You are on page 1of 23

Manual skill, hand skill asymmetry, and cognitive

performances in young children


Georges Dellatolas, Maria De Agostini, Florence Curt, Helgard
Kremin, Alexia Letierce, Jean Maccario, and Joseph Lellouch
INSERM U.472, Epidemiologie et Biostatistiques, Villejuif, France
A total of 1022 children aged 3 to 6 years were examined in their preschools and
27% of them were followed up for 2 years. A computerised version of the peg-
moving task was used repeatedly to assess hand skill of the dominant and the
nondominant hand. Cognitive performance was repeatedly evaluated by tasks
involving speech, vocabulary, phonological memory, and visual-spatial skills.
Results showed that: (i) age, sex, and handedness effects on hand skill asymmetry
generally confirmed previous reports, especially by Annett (2002); (ii) visual-
spatial and vocabulary tasks were significantly related to hand skill but speech and
phonological memory tasks were not, and the role of the dominant and non-
dominant hand were similar; (iii) overall, manual laterality indexes were only
weakly associated to some cognitive abilities; (iv) early manual skill was more
strongly associated to cognitive tasks than later manual skill. These results fit the
assumption of a significant role of early manual behaviour in aspects of cognitive
development not relying exclusively on phonology, and raise questions about
cognitive development and rehabilitation of children with early occurring manual
deficiencies.
There are many studies investigating links between handedness or hand skill
asymmetry and cognitive performance in normal preschool and school children
(Annett, 2002; Bishop, 1990; Crow, Crow, Done, & Leask, 1998; De Agostini &
Dellatolas, 2001; Hardyck, Petrinovich, & Goldman, 1976; Natsopoulos,
Kiosseoglou, Xeromeritou & Alevriadou, 1998). Findings from these studies
remain debatable. When normal children are simply classified into two groups,
left- and right-handers, according to the hand used more often on a hand
LATERALITY, 2003, 8 (4), 317338
Address correspondence to Georges Dellatolas, 16 avenue Paul Vaillant-Couturier, 94807
Villejuif, France, Email: dellatolas@vjf.inserm.fr
This study was supported by grants from the Institut National de la Sante et de la Recherche
Medicale (INSERM), Ministere de la Recherche et de la Technologie (MRT), Caisse Nationale
d'Assurance Maladie (CNAM), and Fondation pour la Recherche Medicale. We thank E. Damia, P.
Myquel, and the staff of the electronics laboratory of Institut Gustave Roussy at Villejuif for con-
structing the computerised peg-moving apparatus, and S. Seaman for English corrections.
# 2003 Psychology Press Ltd
http://www.tandf.co.uk/journals/pp/1357650X.html DOI:10.1080/1357650
preference scale, there are generally no (or questionable) differences in cogni-
tive performance between these two groups (Bishop, 1990; De Agostini &
Dellatolas, 2001; Hardyck et al., 1976). When the continuum of hand skill
asymmetry is considered, the findings are complex and often contradictory.
There seems to be some agreement that phonology-related skills increase with
dextrality (Annett, 2002; De Agostini & Dellatolas, 2001; Natsopoulos et al.,
1998), but findings are conflicting concerning other relationships. Strong right-
hand advantage has been related to lower visual-spatial skills according to
Annett (2002), but to higher visual-spatial skills according to other authors
(Cerone & McKeever, 1999; De Agostini & Dellatolas, 2001). Subjects near the
middle of the sinistralitydextrality continuum might show either higher cog-
nitive skills, especially for vocabulary (Annett, 2002; Annett & Manning, 1989),
or, on the contrary, lower cognitive skills (Crow et al., 1998).
The above contradictions might be partly due to differences in methodology
and battery of tests used for cognitive assessment and measurement of hand
skill. In the verbal domain, Annett suggests that speech-related tasks show quite
a different hemispheric lateralisation from other language tasks (Annett, 2002).
In the visual-spatial domain it could be important to distinguish tasks that
involve directly the use of one (or both) hand(s) (e.g., reproduction of figures,
etc.) and those that do not, when the focus is the correlation of such visual-
spatial tasks with hand skill asymmetry. Concerning hand skill assessment,
previous reports showed that the distribution of hand skill asymmetry in the
population (and also the relationship between hand skill asymmetry and hand
preference) is quite different when one uses the peg-moving task of Annett
(1985) or paper-and-pencil tasks (Crow et al., 1998). These different distribu-
tions could explain why subjects near the middle of the sinistralitydextrality
continuum show either high or low cognitive scores (Curt, Maccario, & Del-
latolas, 1992; De Agostini & Dellatolas, 2001). Another confounding factor
could be the relationship of some cognitive tasks with overall manual skill.
There is a relative rarity of studies in normal preschool and school children
examining potential relationships between cognition and overall manual skill
(Bishop, 1980; Bishop & Edmundson 1987; Smirni & Zappala, 1989). In infants,
on the contrary, the relationships between manual and cognitive performance
have been extensively explored and discussed, and the links between perception
and action repeatedly demonstrated (Bushnell & Boudreau, 1993; Gibson, 1988;
Piaget, 1952; Ruff, 1989; Thelen, 1995). An infant's spontaneous manual
behaviour is considered as having an important ``knowledge gathering aspect''
(Bertenthal & Clifton, 1998). Widely used infant scales for assessing mental
development rely on motor activity such as grasping and manipulating objects
(Bayley, 1993). However, scales of infant development are poor predictors of
later intellectual development (Bornstein & Sigman, 1986; Colombo, 1993;
McCall, Hogarty, & Hurlburt, 1972), assessment of the Intellectual Quotient in
adults does not rely on manual skill tasks (e.g., Wechsler, 1981), and further-
318 DELLATOLAS ET AL.
more, the reliance of infant scales on motor activity has been proposed as a
possible explanation for their poor predictive value (Dougherty & Haith, 1997).
