Content uploaded by Michel Audiffren
Author content
All content in this area was uploaded by Michel Audiffren on Jan 09, 2018
Content may be subject to copyright.
Age-Related Differences in the Preparatory Processes
of Motor Programming
Isabelle Olivier and Michel Audiffren
Faculte´ des Sciences du Sport, Universite´ de Poitiers, France
and
Hubert Ripoll
Faculte´ des Sciences du Sport, Universite´delaMe´ditte´rane´e, France
This article investigates the mechanisms underlying the age-related differences in
information processing in the production of motor responses, especially the development
of feedforward mechanisms. No age-related differences have emerged from developmen-
tal studies aiming at analyzing motor programming. Nevertheless, age effects have seldom
been studied in function of motor preparation. The aim of the present experiment was (1)
to study age differences in motor preparation, and (2) to validate the early maturation of
movement parameters specification. Two conditions were used (1) no advanced informa-
tion on the movement to be made was given to the subject (neutral prime), and (2)
advanced information on which direction should be followed was provided to the subject,
allowing him to prepare a response based on a prime (primed condition). Four age groups
were studied 6, 8, 10 and 22 years. Our results showed mainly an early maturation of the
programming processes. More specifically, (1) beyond 6 years of age, children are capable
of using the information provided by the prime to prepare their movement in advance, (2)
costs and benefits of pre-programming do not vary significantly with age, (3) deprogram-
ming-reprogramming of effector and direction is quite similar across the four age groups.
© 1998 Academic Press
Key Words: Ontogenesis; movement preparation; response programming.
The information processing approach is commonly used to study mental
operations involved in reaction time tasks. In such a theoretical framework, a
multi-linear-stage model including stimulus identification, response selection,
We are grateful to the children, parents, teachers, and students for their participation in the study.
We thank Yannick Blandin (University of Poitiers), Michael Broderick (Pennsylvania State Univer-
sity), Michelle Fleury (Laval University), and Jean-Jacques Temprado (University of Marseille) for
helpful comments on earlier versions of this paper.
Address correspondence and reprint requests to Isabelle Olivier, Faculte´ des Sciences du Sport, 4
alle´e Jean Monnet, 86 000 Poitiers, France. E-mail: Isabelle.Olivier@mshs.univ-poitiers.fr.
JOURNAL OF EXPERIMENTAL CHILD PSYCHOLOGY 69, 49–65 (1998)
ARTICLE NO. CH982433
0022-0965/98 $25.00
Copyright © 1998 by Academic Press
All rights of reproduction in any form reserved.
49
and response programming is often presented (e.g., Sanders, 1990; Schmidt,
1988; Theios, 1975). Ontogenetic research reported in this paper is mainly
concerned with the ontogenesis of the response-programming stage.
Originally, research using the chronometric approach showed that information
processing speed increases with age (see Wickens, 1974 for a review). For
instance, Sugden (1980) suggested that the speed of information ranges from an
average of 3.15 bits/s for age 6 years, to 4.29 bits/s for age 8 years, to 5.53 bits/s
for age 12 years. This approach provided evidence for the existence of age
differences located at the identification and selection stages of processing (e.g.,
Welsandt, Zupnick, & Meyer, 1973). For the motor side of information process-
ing, it was shown that certain operations, insuring motor control, occur prior to
response initiation (feedforward processing), and during execution (feedback
processing). Research on motor development predominantly studied on-line
processing which allows the use of feedback during movement execution (Bard
& Hay, 1983; Bard, Fleury, & Gagnon, 1990; Hauert & Pellizzer, 1992; Hauert,
Zanone, & Mounoud, 1990; Hay, 1987), and yielded age-related differences in
the use of feedback processing. These studies excluded the analysis of feedfor-
ward processing. However, according to Thomas (1980), feedforward processing
may provide the best locus of explanation for age-related differences in motor-
skill execution. Feedforward processes conception led to the emergence of many
motor programming models (Schmidt, 1975; Rosenbaum, 1983). The present
study was designed to investigate the maturation of programming, deprogram-
ming and reprogramming processing according to Rosenbaum’s parametric
model of motor programming.
Durand and Barna (1987), had 7- and 11-year olds and adults striking a ball at
a target with a hockey stick, to study the ontogenesis of various information
processes. In addition to age, they manipulated the uncertainty of the ball
trajectory (rolling versus bouncing), the number of alternatives (one vs. two) and
the target size (large vs. small). The previous three factors are liable to affect the
identification, the selection and the programming stage, respectively. The authors
observed two specific and independent interactions: age by uncertainty of ball
trajectory and age by number of alternatives. These results confirmed that
age-related differences in information processing may be attributed to the level
of maturity of the identification and selection stages. They also showed that the
target size did not interact with age, suggesting that the response-programming
stage matures earlier. This study, however, was somewhat limited in scope.
Indeed, it is debatable whether the three task factors were really representative of
the respective operations of identification, selection, and programming. More-
over, this research studied the number of errors in aiming accuracy, and it is well
known that the terminal accuracy of an aiming movement depends on the
programming of impulse duration and amplitude (Schmidt et al., 1979), and/or on
the number of feedback-based corrections (Crossman & Goodeve, 1963/1983).
Therefore, it is impossible to determine precisely the role of feedforward pro-
50 OLIVIER, AUDIFFREN, AND RIPOLL
cessing in motor performance based on performance accuracy. An RT measure
would be more appropriate for inferring feedforward processing.
