The Relationship Between Action Execution, Imagination, and Perception in Children by Emma Jane Yoxon A thesis submitted in conformity with the requirements for the degree of Master of Science Exercise Science University of Toronto © Copyright by Emma Yoxon 2015
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
The Relationship Between Action Execution, Imagination, and Perception in Children
by
Emma Jane Yoxon
A thesis submitted in conformity with the requirements for the degree of Master of Science
Exercise Science University of Toronto
© Copyright by Emma Yoxon 2015
ii
The relationship between action execution, imagination, and
perception in children
Emma Yoxon
Master of Science
Exercise Science
University of Toronto
2015
Abstract
Action simulation has been proposed as a unifying mechanism for imagination, perception and
execution of action. In children, there has been considerable focus on the development of action
imagination, although these findings have not been related to other processes that may share
similar mechanisms. The purpose of the research reported in this thesis was to examine action
imagination and perception (action possibility judgements) from late childhood to adolescence.
Accordingly, imagined and perceived movement times (MTs) were compared to actual MTs in a
continuous pointing task as a function of age. The critical finding was that differences between
actual and imagined MTs remained relatively stable across the age groups, whereas perceived
MTs approached actual MTs as a function of age. These findings suggest that although action
simulation may be developed in early childhood, action possibility judgements may rely on
additional processes that continue to develop in late childhood and adolescence.
iii
Acknowledgments
I am very grateful to have been surrounded by such wonderful people throughout this process.
To my supervisor, Dr. Tim Welsh, thank you for all of the wonderful opportunities your
supervision has afforded me. Your unwavering support created an environment for me to be
challenged but also free to engage in new ideas and interests. Thank you for your
encouragement, constant guidance, confidence and, of course, for making time to answer my
numerous questions and queries. To Dr. Luc Tremblay, thank you for challenging me and
creating many moments for me to think in new and innovative ways. To Dr. Jessica Brian, your
guidance, input and feedback in developing the thesis project were greatly appreciated. To Dr.
Mark Shmuckler, thank you for the helpful comments and criticisms that helped to shape the
final document. To all of my lab-mates in the AA and PMB lab, thank you for the continuous
support, reassurance and stimulating discussion. I would like to extend a special thank you to
Sandra Pacione for being by my side throughout this process and for being a great lab-mate,
friend and shoulder to lean on. Thank you to all of my friends in Toronto and around the world
for your advice, inspiration and friendship. To Drew, thank you for listening to my fears and
doubts, for your advice, guidance, encouragement and all of your love. Finally, thank you to my
family: Mom, Dad, Renée and Ian for all your love and support leading up to and throughout this
process.
iv
Table of Contents
Acknowledgments.......................................................................................................................... iii
Table of Contents ........................................................................................................................... iv
List of Tables ................................................................................................................................. vi
List of Figures ............................................................................................................................... vii
List of Appendices ....................................................................................................................... viii
Introduction ......................................................................................................................................1
Chapter 1 Literature Review ............................................................................................................3
1.1 Common Coding Theory ........................................................................................................3
1.1.1 Neural Simulation of Action ....................................................................................4
1.1.2 Simulation as a Mechanism for Motor Imagery ......................................................5
1.1.3 Simulation as a Mechanism for Action Possibility Judgements ..............................7
1.2 Development .........................................................................................................................11
1.2.1 Typical Motor Development ..................................................................................11
1.2.2 Motor Imagery Development .................................................................................12
1.2.3 Representing the Actions of Others .......................................................................14
1.3 Summary ...............................................................................................................................15
1.4 The Current Project: Purpose and Specific Hypotheses .......................................................16
Chapter 2 The Independent Development of Action Imagination and Perception ........................18
2.1 Abstract .................................................................................................................................18
2.2 Introduction ...........................................................................................................................19
2.2.1 Neural simulation of action ....................................................................................20
2.2.2 Action simulation in children.................................................................................21