Overall, there seems to be some agreement that hand activity plays a major role
in early cognitive development, but a less important role latermainly, in our
literate societies, after acquisition of reading and writing skills. This leads to the
hypothesis that manual skills as assessed in early life might be more related to
the development of cognitive abilities than manual skills assessed later.
Moreover, it might be hypothesised that it is this overall manual skill that
relates differently to right-hemisphere-dependent and left-hemisphere-dependent
cognitive tasks, rather than the asymmetry in hand skill. If right-hemisphere
visual-spatial abilities are related to early bi-manual exploratory behaviour,
tasks assessing such abilities might correlate equally with left-hand skill and
right-hand skill. On the contrary, some left-hemisphere speech-related or pho-
nological memory-related tasks, possibly unrelated to early manual behaviour,
could also be unrelated to measures of right- and left-hand skill in preschool and
school children.
The present report examines relationships between hand skill and some
cognitive performances from a longitudinal study, planned for the investigation
of laterality effects in preschool children (Curt, De Agostini, Maccario, &
Dellatolas, 1995; Dellatolas, Curt, Dargent-Pare, & De Agostini, 1998a; Del-
latolas, Tubert-Bitter, Curt, & De Agostini, 1997; Dellatolas, Viguier, Deloche,
& De Agostini, 1998b). Hand skill was measured repeatedly by a computerised
version of the preg-moving task (CPMT) proposed by Annett (1985). Among the
many cognitive tasks used in this investigation, six tasks were selected on
theoretical criteria (i.e., to assess abilities involving ``speech'', phonological
memory, vocabulary, and visual-spatial skills with and without manual execu-
tion) and on design-related criteria (i.e., no missing data and repeated during the
longitudinal study). The analysis focused on the following: (i) the development
of hand skill and its asymmetry in young children, as assessed by the CPMT,
with examination of possible sex and handedness effects, in order to compare
the present findings with those reported previously, especially by Annett (2002);
(ii) the relationships between hand skill asymmetry and cognitive performance;
(iii) the relationship between overall hand skill and cognitive performance; (iv)
the respective role of early and later hand skill on cognitive performance.
METHOD
Subjects
In France most children attend preschools (``ecoles maternelles'') for 3 years,
from age 3 to 6 years, before entering the first grade of primary school. A total
of 1022 children, 516 boys and 506 girls, aged from 3 to 6 years (mean = 4.51,
SD = 0.83; 34 yrs: n = 309; 45 yrs: n = 384; 56 yrs: n = 329) attending
preschools and schools in the suburbs of Paris were included in the study (Curt
HAND SKILL AND COGNITION IN CHILDREN 319
et al., 1995; Dellatolas et al., 1997, 1998a, 1998b). The proportion of left-
handers was 11.3%, significantly higher among boys (14.3%) than among girls
(8.1%) (w
2
= 6.96; df = 1; p = .002). Of these 1022 children, 30% were followed
up for 2 years. The younger children were followed from age 3 to 5 years, and
the older ones from age 6 to 8 years. Left-handed children were oversampled,
that is, all left-handers were selected for follow-up. Children were examined five
times (T1T5), once every 6 months. The present follow-up sample includes all
children with no missing data for manual skill at T1, T3, and T5. In the present
study, only data from these three examinations will be reported. As Table 1
shows, there are 258 children (25.2% of the initial sample), 162 boys and 96
girls, 26.7% of them are left-handed, and there is a delay of 6 or 7 months
between each examination.
Procedure
Each child was examined individually in his or her (pre) school institution by a
psychologist. The duration of each examination was about 30 minutes for the
first examination and about 1 hour for the other four examinations. For the
younger children the session was split into shorter components when necessary.
All five examinations included assessment of handedness (De Agostini &
Dellatolas, 1988, 2001). Assessment of right- and left-hand skill with a
TABLE 1
Description of the sample
Initial sample
N 1022
Sex (% boys) 50.5
Handedness (% left) 11.3
Age: mean (SD) years 4.51 (0.83)
33.5 157
3.54 152
44.5 184
4.55 200
55.5 179
5.56 150
Follow-up sample
N 258
Sex (% boys) 62.8
Handedness (% left-handers) 26.7
Age at T1: mean (SD) years 4.3 (0.9)
Age at T3: mean (SD) years 5.4 (1.0)
Age at T5: mean (SD) years 6.5 (1.0)
320 DELLATOLAS ET AL.
computerised version of the peg-moving task (CPMT) (Curt et al., 1995), and
the digit span test, were given at T1, T3, and T5. In addition, the first exam-
ination (T1) included oral repetition and vocabulary assessment; the third (T3),
oral repetition and a visual memory task; and the fifth (T5), semantic verbal
fluency and a visual-spatial task involving manual execution.
Hand-preference assessment. On this eight-item task the child was asked
to: draw a line with a pencil, use an eraser, show how to use a comb, a
toothbrush, a hammer, a spoon, cut a sheet of paper with scissors and cut clay
with a knife (De Agostini & Dellatolas, 1988). Right-hand, both-hands, and left-
hand actions were coded 1, 2, and 3, respectively. A total handedness score (HS)
was calculated. HS thus ranges from 8 (extreme right-hander, i.e., 8 right-hand
answers) to 24 (extreme left-hander, i.e., 8 left-hand answers). The scores from
the three assessments were added. Thus, a total HS (range = 2472) was
generated. A child was classified as right-handed if his/her total HS was lower
than 48 (no child obtained a value of 48) and as left-handed if his/her total HS
was greater than 48.
The computerised peg-moving task (CPMT). The peg moving task was
initially proposed by Annett (1985) for the assessment of hand skill in children.