The first chronometric research with children showed no age difference in the
duration of the feedforward mechanisms involved in the production of a motor
response (Henry & Rogers, 1960). They measured 6- and 10-year-olds’ and adults’
RT in three tasks of different movement complexity and found no Age 3Movement
Complexity interaction. However, they did not measured response errors, which are
obviously part of a clear description of response programming. On the other hand,
Clark (1982), Olivier, Ripoll & Audiffren, (1994), and Reilly & Spirduso, (1991)
included errors in their chronometric studies and provided evidence for the early
maturation of the response-programming stage. In Clark (1982) and Reilly &
Spirduso (1991), 7- and 11-year-olds and adults performed a two-choice RT task
under two levels of response-response compatibility (R-R compatibility). This factor
is liable to affect response programming (Kornblum, 1965; Schulman & McConkie,
1973). They showed that R-R compatibility had no significant interaction with age on
RT or response errors, suggesting an early maturation, for response-programming
operations. More recently, Olivier, Ripoll & Audiffren (1994) showed that RT and
errors were similarly affected by motor complexity at age 7, 9, 11, and 22 in
intercepting a ball with a paddle, thus confirming the early maturation of the
programming stage. However, in this earlier study, subjects were not given a
standardized limit of acceptable errors. In the present research, children and adults
were trained till they reached not more than 5% errors in order to further study of the
maturation of response-programming.
Studies concerning the specification of movement parameters frequently use
precuing or priming paradigms. Such paradigms consist of providing information
on the response to be executed, prior to the response signal, thus allowing
subjects to prepare their response in advance. In the precuing paradigm (Rosen-
baum, 1980), information is given about none, some, or all of the parameters of
a forthcoming movement. It is assumed that RT reflects the time necessary to
specify the values that were not precued. From RTs obtained for the same
movements in different precuing conditions, one can make detailed inferences
about the parameter specification process involved in each of the movements.
Rosenbaum’s study concerned the specification of the hand (right or left), the
direction (upward or downward), and the extent (short or long) of hand aiming
movements. He found the longest specification times for the hand, followed by
the direction, the shortest being for the extent. Moreover, these parameters were
specified serially, although not in an invariant order. The most important limit of
the precuing technique concerns the nature of subject’s response preparation
following precue presentation. Subjects are liable to prepare several of the
potential responses allowed by a precue and to select an appropriate response
from this set following the occurrence of the response signal. Differences in RTs
associated with different precues could then be attributed to differences in the
discriminability among a set of tuned responses. Rosenbaum and Kornblum
51
MOTOR PROGRAMMING
(1982) proposed a priming paradigm [derived from LaBerge, Van Gelder, &
Yellott, (1970), and Posner & Snyder, (1975), studies on attentional processing]
to overcome the precuing paradigm limitations and constrained subjects to the
specific tuning of a response. This procedure consists of creating a bias for one
response. The priming technique is, therefore, specifically designed to induce
subjects to adopt fixed preparation strategies. The prime supplies the subject with
complete target information about movement parameters. However, within a
series of trials, the prime is followed either with a high probality by an expected
target (a valid prime) or with a low probality by an unexpected target (an invalid
prime). The high probability of prime validity generally prompts subjects to
prepare the primed response. Pointing movements toward the expected and
unexpected targets can, therefore, differ in one, several, or all parameters,
according to the spatial locations of both kinds of targets. The RT lengthening for
pointing movements to an unexpected target is interpreted as the sum of the times
necessary to deprogram and reprogram movement parameters for which the
prime was invalid (Le´pine, Glencross, & Requin, 1989). A priming technique,
therefore, seems to be appropriate to study the maturation of preparatory and
programming processes.
In the present experiment, we used two conditions, one in which no advance
information about the required movement was given to the subject (neutral
condition), and one in which advance information was given to the subject in
order to prepare response (primed condition). In the neutral condition, RT
measured the delay to detect, identify, and program the required response. In the
primed condition, preparatory processes should occur prior to the response
signal-during the preparatory period-and consisted of programming a response in
advance (movement pre-programming). Reactive processes (Bertelson, 1969)
occur after the signal, during the RT. When the prime is valid, the RT is a
measure of the delay in detecting, identifying and initiating the primed response.
When the prime is invalid, RT is a measure of the delay in detecting, identifying,
and modifying one or more parameters of the primed response in the aim to
respond appropriately (deprogramming and reprogramming of movement param-
eters). Some studies showed that RT is progressively longer to (1) initiate a
pre-programmed movement in the case of a valid prime, (2) program the effector
and the direction of the movement in the case of a neutral prime, and (3) program,
deprogram, and then reprogram, in all or in part, the pre-programmed movement
in the case of an invalid prime (Larish & Frekany, 1985; Le´pine, Glencross, &
Requin, 1989). It would be interesting to use the Posner and Snyder’s (1975) cost
benefit analysis, initially applied to orienting of attention processes, to study
motor preparation. In fact, the term attention is mainly used when attentional
ressources are oriented on perceptual processes (e.g., covert and overt visual
orienting of attention). When ressources are oriented on motor processes, the
term motor preparation is prefered (Requin, 1985). According to Posner and
Snyder, the difference between RT measured in an invalid condition and RT
52 OLIVIER, AUDIFFREN, AND RIPOLL
measured in a neutral condition corresponds to the cost of orienting of attention,
whereas the difference between RT measured in a neutral condition and RT
measured in a valid condition corresponds to the benefit of orienting of attention.