2.3 Methods.................................................................................................................................23
v
2.3.1 Participants .............................................................................................................23
2.3.2 Study Design and Tasks .........................................................................................23
2.4 Results ...................................................................................................................................28
2.4.1 Fitts’ Law ...............................................................................................................28
2.4.2 Group Differences ..................................................................................................30
2.4.3 The relationship between age and simulation congruency ....................................31
2.5 Discussion .............................................................................................................................33
2.5.1 Fitts’ law in imagination and perception ...............................................................33
2.5.2 The relationship between age and action imagination ...........................................34
2.5.3 The relationship between age and action perception .............................................35
2.5.4 Conclusions ............................................................................................................37
Chapter 3 Summary and Conclusions ............................................................................................38
3.1 Summary ...............................................................................................................................38
3.2 Conclusions ...........................................................................................................................39
3.3 Limitations and Future Directions ........................................................................................40
References ......................................................................................................................................43
Appendices .....................................................................................................................................51
Appendix A: Edinburgh Handedness Questionnaire .....................................................................52
Appendix B: Additional Analysis ..................................................................................................53
vi
List of Tables
2.1 Fitts’ law equations and statistical analysis for the linear regressions calculated between
MT and ID for each of the tasks and groups……………………………………………. 30
vii
List of Figures
2.1. An example of the pictures that were displayed in the perception task. Images 1 (hand on
the right side target) and 2 (hand on the left side target) were alternated at a range of
SOAs to create the apparent motion of the hand. ............................................................ 27
2.2 Linear regressions between index of difficulty (ID) and movement time for each of the
three tasks, for each of the three experimental groups. ………………………………… 29
2.3. Mean imagined MTs for each of the execution, perception and imagination tasks.
Asterisks indicate significant (Tukey’s HSD, p < .05, CV = 79.9) within group
differences between the tasks. ………………………………………………………….. 32
2.4. Difference scores between imagination and execution (A) and perception and execution
(B) as a function of age. Note: This analysis includes only child participants. ...……… 32
4.1. Ln transformation of ratios of imagination and perception MT as a function of actual
execution movement time for each of the groups. ……………………………………... 54
4.2. Ln transformation of ratios of imagination and perception MT (as a function of actual
execution MT), correlated with age. …………………………………………………… 54
viii
List of Appendices
Appendix A: Edinburgh Handedness Questionnaire .....................................................................52
Appendix B: Additional Analysis ..................................................................................................53
1
Introduction
The experiment reported in the present thesis evaluates the nature of internal action
simulation in action imagination and perception. In computer science, simulators focus on
modeling the internal components or state of a system, whereas emulators focus on modeling the
system’s outcomes or outwardly observable behaviours. Similarly, in the neural control of
movement, simulations are said to serve as internal representations of actions; a way for the brain
to covertly rehearse the neural components of action without outwardly performing a movement
(Jeannerod, 2001). One of the more straightforward or recognizable uses of action simulation is
in the imagination of actions or motor imagery (MI). MI is the conscious and purposeful
simulation of a given action without actually performing the movement (Decety, 1996). MI is
also a widely used experimental paradigm that is thought to serve as a window into other motor
cognitive processes that incorporate motor simulations in more unconscious ways (Munzert,
Lorey, & Zentgraf, 2009). For instance, action simulation has been implicated in a variety of
processes including mental practice and observational learning (Jeannerod, 2001), coordination
of one’s actions with another’s (Sebanz & Knoblich, 2009), action possibility judgments
(Chandrasekharan, Binsted, Ayres, Higgins, & Welsh, 2012; Eskenazi, Rotshtein, Grosjean, &
Knoblich, 2012; Grosjean, Shiffrar, & Knoblich, 2007; Welsh, Wong, & Chandrasekharan,
2013) and even in understanding the intentions of another via the perception of their actions
(Blakemore & Decety, 2001). Consequently, the study of explicit MI is thought to provide a
wealth of information on how individuals cognitively represent action and how differences in
these representations influence how we interact with others and our environment.
