Implicit and explicit motor learning in preterms.

Motor skills can be learned implicitly and explicitly. Implicit learning is the
ability to acquire a new skill by doing it unaware of the regularities
governing the task. The procedural knowledge gained is difficult or even
impossible to access consciously and has been shown to be relatively
independent of both age (Meulemans et al., 1998) and IQ (Reber et al., 1991).
In explicit learning, on the other hand, declarative knowledge is used to build
up a set of performance rules that guides motor output. Berry and Broadbent
(1988) demonstrated that the application of declarative knowledge requires the
availability of working memory, whereas, the application of procedural
knowledge does not.
Ample studies on implicit and explicit learning in neurologically intact
individuals have assumed that explicit learning may only proceed with intact
working memory (Maxwell et al., 2003). In the proposed project, we will
examine motor learning in preterm children. These children are likely to
develop working memory deficits, despite the absence of known neurological
disorders (Stewart et al., 1999). Moreover, these children are highly at risk
to develop motor performance and learning problems (Evensen et al., 2004; De
Kleine et al., 2006). There have been no studies in preterms that directly
relate deficits in working memory formation to motor skill learning, despite
its theoretical relevance for understanding the cognitive factors affecting
learning and its applied relevance for rehabilitation and recovery (Krakauer,
2006).

The first aim in the proposed project is to directly test the alleged role of
working memory on (explicit) motor skill learning in preterm children.
Our second aim is to examine the relation between implicit and explicit
learning in preterms. This is especially relevant, as the potential mediating
role of explicit working memory for this interaction can be unveiled.
Specifically, if working memory is indeed an important factor mediating the
(facilitatory or inhibitory) interaction effects found in previous studies, we
do not expect such effects in the group preterms with profound working memory
deficits. Finding such an absence of interference in preterms with deficits in
working memory allows us to draw more specific inferences related to cognitive
substructure of motor learning.
The third aim is to study the formation of memory representations underlying
explicit and implicit learning, until present not systematically been studied
in preterms. It is still an open question how the deficit in working memory
affects the nature of memory representations sub serving both types of
learning. It may be speculated that the memory representation sub serving
explicit learning in the preterms is incompletely formed, or even impossible to
form, owing to a deficit in working memory, and as a consequence these
participants may need to *resort to* effector-dependent memory representations
for this type of sequence learning.

Randomized experimental design with two *experimental* groups and one control
group of the same age range (5-8 years), yielding three groups: -1- preterms
with mild/absent working memory deficits, -2- preterms with severe working
memory deficits and -3- typically developing children matched on age and IQ.
These three groups allow us to make the relevant comparisons of interest
related to our research questions, while at the same time control for the
effect of prematurity per se on learning (addition of group 2).
Implicit and explicit learning will be examined in the three groups separately
on two distinct tasks that are modified versions of the Serial Response Time
Task (SRT). The first task will be performed on a digitizing tablet, and the
second task will be performed on a customized button-box task. First, the SRT
task is probably the most robust paradigm to study implicit and explicit
learning, already successfully applied in clinical groups, second, the use of
pen-displacement data and the concomitant analysis of multiple variables allows
us to more precisely determine the different levels at which learning might be
effective.

In the basic set-up of the SRT tasks the participant is required to react as
quickly as possible to stimuli that are displayed on the computer monitor by
either moving the pen to the corresponding location on the digitizing tablet,
or by pressing the corresponding key on the button-box.
During implicit learning trials the sequence of stimuli follows a structured
pattern, i.e., is governed by a set of rules or is completely random. Trials
with random sequence are intermixed with those having a structured sequence,
such that discovery of the structured sequence is made less probable.
During explicit learning trials children are informed either verbally (e.g.,
numbering of consecutive stimuli) or visually (the figure they have to draw)
about the sequential ordering of the stimuli.

Aim #1: In the first series of experiments we will scrutinize the alleged role
of working memory for learning in preterms, in particular for explicit
learning. Children of each of the three groups will be randomly assigned to two
subgroups with a different task order, yielding a total of 6 groups (n=15 for
each group). All children learn both tasks implicitly and explicitly, the order
of which is counterbalanced. During the learning phase, all children will learn
both tasks in 2 consecutive days. As explicit knowledge about sequence
regularities on a specific task may disturb implicit learning, all children
will start each task in an implicit learning condition, either in the first
session or in the second session following a pause period of 1 month (to
obviate possible carry-over learning effects). After learning-day 2 in the
first session and learning-day 4 in the second session participants enter the
test session after a pause of half an hour.

In the test session, participants will first repeat both tasks to assess the
extent to which learning had taken place under implicit and explicit
conditions. For this purpose, participants also have to perform both tasks
under secondary task loading (e.g., counting, or random letter generation) to
test the assumption that performance is more robust for the *implicit learners*
under secondary task loading.

Aim #2: In the second series of experiments we focus on the interaction between
implicit and explicit learning, specifically, on the effect of explicit
information on an implicitly learned task. This is examined for one of the two
experimental tasks described above, the choice of which is based on the results
of the first series of experiments. Again the three experimental groups are
split in half, yielding six experimental groups (n=15). Participants are
randomly assigned to either the *pure* implicit learners (control groups), or
implicit learners that receive explicit learning trials also (experimental
groups). For all groups, learning will take place on 3 consecutive days. On day
2 and 3 of learning, the experimental groups are given explicit information
about the task. Specifically, they are informed on the repeating sequence in
some of the trials, or are provided with a schematic drawing of the repeating
sequence. Participants in the control groups continue to learn implicitly on
days 2 and 3. At the end of day 3, participants enter the test session which
will be similar to test session described in the first series of experiments
(Aim #1). After 1 week, retention testing is performed on day 4 for all
participants.

Aim #3: In the third series of the experiments, we will examine the nature of
the memory representations sub serving implicit and explicit learning. For this
purpose, the button-box task is used. As in the first series of experiments
participants are randomly assigned to the implicit or explicit learning group.
Participants will learn a button-box task, i.e. learning a particular sequence
of the keys, on 3 consecutive days. One of the crucial measurements in the
third series of experiments is the transfer test that is administered following
each day of learning, and at day 4. Specifically, transfer of learning from a
learned condition to a novel condition is evaluated, the methodology of which
closely follows Bapi et al. (2000). These transfer tests allows us to infer if
memory representations are coded in spatial or motor coordinates.

Three series of experiments are conducted. In each series of experiments a new
number of children are recruited. So this means that each child participates a
maximum of 4 days in one experiment. The pause time differs in the first,
second and last series. The duration of each session is 30-45 minutes, so a
total time of 2-3 hours. The tasks are age related and children like to learn
such computer tasks.
There is no risk in these task conditions. For the parents of the children some
burden is present because at this age they have to transport them or to reserve
time in their own time schedule for the visits.
However, insight in learning processes in this group of children is of great
importance to build up a body of knowledge in the rehabilitation programs.

The control children will be tested at school. These children will perform the
some neuropsychological tests to allow comparison with the preterms. So for
them the time investment per child is prolonged with 2 hours. However, also
these children like to do the tests, so no burden or risk is present. However,
interference with school activities can be present. We will arrange the
organisation of the study activities in consultation with teachers and parents
to reduce the burden as most as possible.