In our experiment, cost would represent duration of deprogramming-reprogram-
ming process and benefit, duration of pre-programming process.
Two main questions were addressed in the present experiment. First, are there
age differences in motor preparation? This question was answered by comparing
the magnitude of motor preparation benefits and deprogramming-reprogramming
costs from 6 to 22 years of age. Second, does programming of both effector (right
or left-hand) and direction (up and down) parameters mature early? This question
was answered by comparing the evolution of the costs of deprogramming side,
deprogramming-reprogramming direction and deprogramming-reprogramming
both parameters from 6 to 22 years of age.
METHOD
Subjects
Forty-three subjects divided into four groups according to their age partici-
pated in the experiment. The first group was composed of 10 subjects (5 girls, 5
boys) averaging 6 years, 5 months (range: 6,2–6,10). The second group was
composed of 12 subjects (8 girls, 4 boys) averaging 8 years, 5 months (range: 8–
8,10). The third group was composed of 9 subjects (4 girls, 5 boys) averaging 10
years, 5 months (range: 10,3–10,10), and the fourth group was composed of 12
subjects (5 females, 5 males) averaging 22 years, 5 months (range: 19,3– 28,9).
All subjects were right-handed and naive as the purpose of the experiment.
Apparatus
Subjects were seated at a table facing a cathode-ray-tube (CRT) screen placed
1 m away (see Figure 1). Two handles (right and left), that could be moved up
and down, were mounted on the table. Subjects placed their forearms pointing
forward, holding a handle in each hand. Wrist extensions and flexions moved the
handles up and down, respectively. The pivot of each handle was equipped with
a potentiometer that sensed the position of the handles at all times. This position
was represented by two Xs displayed on the screen to indicate the position of the
X when the handles were in the starting position. Stimuli were displayed on a
green LED device fixed in the center of the CRT screen. The response signal was
an arrow (1 cm long) located at each side of a vertical line which pointed either
up or down. The primes were short horizontal lines located at each side of the
vertical line at the tip of each arrow. The lighting of the four short horizontal lines
corresponded to the neutral prime, whereas the lighting of one short horizontal
line only corresponded to a valid or an invalid prime.
The task consisted of moving the left or right handle, thus controlling the Xs
displacements on the screen, in an upward or downard direction depending on the
direction of the arrow. Thus, subjects performed a choice reaction task with four
53MOTOR PROGRAMMING
alternatives (two hands and two directions). The CRT screen also provided
subjects with knowledge of results on their RT after each trial.
Procedure
The experimenter explained the task to the subjects. They were required to
move the handles toward the signal as quickly as possible without making errors.
An auditory signal marked the start of the trial, when the subject got into the
starting position. A neutral prime or a valid or invalid prime signal occurred and
TABLE 1
Each Cell Represents the Probability of Occurrence of the Corresponding Response Signal
Response signal
Priming signal
Right up Right down Left up Left down Neutral
Right up 0.50 0.17 0.17 0.17 0.25
Right down 0.17 0.50 0.17 0.17 0.25
Left up 0.17 0.17 0.50 0.17 0.25
Left down 0.17 0.17 0.17 0.50 0.25
FIG. 1. Experimental setup. A) Subject executing a right up response. B) Side view of hand
position.
54 OLIVIER, AUDIFFREN, AND RIPOLL
was followed by a 2 sec foreperiod. At the end of the preparatory period, one of
the four arrows went on. The probability of any particular arrow going on was
0.25 in the neutral prime condition, and the probability of an arrow to correspond
to the priming signal was 0.5 (see Table 1, for more information).
At the end of each trial, during the inter-trial interval, subjects received
feedback on their RT. Moreover, at the end of each block of 36 trials, subjects
were given feedback on their average RT and on the number of errors for that
block. RT was the time elapsed from the onset of the arrow to the X departure
from the starting position. Any RT less than 150 ms was considered to be an
anticipation, and the word ‘‘Anticipation’’ appeared on the CRT screen. When
RT exceeded 600 ms, the message ‘‘Too Long’’ appeared on the CRT. In case
of an error of response programming (side or direction), the type of error was
indicated on the CRT screen; i.e., ‘‘Direction Error’’, ‘‘Side Error’’ or ‘‘Double
Error’’, respectively. Finally if the non-responding hand failed to stay in the
starting position during the response of the other hand, the message ‘‘Moved’’
appeared on the CRT. Whenever any of these errors occurred, the trial was
rejected, and repeated at random within the remaining trials. Each block thus
consisted of 36 error-free trials.
Subjects began the experiment by a training session aimed at achieving a
criterion of accuracy (the subject was required to execute the task with less than
5% errors) and a criterion of stability (less than 15% standard deviation of the RT
mean). Those criteria have been established by Sanders (1980, 1990) in studies
on adults. Subjects were submitted to four blocks of 36 trials, each composed of
12 trials for each of the three prime conditions: 12 valid trials, 12 neutral trials
and 12 invalid trials, for a total of 144 trials. Subjects were tested individually
during two sessions of approximately 30 minutes, including two blocks of 36
trials. The duration of the training session depended on the number of periods,
that is the number of trial blocks, necessary to achieve the criteria.