Interestingly, while the developmental trajectory of MI has been of interest to researchers
for some time, there has been little experimental evidence to link this trajectory to the
development of other motor cognitive processes such as action possibility judgements or
interpersonal action coordination. This is a meaningful theoretical question as an individual’s
ability to engage in MI is thought to reflect their ability to simulate action, which can impact the
processes described above. The purpose of the research presented in this thesis, therefore, was to
answer this question by investigating how the relationship between imagination, perception and
execution develops across childhood. Prior to outlining the specific hypotheses of the thesis
2
experiment, Chapter 1 consists of a review of current literature on action simulation and motor
development. Chapter 2 then presents an academic paper describing the thesis experiment that is
formatted for submission. This is followed by general conclusions and future directions in
Chapter 3.
3
Chapter 1 Literature Review
1.1 Common Coding Theory
Everyday interactions with one another rely on the understanding and interpretation of
others’ actions. Simple gestures such as those in conversation can not only direct attention but
can also convey emotions or concepts. In a classroom, for instance, a teacher can direct the
class’s attention towards a particular word on the chalkboard simply by pointing towards it. In
theatre, an actor can punctuate their lines with gestures in order to more elaborately convey
emotion. In sports, coaches often use gestures and hand signals to better convey information or
teach skills. It is evident, therefore, that the perception of action affords insight into the mental
states of another, without the need for actually performing the observed action. Similarly, the
imagination of action allows one to experience movements without the need to actually perform
the action. Athletes, for example, will often mentally take themselves through a routine, game or
skill in preparation for competition. This kind of mental practice can facilitate training without
the need for excess physical exertion. Critically, both perception and imagination allow one to
experience an action without actually executing the observed or imagined action. This notion has
been explored a great deal by modern researchers and the results of this work highlight the
interrelated nature of action perception, imagination and execution.
The relationship between action and perception has been of interest to scientists for over
a hundred years (James, 1890; Lotze, 1852). These early scientists put forth the notion that
actions can be represented in terms of the perceptual effects they have on the environment. This
notion forms the basis of modern common coding theory (Hommel, Müsseler, Aschersleben, &
Prinz, 2001; Hommel, 2009; Prinz, 1997). The main tenet of this approach is that there are
certain aspects of perception and action that share a common representational domain such that
action codes are tightly bound to the perceptual codes that represent the effects of those actions.
In this way, actions may be planned in terms of their effects and, in the reverse direction,
perception of effects can lead to the occurrence of a related action (Prinz, 1997). For example,
when driving a car, the desire to slow down the vehicle (a perceptual effect) will activate the
action to step on the brakes. In the reverse direction, the action of stepping on the brakes can
4
allow one to predict when the car will come to a stop. Furthermore, in Hommel’s theory of event
coding (TEC), he states that cognitive representations of actions and perceptual effects are made
up of smaller feature codes that create an overall representation of the event. It is these larger
cognitive representations of action and perception that become bound through experience and
can consequently be related to the same event (Hommel, 2009). Critically, the binding of action
and perception relies heavily on the cognitive representation of an action and its effects, a
process which may be facilitated by the neural simulation of action.
The notion that neural representations of actions and effects are closely related has
implications beyond the planning and control of goal-directed actions in the domain of motor
cognition. In action imagination, for example, it has been suggested that bound neural codes can
be activated at sub-motor threshold levels such that by imagining the perceptual effects of an
action (for example, the visual feedback received) one can effectively rehearse the motor
components of that action (Wong, Manson, Tremblay, & Welsh, 2013; Yoxon, Tremblay, &
Welsh, 2015). Furthermore, it has been demonstrated that the perception of another’s actions can
activate the cognitive representation of that action in the self (van der Wel, Sebanz & Knoblich,
2013). This mechanism allows one to interpret the actions of another to make decisions about the
action possibility, coordinate one’s actions with another or, in a social context, infer intentions or
desires (Blakemore & Decety, 2001). In all of these processes, however, motor output is
activated at levels that are below the threshold for actual actions to emerge. Effectively, in all of
these cases, perceptions or thoughts of perceptions activate simulations of the desired action.