We used a computerised version of this task (Curt et al., 1995). This comprises a
wooden board consisting of two rows of 10 holes. Infrared emittorreceptors at
about 3 mm from the base of each hole are linked to a computer to record the
departure and arrival of a close-fitting wooden peg. The timer was started by the
emittorreceptor when the first peg was raised. The child had to move each of
the 10 pegs, one by one, to the hole in the opposite row, as fast as possible. The
peg-board was placed horizontally and centrally. After a demonstration by the
examiner, the child practised with each hand, moving 10 pegs. The task started
with a hand chosen at random and each hand performed two trials: one trial from
the nearest row to the furthest one, and one trial from the furthest row to the
nearest one. For displacements from the nearest to the furthest row, the child
began on the left side of the board with the left hand and on the right side with
the right hand. On the contrary, for displacements from the furthest row to the
nearest one, the child began on the contralateral side of the involved hand. If a
peg was dropped, the trial was restarted.
Vocabulary (T1). The child was asked to name 14 pictures of objects, to
name 14 pictures of actions, and then, in a multiple-choice task, to show the
picture of an object (16 items) or of an action (16 items) named by the examiner.
There were three controlled distractors, one visual, one semantic, and one
phonological. The score corresponds to the number of correct answers and
pointings (max = 60) (Kremin & Dellatolas, 1995).
HAND SKILL AND COGNITION IN CHILDREN 321
Digit span (T1, T3, T5). The subject had to reproduce an increasing number
of digits (presentation rate: 1/second). The examiner pronounced a series of two,
then three, etc. digits, and the subject had to repeat them. There were two items
for each number of digits. The task stopped after two consecutive errors. The
score was the number of digits produced in the correct sequence.
Oral repetition of words and nonwords (T1, T3). The child was asked to
repeat 16 common short words and 16 short nonwords (mono- or bisyllabic)
(Kremin & Dellatolas, 1995). The score was the number of items correctly
repeated (max = 32).
Visual recognition of abstract patterns (T3). A total of 24 abstract figures
were presented at the rate of one every 5 seconds. Every previously seen figure
was subsequently presented in a multiple choice paradigm with three previously
unseen abstract figures. The child had to recognise the target figure. The score
corresponded to the number of correct recognitions (Jambaque, Dellatolas,
Dulac, Ponsot, & Signoret, 1993; Signoret, 1991).
Embedded figures (T5). In this five-item task, scored from 0 to 5, the child
is asked to show the geometric figure embedded in a more complex geometric
figure, using a pencil (Gottschaldt, 1926, 1929).
Semantic verbal fluency (T5). The child was asked to produce as many
words as possible belonging to the category ``animals'' within 1 minute, and
then to the category ``clothing''. The score was the total number of correct
words produced.
Statistical analysis and data reduction
Statistical analysis was performed using the SAS software (1989). From the
hand-preference assessment, children were simply classified into two groups,
right- and left-handers, as described above. From the CPMT at T1, five times
and two hand skill asymmetry indexes were considered: right-hand time (RH)
i.e., sum of the time needed for the two trials of the right hand; left-hand time
(LH); dominant-hand time (DH), which is the right hand of right-handers and the
left hand of left-handers; nondominant-hand time (NDH), which is the left hand
of right-handers and the right hand of left-handers; total time (RH+LH); right-
hand advantage or ``Dextrality Index'' (DI), defined as (LH7RH)/(LH+RH);
and dominant-hand advantage or ``Laterality Index'' (LI) defined as
(NDH7DH)/(NDH+DH). From the CPMT at T3 and T5, only the total score
was considered here. Regression models were used (with the GLM procedure of
SAS) to reveal linear effects of age (continuous), sex, and handedness (two
levels) on CPMT and on cognitive scores. In some cases parabolic relationships
322 DELLATOLAS ET AL.
were also investigated (for the age effect on total CPMT time at T1; and for the
DI effect on cognitive scores). The age effect was very important, as expected,
both on CPMT times and on cognitive scores. To eliminate this age effect when
investigating relationships between CPMT times and cognitive scores, regres-
sion residuals were used (residuals of the parabolic relationship of age on CPMT
times; and of the linear relationship of age on cognitive scores).
RESULTS
Table 2 shows that age effect was highly significant for all CPMT times,
F(1, 1018) = 49.7, and the relationship was parabolic: age
2
was also significant,
F(1, 1017) = 21.7 (Figure 1). However, there was no significant age effect on
right-hand advantage (DI) and dominant-hand advantage (LI). When age and
handedness effects were partialled out, girls were slightly faster than boys with
the right hand and the dominant hand, F(1, 1017) = 4.3 and 4.6, but not with the
left or the nondominant hand. The between-hands discrepancies (DI and LI)
were bigger for girls than for boys, F(1, 1017) = 8.6 and 8.4. Total time was not
significantly different between boys and girls (Figures 2a and 2b). When age and
sex effects were partialled out, right-handers were faster with the right hand and
the dominant hand (i.e., when the right hand of the right-handers was compared
to the left hand of the left-handers), and left-handers were faster with the left
hand and, marginally, the nondominant hand. Both DI and LI were greater in
right-handers than in left-handers, F(1, 1017) = 245.1 for DI, and 40.5 for LI.
