
self-reported school marks]. Z Pa
¨
dagog Psychol. 19:219–24.
Diseth A. (2003). Personality and approaches to learning as
predictors of academic achievement. Eur J Pers. 17:143–55.
Duckworth AL, Seligman ME, Martin EP. (2006). Self-discipline
gives girls the edge: Gender in self-discipline, grades, and
achievement test scores. J Educ Psychol. 98:198–208.
Furnham A, Chamorro-Premuzic T, McDougall F. (2002). Person-
ality, cognitive ability, and beliefs about intelligence as pre-
dictors of academic performance. Learn Individ Differ. 14:47–64.
Giannotti F, Cortesi F. (2002). Sleep patterns and daytime func-
tioning in adolescence: An epidemiological survey of an Italian
high school student sample. In: Carskadon MA, ed. Adolescent
sleep patterns. Biological, social, and psychological influences.
Cambridge: Cambridge University Press, pp. 132–47.
Gibb SJ, Fergusson DM, Horwood LJ. (2008). Gender differences in
educational achievement to age 25. Aust J Educ. 52:63–80.
Gose A, Wooden S, Muller D. (1980). The relative potential of
self-concept and intelligence as predictors of achievement.
J Psychol. 104:279–87.
Gottfried AE. (1990). Academic intrinsic motivation in young
elementary school children. J Educ Psychol. 82:525–38.
Gustafsson J, Undheim JO. (1996). Individual differences in
cognitive functions. In: Berliner DC, Calfee RC, eds.
Handbook of educational psychology. New York: Prentice
Hall, pp. 186–242.
Heaven PC, Mak A, Barry J, Ciarrochi J. (2002). Personality and
family influences on adolescent attitudes to school and self-
rated academic performance. Pers Indiv Differ. 32:453–62.
8 T. Arbabi et al.
Chronobiology International
Chronobiol Int Downloaded from informahealthcare.com by 217.224.209.130 on 11/13/14
For personal use only.
Hedges L, Nowell A. (1995). Sex differences in mental test scores,
variability, and numbers of high-scoring individuals. Science.
269:41–5.
Hejazi E, Shahraray M, Farsinejad M, Asgary A. (2009). Identity
styles and academic achievement: Mediating role of academic
self-efficacy. Social Psychol Educ. 12:123–35.
Jacobs JE, Lanza S, Osgood DW, et al. (2002). Changes in children’s
self-competence and values: Gender and domain differences
across grades one through twelve. Child Dev. 73:509–27.
Killgore WDS, Killgore DB. (2007). Morningness–eveningness
correlates with verbal ability in women but not in men.
Percept Motor Skill. 104:335–8.
Killgore WDS, Killgore DB, Day LM, et al. (2007). The effects of 53
hours of sleep deprivation on moral judgement. Sleep. 30:
345–52.
Laberge L, Petit D, Simard C, et al. (2001). Development of sleep
patterns in early adolescence. J Sleep Res. 10:59–67.
Leeson P, Ciarrochi J, Heaven PC. (2008). Cognitive ability,
personality, and academic performance in adolescence. Pers
Indiv Differ. 45:630–5.
Lounsbury JW, Sundstrom E, Loveland JL, Gibson LW. (2002).
Broad versus narrow personality traits in predicting aca-
demic performance of adolescents. Learn Individ Differ. 14:
65–76.
Mayes SD, Calhoun SL, Bixler EO, Zimmerman DN. (2009). IQ and
neuropsychological predictors of academic achievement. Learn
Individ Differ. 19:238–41.
McGhee RL, Ehrler DJ, Buckhalt JA. (2007). FFPI-C: Five-Factor
Personality Inventory — Children. Austin, TX: PRO-ED.
Meece JL, Glienke BB, Burg S. (2006). Gender and motivation.
J School Psychol. 44:351–73.
Poropat AE. (2009). A meta-analysis of the five-factor model of
personality and academic performance. Psychol Bull. 135:
322–38.
Portaluppi F, Smolensky MH, Touitou Y. (2010). Ethics and
methods for biological rhythm research on animals and
human beings. Chronobiol Int. 27:1911–29.
Preckel F, Lipnevich AA, Schneider S, Roberts RD. (2011).
Chronotype, cognitive abilities, and academic achievement:
A meta-analytic investigation. Learn Individ Differ. 21:483–92.
Preckel F, Lipnevich AA, Boehme K, et al. (2013). Morningness–
eveningness and educational outcomes: The lark has an
advantage over the owl at high school. British J Educ Psychol.
83:114–34.
Randler C. (2008). Psychometric properties of the German version
of the composite scale of morningness. Biol Rhythm Res. 39:
151–61.
Randler C. (2011). Age and gender differences in morningness–
eveningness during adolescence. J Genetic Psychol. 172:302–8.
Randler C, Frech D. (2006). Correlation between morningness–
eveningness and final school leaving exams. Biol Rhythm Res.