That is, the activation of a perceptual representation in turn leads to the activation of the internal
components of an action, without acting on the environment. This approach heavily implicates
the neural simulation of action in the success of the motor cognitive processes described.
1.1.1 Neural Simulation of Action
Marc Jeannerod (1995, 2001) introduced the notion that the neural simulation of action
could be the singular mechanism underlying various processes associated with motor cognition.
In his neural simulation theory, he states the neural simulation of action occurs whenever there is
activation of the cortical areas associated with movement in the absence of any voluntary action
(Jeannerod, 2001). These simulation or “s-states” were implicated in a myriad of tasks from the
5
more conscious or deliberate simulation in action imagination to tasks that involve primarily
subconscious simulation such as action possibility judgements. What is of particular interest is
that despite the obvious differences in the consciousness of simulation between these two tasks,
both have been shown to elicit activation of cortical motor areas. It is for this reason that
Jeannerod (2001) suggested neural simulation of action to be a unifying mechanism for such
motor cognitive processes. Moreover, the neural simulation of action may be an important
mechanism for the common representation of action and perception, as proposed in the common
coding theory discussed in the previous section. The following subsections recount both the
neurophysiological and behavioural evidence that supports the presence of simulation states in
motor imagery (MI) and in action possibility judgements.
1.1.2 Simulation as a Mechanism for Motor Imagery
Evidence for the neural simulation of action in MI is supported considerably by
neurophysiological research demonstrating overlapping neural networks in MI and action
execution. The vast majority of this supporting evidence comes from neuroimaging work over
the last decade. It was proposed early on that MI activates cortical regions that are implicated in
the planning and control of voluntary movement and that these areas also overlap with other
processes considered in motor cognition (Grèzes & Decety, 2001; Jeannerod, 2001). Recently,
meta-analyses of neuroimaging studies have clearly established that MI activates a fronto-
parietal region similar to that of actual execution. The cortical areas that have been the most
clearly implicated are the supplementary motor area (SMA), pre-motor cortex (PMC), and the
primary motor cortex as well as various areas of the parietal cortex. Importantly, all of these
areas are heavily implicated in the planning, execution and online control of movement (Hétu et
al., 2013; Munzert et al., 2009). Furthermore, recent work has demonstrated that within
participantsthere are many overlapping features of the neural networks for execution and
imagination (Sharma & Baron, 2013). Overall, the body of neuroimaging research on this topic
supports the notion that deliberate imagination of actions can lead to activation of neural
networks similar to that for actual execution, supporting the notion that action imagination
involves a simulation of the internal components of movement.
6
Moreover, dozens of studies utilizing transcranial magnetic stimulation (TMS) have
provided evidence consistent with the results of neuroimaging studies. The overall finding of this
body of literature is that the activity of the corticospinal tract is increased when an individual
imagines themselves performing movements (Clark, Tremblay, & Ste-Marie, 2004). Specifically,
it has been shown that motor evoked potentials (MEP) are greater in action imagination than at
rest or in other non-motor tasks, such as strictly visual imagery or mental counting (see Munzert
et al., 2009 for review). These activation effects are known to be specific to the musculature to
be recruited in the imagined movement and to be stronger when stimulation is contralateral to
movement (Fadiga et al., 1998). Additionally, a recent study combining transcranial direct
current stimulation (tDCS) with TMS has implicated the parietal cortex, much like the previously
described neuroimaging studies (Feurra et al., 2011). Essentially, the Feurra et al. (2011) study
demonstrated that the decreased or increased activation of the parietal cortex (via tDCS)
modulated the MEPs recorded from hand muscles when TMS was applied over primary motor
cortex during action imagination. This result reflects the use of ipsilateral connections between
parietal and motor cortex in action imagination, similar to the use of this network in actual
execution. The results of these TMS studies highlight that MI makes use of a similar neural
network to that of execution. Altogether, the findings of neurophysiological examinations of MI
support the simulation of cortical motor networks in explicit MI.