Total time was not significantly different between right- and left-handers
(Figures 2c and 2d). As Figure 1 suggests, not only mean CPMT time but also
the standard deviation (SD) of the CPMT time decreased with age. This decrease
of variability was not simply proportional to the mean. Indeed, when the
logarithm of CPMT rather than raw CPMT was analysed, SD still decreased
TABLE 2
T1: Age, sex, and handedness effects on CPMT times, right-hand advantage (DI), and
dominant-hand advantage (LI)
Age (Age
2
) Sex Handedness
Right hand (RH) time p < .0001 (p < .0001) p = .04 (F < M) p < .0001
Left hand (LH) time p < .0001 (p < .001) p = .76 p < .0001
Total (RH+LH) time p < .0001 (p < .0001) p = .23 p = .81
DI = (LH7RH)/Total p = .92 p = .003 (M < F) p < .0001
Dominant hand (DH) time p < .0001 (p < .0001) p = .03 (F < M) p = .007
Nondominant-hand (NDH) time p < .0001 (p < .001) p = .82 p = .06
LI = (NDH-DH)/Total p = .29 p = .004 (M < F) p = .0001
n = 1022
HAND SKILL AND COGNITION IN CHILDREN 323
Figure 1. Plot of total CPMT time (in seconds) at T1 versus age of the child (in years), with a
parabolic fit (n = 1022).
2a LI for boys and girls
Figure 2a2d. Histograms of laterality index (LI) and dextrality index (DI) for all children with
separate normal adjustments for boys (solid) and girls (dashed) or for right (solid) and left (dashed)
handed children.
n = 1022
324
2b DI for boys and girls
2c LI for left- and right-handed children
325
regularly with agefrom 18.7 in the younger to 10.1 in the older children;
Levene test of homogeneity of variance: F(5, 1016) = 10.7, p < .001.
A similar and highly significant effect of age on the three tasks of T1 was
observed, F(1, 1018) = 260, 271, and 384, for oral repetition, digit span, and
vocabulary respectively, as well as a slight advantage of girls compared to boys
on oral repetition and digit span, F(1, 1018) = 5.6, and 4.9. There was no
significant difference between right- and left-handers (Table 3).
Table 4 shows age-adjusted levels of significance (p-values) of the associa-
tions between CPMT scores and the three tasks of T1, and values of partial
correlation coefficients. Oral repetition was not related to any CPMT score (all
F-ratios < 0.5). Digit span was slightly related to CPMT times, F(1, 1020) values
between 25 and 31, but not to differences between hands (DI and LI). Voca-
bulary was related to CPMT times, F(1, 1020) values between 62 and 116, and
2d DI for left- and right-handed children
Figure 2 (continued).
TABLE 3
T1: Age, sex, and handedness on repetition, digit span, and vocabulary
Repetition Digit span Vocabulary
Age p < .0001 p < .0001 p < .0001
Sex p = .02 (M < F) p = .003 (M < F) p = .79
Handedness p = .87 p = .52 p = .98
n = 1022
326 DELLATOLAS ET AL.
showed a linear relationship with LI, i.e., dominant-hand advantage, F(1, 1020)
= 9.2, and a parabolic relationship with DI (i.e., right-hand advantage): voca-
bulary scores tended to be higher in children with reduced advantage of the
dominant hand than in children with important advantage of the dominant hand
(Figures 3a, 3b, 3c, 4a, 4b, 4c). When the same analyses were performed
separately for each age-group (three groups: 34 years, 45 years, 56 years),
and separately in boys and girls, the above relationships remained.
TABLE 4
T1: CPMT times on repetition, digit span, and vocabulary (age-adjusted associations)
CPMT (T1) Repetition Digit span Vocabulary
Right hand (RH) p = .91 p < .0001, r = 7.17 p < .0001, r = 7.25
Left hand (LH) p = .79 p < .0001, r = 7.15 p < .0001, r = 7.30
Total (RH+LH) p = .83 p < .0001, r = 7.17 p < .0001, r = 7.30
DI = (LH7RH)/Total (DI
2
) p = .73 p = .28 p = .06 (p = .006)
Dominant hand (DH) p = .49 p < .0001, r = 7.16 p < .0001, r = 7.24
Nondominant hand (NDH) p = .82 p < .0001, r = 7.16 p < .0001, r = 7.32
LI = (NDH7DH)/Total p = .35 p = .71 p = .003, r = 7.09
n = 1022
(a) Vocabulary
Figure 3a3c (above and overleaf). Graphs of the residuals on age of three cognitive tests at T1
(vocabulary, digit span, and oral repetition) versus the residuals on age of the total CPMT at T1 n =
1022). A high value (positive) of the residual of a cognitive test means that the child performed better
than the mean performance of children of the same age. A negative value of the residual of CPMT
means that the child performed faster than the mean performance of children of the same age. Data
are fitted with a linear model. A slope equal to zero means no association between the cognitive test
and CPMT time.
n = 1022
HAND SKILL AND COGNITION IN CHILDREN 327
Table 5 shows age-adjusted levels of significance (p-values) of the associa-
tions between CPMT scores at T1 and the three tasks of T3, and values of partial
correlation coefficients. CPMT times were related to visual recognition only.
Right-hand advantage (DI) was related to visual recognition and (slightly) to
oral repetition. Boys were slightly better than girls on semantic fluency (p < .02).
No other sex or handedness effect on the three cognitive tasks of T3 was
observed.
(b) Digit span
(c) Oral repetition
Figure 3 (continued).
n = 1022
n = 1022
328 DELLATOLAS ET AL.
Table 6 shows age-adjusted levels of significance (p-values) of the associa-
tions between CPMT scores at T1 and the three tasks of T5, and values of partial
correlation coefficients. Digit span was not related to any CPMT score.
Semantic fluency was slightly related CPMT times but not to discrepancies
between hands. Embedded figures was related to CPMT times and slightly to
(a) Vocabulary
(b) Digit span
Figure 4a4c (above and overleaf). Graphs of the residuals on age of three cognitive tests at T1
(vocabulary, digit span, and oral repetition) versus the dextrality index (DI) at T1 (n = 1022). DI is
not age-adjusted because it does not depend on age. Data are fitted with a linear or a parabolic model.
n = 1022
n = 1022
HAND SKILL AND COGNITION IN CHILDREN 329
right-hand advantage (DI). Right-handers were slightly better than left-handers
on embedded figures (p < .03). No other sex or handedness effect on the three
cognitive tasks of T5 was observed.