Statistical Analyses
We selected a multivariate (MANOVA) to study the effect and interactions of
factors with repeated measures (Mc Call & Appelbaum, 1973; Keselman, &
Keselman, 1990; O’Brien & Kister Kaiser, 1985; Rogan, Keselman, & Mendoza,
1979). Box (1974) showed that ANOVAs, traditionally used to analyse repeated
measures designs, often violate the multisample sphericity assumption. There are
two parametric approaches that avoid sphericity assumptions for repeated mea-
sures analysis: modifying the traditional mixed-model method and using a more
conservative univariate F test, like Greenhouse-Geiser (1958, 1959) or Huynh-
Feldt (1976) procedures, or using a MANOVA approach, free of sphericity
assumptions. A recent paper (Keselman & Keselman, 1990) showed that for
balanced designs containing a large number of repeated measurements, the
MANOVA procedure is generally preferable to the Huynh-Feldt procedure. In
addition, Olson (1974) compared the type I error and power rates of six multi-
55MOTOR PROGRAMMING
variate tests including Roy’s (1953) largest root criterion, Hotelling’s (1951)
trace criterion, Wilk’s (1932) likelihood ratio and Pillai-Bartlett’s (Bartlett, 1939;
Pillai, 1955) trace statistic. Based on his simulation results, Olson (1974) recom-
manded the Pillai-Bartlett trace criterion V to assess the within-subjects interac-
tion hypothesis. As suggested by Olson (1974), we used the Pillai-Bartlett
procedure, provided by Statistical package, to study the effect of Prime Type,
Prime Invalidity, Errors Type and their interactions with others factors.
RESULTS
Training Session
An ANOVA was conducted, with Age as a between-subject factor, on the
number of block of trials necessary to achieve the two criteria. The ANOVA
showed that the number of blocks necessary to achieve both criteria decreased
with age, F (3, 41) 510.01, p ,.001 (see Table 2). The post-hoc analysis using
the Newman-Keuls test, showed no significant difference between 6- and 8-year-
old subjects nor between 10- and 22-year-old subjects. All other differences were
significant. The number of blocks necessary to achieve only the accuracy crite-
rion alone decreased with age, F (3, 41) 53.25, p ,.05. However, a Newman-
Keuls test showed a difference between 6 and 22 years only. The number of
blocks necessary to achieve the stability criterion also decreased with age, F (3,
41) 58,81, p ,.001. A Newman-Keuls post hoc showed a significant effect
between 6-year olds and all other groups.
An ANCOVA was conducted with Age as a between-subject factor, Practice
(first block vs last trial block in which the two criteria were achieved) as a
within-subject factor, and the Number of Blocks of trials necessary to achieve the
two criteria as a fixed covariate. We chose an ANCOVA rather than an ANOVA
because we suspected that the differences in number of trial blocks necessary to
achieve the two criteria among participating subjects greatly contributed to the
random variability of the dependent measure. There was a significant interaction
Age and Practice, F (3,40) 511.49, p ,.001. The effect of practice decreased as
age increased. A Newman-Keuls test showed that Practice effect reached signif-
TABLE 2
Mean Number of Blocks Necessary to Achieve the Criterion
at Training Session According to Age
Age
Number of blocks necessary to achieve criterion
Accuracy Stability Both
6 yr. 3, 08 5, 5 7, 25
8 yr. 2, 83 3, 17 5, 08
10 yr. 2 1, 78 2, 44
22 yr. 1, 42 1, 5 1, 67
56 OLIVIER, AUDIFFREN, AND RIPOLL
icance at age, 6, 8, and 10 but not at 22. Mean RT in the first and last training
blocks are reported in Table 3 from age 6 years to 22.
Experimental Session
An analysis of variance was applied to RT, with Age as a between-subject
factor and with Prime Type (three levels: valid, neutral, and invalid), Side (two
levels: right, and left), and Direction (two levels: up, and down) as repeated-
measures factors. Another analysis of variance was realized for arcsine trans-
formed percentage error rates, with Age as a between-subject factor and with
Prime Type (three levels: valid, neutral, and invalid) as repeated-measures factor.
In addition, for the invalid prime condition, an analysis of variance was applied
to RT with Age as a between-subject factor and with Prime Invalidity (three
levels: invalid hand, invalid direction, and both), Side and Direction as repeated-
measures factors. Finally, an analysis of variance was conducted for arcsine
transformed percentage error rates, with Age as a between-subject factor and
with Prime Invalidity and Errors Type (six levels: side error, direction error,
double error, anticipation, too long, and moved) as repeated-measures factors.
Newman-Keuls tests were used for post-hoc comparisons. Preliminary analyses
of each type included sex as a source of variance were applied to RT and
percentage of error, but none revealed a significant main effect of sex nor any
significant interactions with sex.
Reaction Time
The main effect of Age was significant, F(3,39) 551.49, p,.001. RT
decreased as age increased (see Figure 2). The post-hoc analysis showed no
difference in RT between 6- and 8- year olds. All other post-hoc comparisons
were significant. A significant effect of Side was found, F(1,39) 58.69, p,.05.
RTs for right responses (423 ms) were shorter than those for left responses (436
ms). The MANOVA revealed that the main effect for Prime Type was signifi-
cant, V(2,38) 540.29, p,.001. Average RTs were shorter when the prime was
valid (400 ms) than when the prime was neutral (432 ms) or invalid (458 ms).