In addition to the wealth of neurophysiological research on the neural underpinnings of
action imagination, there has also been an accumulation of behavioural data demonstrating the
consistencies between real and imagined movements. Much of this evidence uses experimental
paradigms employing mental chronometry. These paradigms compare the amount of time
necessary to complete an imagined movement to the amount of time needed to actually execute
the movement. One of these first experiments demonstrated that the time necessary to imagine a
graphic tracing task was similar to the time necessary to actually complete the task (Michel &
Decety, 1989). These authors concluded that the temporal similarity in actual and imagined tasks
was evidence for similarity in the cortical mechanisms involved. Decety and Jeannerod (1995)
also supported this principle by demonstrating the presence of a well-characterized speed-
accuracy trade-off, known as Fitts’ law, in imagined movements. Fitts’ law characterizes the
linear relationship between the amount of time necessary to complete a task and the accuracy
7
demands of that task. This relationship is denoted in the formal equation: MT = a + b(ID), where
MT is movement time, ID is the index of difficulty (i.e., movement demand) and “a” and “b” are
constants relating to an individual’s base MT and the unit increase in MT as a function of ID,
respectively. In Fitts’ original task, participants touched back and forth between targets of
varying width and movement amplitude. In this context, the ID component of Fitts’ equation is a
function of the target width and movement amplitude and is broken down as ID = log2(2A/W),
where A is the movement amplitude and W is the target width (Fitts, 1954). In Jeannerod and
Decety’s (1995) task, participants imagined themselves walking toward doors of varying
openings and walking distances. Critically, imagined walking paths conformed to Fitts’ law.
That is, imagined walking paths took a longer amount of time when the opening was narrower
and/or the walking path was longer. Similarly, Sirigu and colleagues (1995) replicated the
presence of Fitts’ law in imagined movements when they had a patient with a unilateral lesion
both imagine and move a pen from a starting position to a target of varying widths. Recent work
has corroborated this result in neurotypical individuals (Wong et al., 2013; Yoxon et al., 2015).
Interestingly, the majority of studies that have compared actual and imagined MTs have
found that imagined MTs are consistently greater than actual MTs (e.g.Wong et al., 2013;
Young, Pratt, & Chau, 2009; Yoxon et al., 2015). It is unclear why this pattern of MTs
consistently emerges, but the critical finding of these studies is the conservation of the
relationship between MT and ID. In the case of a pointing movement, ID is related to properties
of the stimuli that influence the parameters of the actual movement. Therefore, the presence of
this law in imagined movements suggests that the motor system is likely engaged. Although it
remains largely unknown why imagined MTs tend to be greater than actual MTs, some
researchers have suggested that it may be due to the cognitive effort required to consciously
simulate an action (Young, Pratt & Chau, 2009). Overall, both neurophysiological and
behavioural work supports the neural simulation of movement in action imagination, implying
that execution and imagination share common neural substrates.