When correlations of T3 tasks with CPMT scores at T3, and of T5 tasks with
CPMT scores at T3 and T5, were examined, the pattern was similar to that
observed with CPMT scores at T1 (e.g., significant correlations for the two visual-
spatial tasks and semantic fluency; no significant relationship with digit span and
oral repetition). Partial age-adjusted correlations between the three total CPMT
scores were 0.58 between T1 and T3, 0.53 between T3 and T5, and 0.38 between
T1 and T5. Both the initial (T1) and the final (T3 or T5) total CPMT time were
tested as explanatory variables of the three tasks related to overall manual skill,
(c) Oral repetition
Figure 4 (continued).
TABLE 5
CPMT times (T1) on repetition, digit span, and vocabulary at T3 (age-adjusted
associations)
CPMT (T1) Repetition (T3) Digit span (T3) Visual recognition (T3)
Right hand (RH) p = .09 p = .17 p < .0001, r = 7.31
Left hand (LH) p = .83 p = .30 p = .03, r = 714
Total (RH+LH) p = .38 p = .22 p = .0002, r = 7.23
DI = (LH7RH)/Total p = .01, r = .16 p = .55 p = .002, r = .20
Dominant hand (DH) p = .31 p = .34 p = .0002, r = 7.23
Nondominant hand (NDH) p = .71 p = .14 p = .0007, r = 7.21
LI = (NDH7DH)/Total p = .24 p = .46 p = .84
n = 250
n = 1022
330 DELLATOLAS ET AL.
that is visual recognition (T3), embedded figures (T5), and semantic fluency (T5),
to examine whether the effect of the first remained significant when the effect of
the second was partialled out, and vice versa. Table 7 shows that for all three
tasks, only the initial CPMT time remained significant.
DISCUSSION
The findings of this study on development of hand skill, its asymmetry, and its
associations can be compared with those reported previously by Annett (2002),
with a manual version of the same peg-moving task. Both sources show that
there is a fall of peg-moving time of about 40% between 3 and 6 years of age
(Annett, 2002, p. 111). Within this age-range, the fall is more rapid in the
younger (34.5 years) than in the older (4.56) children. In addition, the present
study shows that not only does mean hand skill increase, but also the dispersion
of the scores decreases with age. This suggests that the more clumsy younger
children catch up with their peers 1 or 2 years later. Schooling possibly favours
decreasing of dispersion, and such age-dependent effects on dispersion are
probably related to the well-known difficulty in making ``predictions'' in very
TABLE 6
CPMT times (T1) on embedded figures, digit span, and semantic verbal fluency at T5
(age-adjusted associations)
CPMT Embedded figures Digit span Semantic fluency
Right hand (RH) p < .0001, r = 7.33 p = .62 p = .002, r = 7.19
Left hand (LH) p = .001, r = 7.20 p = .18 p = .0015, r = 7.20
Total (RH+LH) p < .0001, r = 7.28 p = .26 p = .0015, r 70.20
DI = (LH7RH)/Total p = .01, r = .15 p = .27 p = .95
Dominant hand (DH) p < .0001, r = 7.29 p = .52 p = .0007, r = 7.21
Nondominant hand (NDH) p < .0001, r = 7.25 p = .20 p = .003, r = 7.19
LI = (NDH7DH)/Total p = .20 p = .28 p = .22
n = 256
TABLE 7
Respective role of initial (T1) and simultaneous (T3 or T5) CPMT total time on the score
of three tasks
CPMT (T1) CPMT (T3) CPMT (T5)
Visual recognition (T3) t = 73.25, p = .0013 t = 7.32, p = .75
Embedded figures (T5) t = 74.01, p < .0001 t = 71.30, p = .19
Semantic fluency (T5) t = 72.90, p = .004 t = 7.60, p = .55
n = 254
HAND SKILL AND COGNITION IN CHILDREN 331
young children when pathology is absent (Bornstein & Sigman, 1986; Colombo,
1993; McCall et al., 1972).
Right minus left asymmetry does not change with age (Annett, 2002, p. 112).
This is confirmed by the absence of any significant age effect either on DI
(which is given as ``right minus left percent'' by Annett, and ``right-hand
advantage'' in the present study), or on LI (``dominant-hand advantage'').
Mean asymmetry is greater for females than for males (Annett, 2002, p. 112).
The present data in preschool children clearly confirm this statement. However,
there is an apparent discrepancy with Annett when the sex effect is considered
separately for the right hand and for the left hand. Annett observed that for the
right hand there was no significant sex difference, but for the left hand there was
a highly significant sex difference in favour of males. We observed exactly the
opposite: right hand slightly faster in females (p = .04 with n = 1022) but no sex
effect for the left hand. The discrepancy is likely to be due to major differences
in age-range and size of the samples. Annett examined an age-range 363 years.
She notes that, for the right hand, ``females tended to be slightly faster than
males up to about 8 years but the sexes were similar thereafter'' (p. 111). Thus,
there is no real discrepancy concerning the sex effect for the right hand. For the
left hand, it is possible that an advantage of males appears after 8 years of age.
The present data do not favour the presence of a ``threshold for sinistrality''
on the distribution of hand skill asymmetry at the peg-moving task (Annett,
2002, p. 74). However, they confirm that the mean asymmetry is much reduced
in left-handers compared to right-handers.