Moreover, a difference in RT was found between neutral and invalid primed
TABLE 3
Mean Reaction Time (ms) at Training Session According to Age
Age
Reaction time
First block Last block Difference
6 yr. 678 518 160
8 yr. 617 524 95
10 yr. 474 417 57
22 yr. 326 325 1
57MOTOR PROGRAMMING
conditions (see Figure 3). The interaction Age by Prime Type was not significant,
V(6,78) 51.81, p5.11. Costs had a significant effect at all ages, whereas
Benefits had a significant effect at age 6 years, 10 and 22 but not in the neutral
condition. Prime Type interacted significantly with Direction, V(2,38) 59.22, p
,.001. The post-hoc analysis showed an effect of Direction in the valid and
invalid conditions, but not in the neutral condition. RTs were shorter for down
than for up responses.
FIG. 2. Reaction time according to Age.
FIG. 3. Reaction time for each level of prime conditions (valid, neutral, invalid side, invalid
direction, and invalid both).
58 OLIVIER, AUDIFFREN, AND RIPOLL
In the invalid prime condition, there was a significant effect of Age, F(3,39)
540, p,.001, similar in direction to that observed in the previous analysis. Side
had also a significant effect, F(1,39) 59.34, p,.05. RTs for right responses (447
ms) were shorter than those for left responses (463 ms). The MANOVA revealed
a main effect of Prime Invalidity (invalid hand, invalid direction, and both),
V(2,38) 56.31, p,.05. A post-hoc analysis showed that average RTs observed
in the invalid side condition (464 ms) were longer than those recorded in the
invalid direction condition (446 ms). We did not observe any significant differ-
ence between invalid both condition and invalid side or direction conditions (see
Figure 3). No other comparisons including interactions reached significance.
Percentage of Error
The main effect of Age was significant, F(3,39) 56.73, p,.001. The post-hoc
analysis showed a difference between 10-year olds (1.8%) and all other groups
(around 1%). Age interacted with Prime Type, V(6,78) 54.06, p,.01 (see Table
4). The percentage of errors was higher for the invalid than for the valid and
neutral primes at age 10 years and 22, but not at 6 and 8.
The effect of Error Type was significant, V(5,35) 536.67, p,.001. The
post-hoc analysis showed that subjects made more anticipation and direction
errors than side, double, too long and moved errors. The Age 3Error Type
TABLE 5
Mean Errors (%) According to Age for Each Level of Error Type
Age
Error type
Side Direction Double Anticipation Too long Moved
6 yr. 1.7 2.5 0.3 0.7 0.7 0.3
8 yr. 1.6 2.8 0.6 1.0 0.2 0.3
10 yr. 1.5 2.6 0.7 5.3 0.6 0.2
22 yr. 0.4 2.3 0.2 1.6 0.0 0.1
TABLE 4
Mean Errors (%) According to Age for Each Level of Prime Type
Age
Prime type
Valid Neutral Invalid
6 yr. 0.9 1.3 1.0
8 yr. 1.0 0.9 1.3
10 yr. 1.3 1.7 2.4
22 yr. 0.6 0.3 1.4
59MOTOR PROGRAMMING
interaction was significant, V(15,111) 53.70, p,.001 (see Table 5). At age 10
years, the percentage of anticipations was higher than at all other ages. At age 6
years, 8 and 22, percentage of direction errors was higher than the percentage of
all other types of errors. Prime Type interacted significantly with Error Type,
F(10,390) 53.38, p,.001).
In the invalid prime condition, there was a significant effect of Age, F(3,39)
54.56, p,.01. Similarly to the previous analysis, a post-hoc analysis showed
a difference between 10-year olds (2.3%) and all other groups (around 1%). We
observed an effect of Prime Invalidity, V(2, 38) 54.96, p,.05. Subjects made
more errors in invalid hand and invalid direction conditions (1.7%) than when
faced with both invalidities (1%). There was a significant effect of Error Type,
V(5, 35) 514.82, p,.001. Subjects made fewer double, too long and moved
errors than anticipation and side errors, whereas most errors were made in
direction. Prime Type interacted significantly with Error Type, F(10, 390) 5
6.66, p,.001 (since the data matrix was nearly singular and could not be
inverted, the results from the ANOVA were presented).
DISCUSSION
The main aims of this experiment were to (1) study age differences in motor
preparation, and (2) validate the early maturation of effector and direction
specification. We will first discuss results related to those aims, then we will
consider our results on the speed of information-processing related to age and
finally, we will address the issue of the amount of practice necessary to achieve
stability and accuracy at different ages.
Costs and Benefits of Motor Preparation vs. Age
Motor preparation can be analyzed within either a preprocessing or a preset-
ting framework. In preprocessing (pre-programming), the shortening of RT with
advanced information is interpreted as a shift of the motor programming stage
from before to after the response signal. In presetting, the shortening of RT
results from a change in a functional stage of motor programming stage, for
example in the activation of subroutines that form the motor program (Requin,
1985). Similarly to the results obtained by Larish & Frekany (1985) and Lepine,
Glencross & Requin (1989), our results support the main assumption of the
Rosenbaum’s model (Rosenbaum, 1980, 1983), that is the preprocessing of
motor preparation. Subjects needed progressively more time to (1) initiate a
pre-programmed movement, in the case of a valid prime (2) program the effector
and the direction of the movement during the RT, in the case of a neutral prime,
and (3) deprogram and reprogram the pre-programmed movement in the case of
an invalid prime.