37:233–9.
Randler C, Frech D. (2009). Young people’s time-of-day prefer-
ences affect their school performance. J Youth Stud. 12:653–67.
Randler C, Truc Y. (2014). Adaptation of the composite scale of
morningness for parent report and results from kindergarten
children. Swiss J Psychol. 73:35–9.
Rindermann H, Neubauer AC. (2004). Processing speed, intelli-
gence, creativity, and school performance: Testing of causal
hypotheses using structural equation models. Intelligence. 32:
573–89.
Roberts RD, Kyllonen PC. (1999). Morningness–eveningness and
intelligence: Early to bed, early to rise will likely make you
anything but wise! Pers Indiv Differ. 27:1123–33.
Roenneberg T, Kuehnle T, Pramstaller P, et al. (2004). A marker for
the end of adolescence. Curr Biol. 14:R1038–9.
Roenneberg T, Wirz-Justice A, Merrow M. (2003). Life between
clocks: Daily temporal patterns of human chronotypes. J Biol
Rhythm. 18:80–90.
Roeser K, Schlarb AA, Ku
¨
bler A. (2013). The Chronotype-Academic
Performance Model (CAM): Daytime sleepiness and learning
motivation link chronotype and school performance in adoles-
cents. Pers Indiv Differ. 54:836–40.
Rothstein M, Paunonen S, Rush J, King G. (1994). Personality and
cognitive ability predictors of performance in graduate business
school. J Educ Psychol. 86:516–30.
Russo PM, Bruni O, Lucidi F, et al. (2007). Sleep habits and
circadian preference in Italian children and adolescents. J Sleep
Res. 16:163–9.
Schicke M, Fagan TK. (1994). Contributions of self-concept and
intelligence to the prediction of academic achievement among
grade 4, 6, and 8 students. J School Psychol. 10:62–9.
Smith CS, Reilly C, Midkiff K. (1989). Evaluation of three circadian
rhythm questionnaires with suggestions for an improved
measure of morningness. J Appl Psychol. 74:728–38.
Spinath B, Freudenthaler HH, Neubauer AC. (2010). Domain-
specific school achievement in boys and girls as predicted by
intelligence, personality and motivation. Pers Indiv Differ. 48:
481–6.
Spinath B, Spinath FM, Harlaar N, Plomin R. (2006). Predicting
school achievement from general cognitive ability,
self-perceived ability, and intrinsic value. Intelligence. 34:
363–74.
Spinath B, Stiensmeier-Pelster J, Scho
¨
ne C, Dickha
¨
user O. (2002).
SELLMO: Skalen zur Erfassung der Lern- und
Leistungsmotivation. Manual [SELLMO: Scales for the assess-
ment of learning motivation and performance motivation.
Manual]. Go
¨
ttingen: Hogrefe.
Spinath FM, Spinath B, Plomin R. (2008). The nature and
nurture of intelligence and motivation in the origins of sex
differences in elementary school achievement. Eur J Pers. 22:
211–29.
Steinmayr R, Spinath B. (2008). Sex differences in school achieve-
ment: What are the roles of personality and achievement
motivation? Eur J Pers. 22:185–209.
Steinmayr R, Spinath B. (2009). The importance of motivation as
a predictor of school achievement. Learn Individ Differ. 19:
80–90.
Swoboda E. (2010). Motivation via natural differentiation in
mathematics (Project report). Institute of Mathematics,
University of Rzeszo
´
w, Rzeszo
´
w, Poland.
Tonetti L, Pasquini F, Fabbri M, et al. (2008). Comparison of two
different actigraphs with polysomnography in healthy young
subjects. Chronobiol Int. 25:145–53.
Vollmer C, Michel U, Randler C. (2012). Outdoor light at night
(LAN) is correlated with eveningness in adolescents.
Chronobiol Int. 29:502–8.
Vollmer C, Po
¨
tsch F, Randler C. (2013). Morningness is associated
with better gradings and higher attention in class. Learn Individ
Differ. 27:167–73.
von Maurice J, Do
¨
rfler T, Artelt C. (2014). The relation between
interests and grades: Path analyses in primary school age. Int J
Educ Res. 64:1–11.
Weib RH. (2008). Grundintelligenztest Skala 2 (CFT 20-R).
Handanweisung [Culture Fair Test, Scale 2 (CFT 20-R)].
Handbook]. Go
¨
ttingen: Hogrefe.
Wittmann M, Dinich J, Merrow M, Roenneberg T. (2006). Social
jetlag: Misalignment of biological and social time. Chronobiol
Int. 23:497–509.
Wolfe R, Johnson S. (1995). Personality as a predictor of college
performance. Educ Psychol Meas. 55:177–85.
Chronotype, intelligence and academic achievement in primary school 9
!
Informa Healthcare USA, Inc.
Chronobiol Int Downloaded from informahealthcare.com by 217.224.209.130 on 11/13/14
For personal use only.