According to Annett (2002, pp. 189190), the left hemisphere has some
special role in facilitating speech learning, but the two hemispheres could be
equipotential for language, as Lenneberg (1967) suggested. Is strong dextrality
associated with better speech? In the present data, two observations could be
considered as being in favour of this hypothesis. First, the sex effect: girls, who
show stronger dextrality compared to boys, also show better speech (see also,
Bishop, 1990; De Agostini, Pare, Goudot, & Dellatolas, 1992; Nalcacy,
Kalaycioglu, Cicek, & Genc, 2001; Natsopoulos et al., 1998). Second, there is
one significant positive correlation between DI (right-hand advantage) at T1 and
oral repetition at T3. However, there was no sign of a relationship between DI at
T1 and oral repetition at T1.
The present data do not support the hypothesis of Annett that spatial ability
declines from left to right across the laterality spectrum (Annett, 2002, p. 235),
at least when this spectrum is measured by the DI of the peg-moving task. In
contrast, the relationships between DI and the two visual-spatial tasks
(embedded figures and visual recognition) were in the opposite direction to
those suggested by Annett: the stronger the right-hand advantage, the better the
performance (in accordance with Cerone & McKeever, 1999).
Strong dextrality and strong sinistrality could be associated with lower per-
formances at vocabulary testing (Annett, 2002, p. 206). The LI index, or
332 DELLATOLAS ET AL.
``dominant-hand advantage'', was constructed to test this hypothesis: a high LI
means either strong dextrality in right-handers or strong sinistrality in left-
handers, a LI equal to 0 means equal hand skill, and a negative LI means
discrepancy between hand preference and hand skill. In accordance with Annett,
vocabulary was the only task associated (negatively) with LI (Annett & Man-
ning, 1989). The parabolic relationship of vocabulary with DI tends to confirm
this observation. However, it should be noted that: (i) the correlation is weak and
its significance is related to the size of the sample (r = 7.09, R
2
= .008);
(ii) vocabulary is much more strongly related to CPMT times of the right hand
(r = .25, R
2
= .06) and of the left hand (r = .30, R
2
= .09). The latter observation
could accentuate the ``inverted U-shaped'' distribution of vocabulary in the right
minus left (R7L) continuum, if R7L is not divided by total time. In this case it
is likely that extreme R7L values would be observed in subjects with the longer
R and L times. We propose here that a close look at the associations of the
cognitive tasks with R (Right hand time), L (Left hand time), and R+L, is a
prerequisite for the study of the associations of the cognitive task with R7L,
and even with (R7L)/(R+L).
Oral repetition was not related at all to manual skill. Oral repetition included
both words and nonwords, however the total repetition performance was more
dependent on repetition of nonwords, given the presence of ceiling effects for
repetition of words.
A weak association was observed between manual skill and digit span in the
initial sample, which was significant given the large number of subjects. This
weak association was not confirmed in the longitudinal part of the study for this
particular task. Early digit span might not depend on exactly the same cognitive
components as later digit span. The correlations between digit span at T1 and the
same task at T3 and T5 were relatively low (r = .32 and r = .37 respectively; but
r = .55 between T3 and T5; adjusted for age). It is possible that familiarisation
with the visual form or the meaning of the numbers plays a more important role
in early digit span than in later digit span. In other words, digit span might be
clearly a ``phonological loop'' dependent task (Gathercole & Baddeley, 1990) in
the older children, but not so in the younger, in whom memory ``strategies''
such as subvocal rehearsal are not well developed (Bjorklund & Douglas, 1997;
Cowan, 1997).
A stronger association was observed between manual skill and vocabulary in
the initial sample. Although stronger and clearly significant, manual skill
explained only 9.6% of the variance in vocabulary performance. This relatively
low link is however expected in normally developing children, whose vocabu-
lary development certainly depends on many other factors in addition to manual
skill. The link between total CPMT and vocabulary had a roughly linear aspect,
suggesting that it was not exclusively due to the particularly clumsy or the most
skilled children. This remains true for the other cognitive tasks of the study
significantly related to CPMT. It is worth noting that vocabulary was assessed
HAND SKILL AND COGNITION IN CHILDREN 333
using figures (naming and multiple-choice pointing tasks) which might be dif-
ferent from spontaneous oral language vocabulary.
However, a vocabulary-dependent task without direct use of figures, that is
the semantic fluency task, was also related to hand skill, and notably with hand
skill assessed 2 years earlier, as were visual-spatial tasks, with (embedded fig-
ures) and without (visual recognition memory) important manual involvement.
As already mentioned, the strong evidence for a ``knowledge-gathering
aspect'' of the hands in infants (Bertenthal & Clifton, 1998) contrasts with the
absence of manual skill tasks in IQ batteries for adults (Wechsler, 1981). This
suggests that the younger the child, the stronger the relationship between hands
and cognition. The present investigation concerns a relatively narrow range of
ages and did not succeed in showing clear evidence, in the initial sample, for a
differential effect of age on the handscognition relationship. Vocabulary at T1,
for instance, remained associated to CPMT in all three age-groups. However, the
longitudinal part of the study does provide evidence that the first CPMT
examination only had a specific effect on later cognitive tests when the effect of
later CPMT examination was partialled out.
Manual skill probably has a stronger cognitive component in young children
than in adults. The peg-moving task is attractive for the young child, as men-
tioned by Annett (1985). There were no refusals and some young children even
asked to be allowed to take our apparatus home. As the results show, CPMT was
a very demanding task for the younger children and probably a more auto-
matised and less attractive task for the older children. The degree of attrac-
tiveness of the task might be considered as indirect evidence of its ``cognitive''
character, and this could also explain why the first CMPT had a stronger rela-
tionship with cognitive scores than later CPMT.