1.1.3 Simulation as a Mechanism for Action Possibility Judgements
Action possibility judgements involve the observation and perception of a given action,
followed by a prediction or judgement about the feasibility of the action. It is thought that to
8
accurately predict the consequences of an action, one must first neurally simulate this action,
effectively mapping the observed movement onto one’s own motor system (van der Wel, Sebanz
& Knoblich, 2013). This mechanism is assumed to be fundamentally similar to that of action
imagination, although the processes differ in that simulation in perception is often non-conscious
(Jeannerod, 2001). Again, much of the evidence for simulation in action perception lies in the
activation of cortical motor networks while individuals observe movements or are asked to make
predictions or judgements based on observed movements. First, early research using single-cell
recordings in monkeys revealed that there is activation of the premotor cortex during the simple
observation of goal-directed actions (Gallese, Fadiga, Fogassi, & Rizzolatti, 1996; Rizzolatti,
Fadiga, Gallese, & Fogassi, 1996). This result was later corroborated by neurophysiological
studies of humans, suggesting that a network of motor areas of the brain is active in the
perception of action, in particular areas of the premotor cortex (Rizzolatti & Craighero, 2004). It
should be noted that this earlier work established the activation of neural motor networks in
simple action observation, without the need for a subsequent judgement on the observed action.
Furthermore, the neural networks described in the previous studies differ slightly from the
networks discussed in the studies to be presented. Although observation of an action is clearly an
important precursor to making any sort of action judgment, action prediction or judgment likely
relies more heavily on simulation.
For instance, Stadler and colleagues (2012) used TMS to temporarily lesion the dorsal
premotor cortex (PMd) while participants observed (and made a perceptual judgement about) an
actor executing tasks of everyday living (e.g., loading the dishwasher). These tasks were
occluded at varying time increments and participants were asked to make a decision about
whether the timing of the video was realistic or if it had been manipulated. Critically,
participants received TMS either early (at the initiation of action prediction) or late (during
ongoing action prediction). The authors found that the temporary lesion of PMd via TMS
affected the accuracy of action predictions only when it was delivered early in action
observation. Thus, while it is likely that PMd is implicated in the initial observation and
embodiment of action, it may not be implicated in the ongoing simulation of action for
prediction.
9
Consistent with the notion that action perception activates a motor simulation network,
Eskenazi and colleagues demonstrated in an fMRI study that a neural network involving the
SMA, primary motor cortex and the globus pallidus (GP) was modulated by the difficulty of an
observed movement. In this experiment, participants were shown alternating images of a finger
pointing at rectangles of various widths. The alternating of the still photos created the apparent
motion of the finger moving between the two locations at a set movement time. Critically, the
target widths and distances were scaled to Fitts’ law and represented movements with IDs of 2, 3
and 4. In 20% of the trials, participants made a judgement about the speed of the movement. It
was found that activation of a motor simulation network was not only apparent, but it was also
modulated by the ID of the task. That is, movements of a higher difficulty were positively
correlated with higher activation of the areas described. Furthermore, these results are in line
with the findings of patient studies that have demonstrated deficits in the perceptual judgements
of action in patients with stable motor lesions (Eskenazi, Grosjean, Humphreys, & Knoblich,
2009; Serino et al., 2010). The results of these neurophysiological and neuropsychological
studies are compelling evidence that action perception goes beyond the simple recognition or
observation of action and implicates motor simulation as a mechanism for action possibility
judgements.
Behaviourally, evidence of simulation in action possibility judgements is similar to that
for action imagination. For example, the laws that describe the actual execution of movement are
also seen in the perception of action, such as the two-thirds power law and Fitts’ law (described
above) (Knoblich, 2008). The two-thirds power law describes the relationship between the
curvature of a movement trajectory and angular movement velocity. Essentially, the velocity at
which a movement is executed increases with the degree of curvature of that movement
(Lacquaniti, Terzuolo, & Viviani, 1983). Accordingly, this law has been shown to govern the
predictions we make about observed actions such that accuracy in predicting the outcome of a
movement is better when the observed movement is consistent with the two-thirds power law
and therefore, follows the constraints of actual human movement (Flach, Knoblich, & Prinz,
2004). Similarly, Grosjean and colleagues demonstrated that when participants were asked to
make a judgement about the possibility of observed movements between two targets, these
judgements were constrained by the speed-accuracy trade-off described by Fitts’ law (Grosjean
10
et al., 2007). This result has been replicated a number of times and has been shown to be
modulated by current body state, experience with the task and the perceived abilities of actors
completing the task (Chandrasekharan et al., 2012; Manson et al., 2014; Welsh et al., 2013). This
behavioural evidence again suggests that action perception and judgement utilizes a mechanism
that involves the motor system itself, lending support to the tenet of common representation
described in common coding theory.