According to the costs and benefits viewpoint (Posner & Snyder, 1975), costs
were equal to RTs observed with an invalid prime, minus those observed with a
neutral prime, whereas benefits were equal to RTs observed with a neutral prime,
60 OLIVIER, AUDIFFREN, AND RIPOLL
minus those observed with a valid prime. A decrease in costs and benefits
induced by motor preparation with age would indicate that adults are more
conservative than children, that is adults are more reticent to engage in the
pre-programming of a movement. Research, studying the development of con-
trolled covert orienting of visual attention across age 5 to 25 (Akhtar & Enns,
1989; Enns & Brodeur, 1989; Nougier et al., 1992; Pearson & Lane, 1990),
showed that (1) observers of all ages oriented attention voluntarily to the cued
locations, and that (2) costs and benefits tended to decrease with increasing age.
Considering that attention is related to perceptual processes like preparation is
related to motor processes, similar results could be expected for the development
of voluntary orienting attention and motor preparation. The non-significant
interaction between Age and Prime Type on RT suggests that (1) beyond 6 years
of age, children are capable of using the information provided by the prime to
prepare their movement in advance, (2) costs and benefits of pre-programming do
not vary significantly with age, that is, all subjects, whatever their age, pre-
program response prior to the occurrence of the response signal, (3) the duration
of response deprogramming-reprogramming is quite similar across the four age
groups. These results are in agreement with those of previous studies, which
suggested an early maturation of the programming processes (Clark, 1982, 1987;
Durand & Barna, 1987; Reilly & Spirduso, 1991). Nevertheless, at age 10 years
and 22, subjects made more errors in the invalid prime condition, suggesting
that movement preparation is more important at age 10 years and 22 than at age
8 years.
Deprogramming-Reprogramming of Side and Direction vs. Age
According to the parametric conception of motor programming (Rosenbaum,
1983), the RT lengthening for pointing movements at unexpected target, ob-
served with the priming technique, is interpreted as the summation of the times
necessary to despecify and to respecify the movement dimensions for which the
prime was invalid. By comparing RTs observed in conditions that differ in
number and nature of unprimed dimensions, the time taken for each movement
parameter specification can be inferred. Movement parameters thus are con-
ceived as dynamic and spatial variables that an observer can describe in the
Euclidian geometrical space. In our experiment, two spatial parameters were
manipulated: effector (right or left hand) and direction (up and down). Three
conditions of invalidity were compared wherein: (1) side was invalid when
direction was valid, that is, subjects had to despecify and respecify side only; (2)
direction was invalid when side was valid, that is, subjects had to despecify and
respecify direction only; and finally (3) side and direction were invalid, that is,
subjects had to despecify and respecify both side and direction. Condition 1
allowed us to infer the duration of side deprogramming-reprogramming, condi-
tion 2, the duration of direction deprogramming-reprogramming, and condition 3,
the duration of side 1direction deprogramming-reprogramming. If movement
61MOTOR PROGRAMMING
programming is serial, the longest RTs should be found when subjects depro-
gramm and reprogramm both side and direction. Results showed that the depro-
gramming and reprogramming of direction were shorter than that of side, but not
than that of both side and direction. This last result does not support a serial view
of the programming process. Similar results led Le´pine, Glencross & Requin
(1989) to propose an alternative model in which, when all movement dimensions
have to be deprogrammed and then reprogrammed, an association among
unprimed dimension values allows a joint or parallel process, which is faster than
a serial deprogramming-reprogramming of all parameters.
Finally, the non-significant interaction between Age and Prime Invalidity on
RT suggests that deprogramming and reprogramming of effector and direction
are identical during the course of child development.
Improvement of Information Processing Speed vs. Age
Results show that RTs decreased as age increased particularly from 8 to 22.
There are two potential explanations for this result: (1) information processing
speed improves throughout ontogenesis or (2) a speed-accuracy trade-off occurs
during the course of development. The first hypothesis should be supported by a
constant error rate from 6 to 22 years of age, whereas the errors rate should
increase with age if a speed-accuracy trade-off is taking place. Results showed
that error rate remained stabilized around 1% from 6 to 22, except for the
10-year-old group. At 10, the shortening of RT and the high percentage of errors,
particularly anticipation errors, suggest a change in the speed-accuracy criterion,
then ten years old children privileging rapidity of response. Similar findings were
reported by others for on-line processing also (Bard & Hay, 1983; Bard, Fleury,
& Gagnon, 1990).
The decreasing RT from 8 to 22 can be rather ascribed to an improvement in
information processing speed rather than to a change in strategy. According to
our results, a critical period in the maturation of information processing is taking
place at age 8 years. These results are in agreement with those of Hauert &
Pellizer (1992), who considered that, at age 8 years, children transfer from spatial
to motor coordinates while aiming at visual targets. A more structural factor
could also induce a decrease in RT with age: myelination in the human cortex
throughout childhood (Benes, 1989; Hatta & Moriya, 1988). Myelination, there-
fore, seems to be a good candidate to explain the improvement in information
processing speed with maturation.
Magnitude of Practice Effect vs. Age
In the present study, children’s performance was stabilized before the exper-
iment, allowing the collect of data from well practiced subjects and the respect
of the processing stage output during the experiment (Sanders, 1980, 1990). As
mentioned above, during the training session, children had to achieve both a
62 OLIVIER, AUDIFFREN, AND RIPOLL
criterion of accuracy and of stability. Such a procedure yielded results worth
commenting.