There is now good evidence that hemispheric specialisation is different in
young children than in adults (see Bates, 1999, for a review), and that cortical
differentiation is largely the product of input to the cortex, albeit with certain
constraints (Johnson, 1997). Looking for differential cognitive correlations of
manual laterality in young children relies excessively on the topographic aspects
of adults hemispheric specialisation and neglects its developmental aspects. It
seems reasonable to assume, for instance, that activity of both hands (not only
activity of the left hand) in children contributes to the specialisation of the right
hemisphere for visual-spatial skills. Also, the present results from normally
developing children seem congruent with brain-imagery studies suggesting that
the right brain hemisphere is dominant in human infants (Chiron et al., 1997), as
well as with observations from children with early brain damage suggesting a
major role of the right hemisphere in verbal comprehension (Bates, 1999).
We hypothesised that in young children, common cognitive components are
involved in early CPMT performance, in the development of visuo-spatial skills,
but also in the understanding of the external word. These manual-activity-
dependent components are initially different from those involved in phonolo-
334 DELLATOLAS ET AL.
gical skills and in the development of phonological memory. It is likely that
schooling stimulates more phonological skills (e.g., learning to read) and manual
specialisation for a limited number of tasks, such as using a pencil for drawing
and writing, than general-purpose manual skill. As a result, the initial relative
independence of phonology from the other cognitive components tends to dis-
appear with the child's development.
Thus, one important account of the main findings of this study could be
summarised as follows: due to genetic and environmental factors, normal infants
and young children show important variability of their manual exploratory
behaviour (MEB). Performance on demanding manual skill tasks, such as those
using Annett's (1985) peg board, is a behavioural marker of previous MEB.
Early MEB plays an important role in the development of visuo-spatial skills
and the knowledge of the environment. Early MEB does not play an important
role in the development of phonological skills, short-term verbal memory, or
reading acquisition, which become progressively major determinants of cogni-
tive development and academic performance.
An alternative hypothesis is that some other underlying factor, which is not
manual-activity-dependent, such as motivation, memory, or speed of informa-
tion processing (Rose & Feldman, 1997), could influence both CPMT and
cognitive tests and explain the association. Such alternative explanations have to
be explored further, to examine specifically whether they could account for the
differential correlations observed.
According to the MEB hypothesis, an early occurring bimanual deficiency
can be a negative factor for specific aspects of cognitive development, such as
visual-spatial skills and comprehension, but not for others, such as phonological
memory or expressive language. This should be observed independently of the
origin of the manual deficiency (e.g., absence or presence and topography of a
brain lesion). Investigations of the cognitive profiles of children with early
bimanual deficiencies of peripheral or central (brain-damage) origin could be
critical for this view.
Finally, it must be noted that the above account of our findings in normally
developing children seems in conflict with recent studies in children with spe-
cific language impairment (SLI), reporting shared genetic variance for impair-
ments on peg-moving tasks and on a test of nonword repetition (Bishop, 2002).
This study found exactly the opposite relationships: in children with SLI, it is
when speech production is affected that the association with motor immaturity is
most evident. In addition, the latter association looks quite expected as both
speech production and peg moving share motor behaviour. Pending further
studies, which will elucidate the complex relationship between hand skill and
cognitive performance, a provisional conclusion might be that: (i) cognitive
associations of motor development could be different in normally developing
children and in populations of children affected by neurological pathologies.
Children with cerebral palsy, for instance, might show speech-production
HAND SKILL AND COGNITION IN CHILDREN 335
difficulties related to their motor deficiencies. (ii) Hemispheric specialisation
could lead to conflicting interpretations of such associations, as hand skill
development might be associated more strongly to right-hemisphere-dependent
visual-spatial skills, or on the contrary more strongly to left-hemisphere-
dependent motor-speech-production skills; and (iii) if speech production and
hand skill are relatively independent in normally developing children, a child
with impairment in both abilities has a high probability of being subject to a
developmental or acquired disorder.
Manuscript received 18 March 2003
REFERENCES
Annett, M. (1985). Left, right, hand and brain: The right shift theory. Hove, UK: Lawrence Erlbaum
Associates Ltd.
Annett, M. (2002). Handedness and brain asymmetry: The right shift theory. Hove, UK: Psychology
Press.
Annett, M., & Manning, M. (1989). The disadvantages of dextrality for intelligence. British Journal
of Psychology, 80, 213226.
Bates, E. (1999). Plasticity, localization and language development. In S. H. Broman & J. M.
Fletcher (Eds.), The changing nervous system: Neurobehavioral consequences of early brain
disorders (pp. 214253). New York: Oxford University Press.
Bayley, N. (1993). Bayley Scales of Infant Development (2nd ed.). San Antonio, CA: Psychological
Corp.
Bertenthal, B. I., & Clifton, R. K. (1998). Perception and action. In W. Damon, D. Kuhn, & R. S.
Siegler (Eds.), Handbook of child psychology: Vol. 2 (pp. 51102). New York: Wiley.
Bishop, D. V. (2002). Motor immaturity and specific speech and language impairment: Evidence for
a common genetic basis. American Journal of Medical Genetics, 114, 5663.
Bishop, D. V. M. (1980). Handedness, clumsiness and cognitive ability. Developmental Medicine
and Child Neurology, 22, 569579.
Bishop, D. V. M. (1990). Handedness and developmental disorder. Hove, UK: Lawrence Erlbaum
Associates Ltd.
Bishop, D. V. M., & Edmundson, A. (1987). Specific language impairment as a maturational lag:
Evidence from longitudinal data on language and motor development. Developmental Medicine
and Child Neurology, 29, 442459.
Bjorklund, D. F., & Douglas, R. N. (1997). The development of memory strategies. In N. Cowan &
C. Hulme (Eds.), The development of memory in childhood (pp. 201246). Hove, UK: Psy-
chology Press.
Bornstein, M. H., & Sigman, M. S. (1986). Continuity in mental development from infancy. Child
Development, 57, 251274.
Bushnell, E., & Boudreau, J. P. (1993). Motor development and the mind: The potential role of motor
abilities as a determinant of aspects of perceptual development. Child Development, 64,
10051021.
Cerone, L. J., & McKeever, W. F. (1999). Failure to support the right-shift theory's hypothesis of a
`heterozygote advantage' for cognitive abilities. British Journal of Psychology, 90, 109123.
Chiron, C., Jambaque, I., Nabbout, R., Lounes, R., Syrota, A., & Dulac, O. (1997). The right brain
hemisphere is dominant in human infants. Brain, 120, 10571065.
Colombo, J. (1993). Infant cognition: Predicting later intellectual functioning. Thousand Oaks, CA:
Sage.
336 DELLATOLAS ET AL.
Cowan, N. (1997). The development of working memory. In N. Cowan & C. Hulme (Eds.), The
development of memory in childhood (pp. 163199). Hove, UK: Psychology Press.
Crow, T. J., Crow, L. R., Done, D. J., & Leask, S. (1998). Relative hand skill predicts academic
ability: Global deficits at the point of hemispheric indecision. Neuropsychologia, 36, 12751282.
Curt, F., De Agostini, M., Maccario, J., & Dellatolas, G. (1995). Parental hand preference and
manual functional asymmetry in preschool children. Behavior Genetics, 25, 525536.
Curt, F., Maccario, J., & Dellatolas, G. (1992). Distributions of hand preference and hand skill
asymmetry in preschool children: Theoretical implications. Neuropsychologia, 30, 2734.
De Agostini, M., & Dellatolas, G. (1988). Une epreuve simple pour evaluer la preference manuelle
chez l'enfant a partir de 3 ans. Enfance, 41, 139147.
De Agostini, M., & Dellatolas, G. (2001). Lateralities in normal children aged 3 to 8 and their role on
cognitive performance. Developmental Neuropsychology, 20, 429448.
De Agostini, M., Pare, C., Goudot, D., & Dellatolas, G. (1992). Manual preference and skill
development in preschool children. Developmental Neuropsychology, 8, 4157.
Dellatolas, G., Curt, F., Dargent-Pare, C., & De Agostini, M. (1998a). Eye-dominance in children: A
longitudinal study. Behavior Genetics, 28, 187195.
Dellatolas, G., Tubert-Bitter, P., Curt, F., & De Agostini, M. (1997). Evolution of degree and
direction of hand preference in children: Methodological and theoretical issues. Neuropsycho-
logical Rehabilitation, 7, 387399.
Dellatolas, G., Viguier, D., Deloche, G., & De Agostini, M. (1998b). Rightleft orientation and
significance of systematic reversal in children. Cortex, 34, 659676.
Dougherty, T. M., & Haith, M. M. (1997). Infant expectations and reaction time as predictors of
childhood speed of processing and IQ. Developmental Psychology, 33, 146155.
Gathercole, S., & Baddeley, A. (1990). Phonological memory deficits in language disordered chil-
dren: Is there a causal connection? Journal of Memory and Language, 29, 336360.
Gibson, E. J. (1988). Exploratory behavior in the development of perceiving, acting, and the
acquiring of knowledge. Annual Review of Psychology, 39, 141.
Gottschaldt, K. (1926). U

ber den Einfluss der Erfahrung auf die Wahrnehmung von Figuren. Psy-
chologische Forschung, 8, 261317.
Gottschaldt, K. (1929). U

ber den Einfluss der Erfahrung auf die Wahrnehmung von Figuren. Psy-
chologische Forschung, 12, 187.
Hardyck, C., Petrinovich, L. F., & Goldman, R. D. (1976). Left handedness and cognitive deficit.
Cortex, 12, 266279.
Jambaque, I., Dellatolas, G., Dulac, O., Ponsot, G., & Signoret, J. L. (1993). Verbal and visual
memory impairment in children with epilepsy. Neuropsychologia, 31, 13211337.
Johnson, M. H. (1997). Developmental cognitive neuroscience. Oxford: Blackwell Publishers.
Kremin, H., & Dellatolas, G. (1995). L'acces au lexique: une etude de standardisation chez l'enfant
d'age pre-scolaire. Revue de Neuropsychologie, 53, 309338.
Lenneberg, E. H. (1967). Biological foundations of language. New York: Wiley.
McCall, R. B., Hogarty, P. S., & Hurlburt, N. (1972). Transitions in infant sensorimotor development
and the prediction of childhood IQ. American Psychologist, 27, 728748.
Nalcacy, E., Kalaycioglu, C., Cicek, M., & Genc, Y. (2001). The relationship between handedness
and fine motor performance. Cortex, 37, 493500.
Natsopoulos, D., Kiosseoglou, G., Xeromeritou, A., & Alevriadou, A. (1998). Do the hands talk on
mind's behalf? Differences in language ability between left- and right-handed children. Brain
and Language, 64, 182214.
Piaget, J. (1952). The origins of intelligence in children. New York: International University Press.
Rose, S. A., & Feldman, J. F. (1997). Memory and speed: Their role in the relation of infant
information processing to later IQ. Child Development, 68, 630641.
Ruff, H. (1989). The infant's use of visual and haptic information in the perception and recognition
of objects. Canadian Journal of Psychology, 43, 302319.
HAND SKILL AND COGNITION IN CHILDREN 337
SAS Institute INC. (1989). SAS/STAT user's guide, version 6, fourth edition. Cary, NC: SAS.
Signoret, J. L. (1991). Batterie d'Efficience Mnesique B.E.M. 144. Paris: Elsevier.
Smirni, P., & Zappala, G. (1989). Manual behavior, lateralization of manual skills and cognitive
performance of preschool children. Perceptual and motor skills, 68, 262272.
Thelen, E. (1995). Motor development: A new synthesis. American Psychologist, 50, 7995.
Wechsler, D. (1981). WAIS-R manual. New York: The Psychological Corp.
338 DELLATOLAS ET AL.

You might also like