The number of trials necessary to achieve the two criteria decreased as age
increased. More precisely, 6- and 8- year olds needed more than twice the
number of blocks of trials than 10- and 22-year olds. It was significantly more
difficult for 6-year-old children to achieve the stability criterion than for all other
groups. Our results are in agreement with those reported in the literature. Indeed,
intra-individual variability has been shown to decrease systematically with age
(Carron, 1971; Eckert & Eichorn, 1977). This may be related to a low degree of
maturation of the different systems, for instance a low degree of control of
muscular fibres. In addition, our results give credit to the existence of a critical
period, characterizing the psychomotor development of children, at approxi-
mately 7 to 8 years of age (Bard & Hay, 1983; Bard, Fleury, & Gagnon, 1990;
Hauert & Pellizzer, 1991).
Let it be added that the effect of practice on information-processing speed
decreased as age increased, suggesting that the younger the subjects, the larger
their capacity to improve their performance. This is probably due to their low
initial level of performance.
REFERENCES
Akhtar, N., & Enns, J. T. (1989). Relations between covert orienting and filtering in the development
of visual attention. Journal of Experimental Child Psychology, 48, 315–334.
Bard, C., & Hay, L. (1983). Etude ontoge´ne´tique de la coordination visuo-manuelle. [Ontogenesis
study of visuo-manual coordination]. Revue Canadienne de Psychologie, 37, 390–413.
Bard, C., Fleury, M., & Gagnon, M. (1990). Coincidence-anticipation timing: An age-related
perspective. In C. Bard, M. Fleury & L. Hay (Eds.), Development of eye-hand coordination
across the life span (pp. 283–305). Columbia: University of South Carolina.
Bartlett, M. S. (1939). A note on tests of significance in multivariate analysis. Proceedings of the
Cambridge Philosophical Society, 35, 180–185.
Benes, F. M. (1989). Myelination of cortical-hippocampal relays during late adolescence. Schizo-
phrenia Bulletin, 15, 585–593.
Bertelson, P. (1968). Quelques proble`mes expe´rimentaux pose´s par l’e´tude de la pre´paration a`re´agir.
[Some experimental problems for study of preparation to react]. Bulletin de Psychologie, 276,
653–659.
Box, G. E. P. (1954). Some theorems on quadratic forms applied in the study of analysis of variance
problems, II. Effects of inequality of variance and correlation between errors in the two-way
classification. Annals of Mathematical Statistics, 25, 484–498.
Carron, A. V. (1971). Motor performance and response consistency as a function of age. Journal of
motor Behavior, 3, 105–109.
Clark, J. E. (1982). Developmental differences in response programming. Journal of Motor Behavior,
4, 247–254.
Clark, J. E. (1987). Age-related differences in programming a movement. In J. E. Clark & J. H.
Humphrey (Eds.), Advances in Motor Development Research I (pp. 95–104). AMS Press, Inc.
New York.
Crossman, E. R. F. W., Goodeve, P. J. (1983). Feedback control of hand-movement and Fitts’ law.
Quarterly Journal of Experimental Psychology, 35, 251–278.
Durand, M., & Barna, R. (1987). Exigences de la taˆche et performances motrices: Etude de´veloppe-
63MOTOR PROGRAMMING
mentale. [Task requirement and motor performances: A developmental study]. In: A. Vom Hoff
& R. Simonnet (Eds.), Sport Psychology Research (pp. 150–161). E. A. P., Paris.
Eckert, H. M., & Eichorn, D. H. (1977). Developmental variability in reaction time. Child Devel-
opment, 48, 452–458.
Enns, J. T., & Brodeur, D. A. (1989). A developmental study of covert orienting to peripheral visual
cues. Journal of Experimental Child Psychology, 48, 171–189.
Geiser, S., & Greenhouse, S. W. (1958). An extension of Box’s results on the use of the F-distribution
in multivariate analysis. Annals of Mathematical Statistics, 29, 885–891.
Greenhouse, S. W., & Geiser, S. (1959). On methods in the analysis of profile data. Psychometrika,
24, 95–112.
Hatta, T., & Moriya, K. (1988). Developmental changes of hemisphere collaboration for tactile
sequential information. International Journal of Behavioral Development, 4, 451–564.
Hauert, C. A., & Pellizzer, G. (1992). De´veloppement du pointage visuo-manuel chez l’enfant:
Donne´es re´centes. [Child development of visuo-manual aiming: Recent views]. In M. Laurent,
J-F. Marini, R. Pfister & P. Therme (Eds.), Research in A.P.S. (Vol. 3, pp. 59–66). Paris: Actio.
Universite´ Aix-Marseille II (UFR STAPS).
Hauert, C. A., Zanone, P. G., & Mounoud, P. (1990). Development of motor control in the child:
Theoretical and experimental approaches. In O. Neumann & W. Prinz (Eds.), Relationships
between perception and action (pp. 325–343). Springer-Verlag, New York, London, Paris.
Hay, L. (1987). Etude ontoge´ne´tique du controˆle d’un mouvement: L’approche manuelle. [Ontoge-
netic study of motor control: Manual approach]. Unpublished doctoral dissertation, Faculty of
Sciences of Luminy, University of Aix-Marseille II, France.
Henry, F. M., & Rogers, D. E. (1960). Increased response latency for complicated movements and
a ‘‘memory drum’’ theory of neuromotor reaction. Research Quarterly, 31, 448–458.
Hotelling, H. (1951). A generalized t test and measure of multivariate dispersion. In J. Neyman (Ed.),
Proceedings of the Second Berkeley Symposium on Mathematical Statistics and Probability (pp.
23–41). Berkeley: University of California Press.
Huynh, H., & Feldt, L. S. (1976). Estimation of the Box correction for degrees of freedom from
sample data in the randomized block and split-plot designs. Journal of Educational Statistics, 1,
69–82.
Keselman, J. C., & Keselman, H. J. (1990). Analysing unbalanced repeated measures designs. British
Journal of Mathematical and Statistical Psychology, 43, 265–282.
Kornblum, S. (1965). Response competition and/or inhibition in two-choice reaction time. Psy-
chonomic Science, 2, 55–56.
LaBerge, D., Van Gelder, P., & Yellott, J. (1970). A cuing technique in choice reaction time.
Perception and Psychophysics, 8, 51–62.
Larish, D. D., & Frekany, G. A. (1985). Planning and preparing expected and unexpected move-
ments: Reexamining the relationships of arm, direction, and extent of movement. Journal of
Motor Behavior, 17, 168–189.
Le´pine, D., Glencross, D., & Requin, J. (1989). Some experimental evidence for and against a
parametric conception of movement programming. Journal of Experimental Psychology: Hu-
man Perception and Performance, 15, 347–362.
McCall, R. B., & Appelbaum, M. I. (1973). Bias in the analysis of repeated-measures designs: Some
alternative approaches. Child Development, 44, 401–415.
O’Brien, R., & Kister Kaiser, M. (1985). MANOVA method for analysing repeated measures
designs: An extensive primer. Psychological Bulletin, 97, 316–333.
Olivier, I., Ripoll, H., & Audiffren, M. (1994). The development of programming and executing of
ball interception in children. Journal of Sport and Exercise Psychology, 16, 92.
Olson, C. L. (1974). Comparative robustness of six tests in multivariate analysis of variance. Journal
of the American Statistical Association, 69, 894–908.
Pearson, D. A., & Lane, D. M. (1990). Visual attention movements: A developmental study. Child
Development, 61, 1779–1795.
64 OLIVIER, AUDIFFREN, AND RIPOLL
Pillai, K. C. (1955). Some new test criteria in multivariate analysis. Annals of Mathematical Statistics,
26, 117–121.
Posner, M. I., & Snyder, C. R. R. (1975). Facilitation and inhibition in the processing of signals’. In
P. M. A. Rabbitt & S. Dornic (Eds.), Attention and performance V (pp. 669–682). Academic
Press: London.
Reilly, M. A., & Spirduso, W. W. (1991). Age-related differences in response programming.
Research Quarterly for Exercice and Sport, 62, 178–186.
Requin, J. (1985). Looking forward to move soon, ante factors selection process in motor behavior.
In M. I. Posner & O. Marin (Eds), Attention and performance XI (pp.). Hillsdale, NJ: Erlbaum.
Rogan, J. C., Keselman, H. J., & Mendoza, J. L. (1979). Analysis of repeated measurements. British
Journal of Mathematical and Statistical Psychology, 32, 269–286.
Rosenbaum, D. A. (1980). Human movement initiation: Specification of arm, direction and extent.
Journal of Experimental Psychology: General, 109, 444–474.
Rosenbaum, D. (1983). The movement precuing technique: Assumptions, applications, and exten-
sions. In R. Magill (Ed.), Memory and the control of action (pp. 231–274). Amsterdam: North
Holland.
Rosenbaum, D., & Kornblum, S. (1982). A priming method for investigating the selection of motor
responses. Acta Psychologica, 50, 223–243.
Roy, S. N. (1953). On a heuristic method of test construction and its use in multivariate analysis.
Annals of Mathematical Statistics, 24, 220–238.
Sanders, A. F. (1980). Stage analysis of reaction processes. In G. E. Stelmach & J. Requin (Eds.),
Tutorials in motor behavior (pp. 331–354). North-Holland Publishing Company.
Sanders, A. F. (1990). Issues and trends in the debate on discrete vs. continuous processing of
information. Acta Psychologica, 74, 123–167.
Schmidt, R. A. (1975). A schema theory of discrete motor skill learning. Psychological Review, 82,
225–260.
Schmidt, R. A. (1988). Motor control and learning: A behavioral emphasis (2nd ed.). Champaign, IL:
Human Kinetics.
Schmidt, R. A., Zelaznik, H. N., Hawkins, B., Franck, J. S., & Quinn, J. T. (1979). Motor-output
variability: A theory for the accuracy of rapid motor acts. Psychological Review, 86, 415–451.
Sugden, D. A. (1980). Movement speed in children. Journal of Motor Behavior, 12, 125–132.
Thomas, J. R. (1980). Acquisition of motor skills: Information processing differences between
children and adults. Research Quarterly for Exercise and Sport, 51, 158–173.
Welsandt, R. F., Zupnick, J. J., & Meyer, P. A. (1973). Age effects in backward visual masking
(Crawford paradigm). Journal of Experimental Child Psychology, 15, 454–461.
Wickens, C. D. (1974). Temporal limits of human information processing: A developmental study.
Psychological Bulletin, 81, 739–755.
Wilk, S. S. (1932). Certain generalizations in the analysis of variance when the correlations have a
certain pattern. Psychometrika, 44(2), 229–233.
Received: September 25, 1996; revised: January 13, 1998
65MOTOR PROGRAMMING