Critically, recent research has also demonstrated the interrelated nature of action
perception, imagination and execution. For instance, Chandrasekharan and colleagues (2012)
demonstrated that action possibility judgements are subject to recent task experience. Using an
experimental paradigm similar to that of Grosjean et al. (2007), the researchers demonstrated that
MTs selected as possible were closer to the participants’ actual execution MTs after physically
experiencing the task. This result suggests that action execution can help to refine internal
simulations of action and in turn lead to action possibility judgements that are more closely
related to actual movement capabilities. Furthermore, Wong et al. (2013) expanded on this idea
by investigating the effect of experience on both action perception and imagination. The critical
finding of the Wong et al. (2013) study was that after actual execution experience, participants’
imagined MTs were closer to actual MTs. Although the researchers found no statistically
significant effect of experience on action perception (as in Chandrasekharan et al., 2012), this
lack of significance was likely due to the varying pattern of effects (i.e. some participants’ MTs
increased while others’ decreased, both towards actual execution MTs). Overall, it is clear that
execution experience plays a role in the refinement of internal representations (simulations) of
action. These results highlight the relationship between these processes and support the notion
that action perception, imagination and execution share a common representational domain, as
common coding theory would suggest.
In summary, the neurophysiological, neuropsychological and behavioural experiments
discussed here suggest that the execution, imagination and perception of action share common
neural substrates. The results of neurophysiological/psychological experiments demonstrate that
there is activation of sensorimotor brain areas during the imagination and observation of action
and particularly when one is asked to make a perceptual decision about an observed action.
Behavioural evidence has demonstrated that the same laws and consistencies observed in actual
11
movement execution are conserved in imagined movements and in action perception. Taken
together, these results highlight the interrelated nature of these three processes but also clearly
demonstrate that action imagination and perception involve the internal simulation of the neural
mechanisms for actual action execution. Therefore, neural simulation of action in imagination
and perception is an important mechanism for the shared representation of action and perceptual
effects, as outlined in common coding theory.
1.2 Development
1.2.1 Typical Motor Development
There are many different biological and psychosocial elements that contribute to the
typical development of motor skills. Biological factors include anthropometric changes (such as
height and body composition) and changes in the physiological systems (in vital capacity and
neuronal growth, for example). These biological changes also interplay with changes in a child’s
physical and social environment, including changes in the type of play that they engage in and/or
have access to (Sugden & Wade, 2013). Although these factors vary on an individual level, there
is a wealth of research that provides a general picture of how motor skills are acquired in
typically developing (TD) children.
The rapid growth and development of the previously mentioned factors in early
childhood means that, by the age of seven, TD children possess an array of fundamental motor
skills such as running and throwing from which to develop more complex skills. Critically,
children around the age of seven begin to interact with the environment in new, more complex
ways, which facilitates development in their spatial and temporal accuracy as well as the
continued improvement in performance of motor skills as a whole (Sugden & Wade,
2013).These improvements are evident in the increased ability of ten to twelve year olds to use
visual information when intercepting an object (e.g., when catching a ball), compared to younger
children aged five to seven years (Chohan, Verheul, Van Kampen, Wind, & Savelsbergh, 2008)
and also in the increased ability of older children to synchronize their movements with that of an
object (e.g., when deciding to cross the road) (Chihak et al., 2010). It is thought that these
differences between younger and older children represent developmental differences in action-
perception coupling (Chohan et al., 2008; Plumert, Kearney, & Cremer, 2007; Plumert, Kearney,
12
Cremer, Recker, & Strutt, 2011). Additionally, decreases in the rate of information processing in
this time period contribute to decreased reaction time and increased performance in manual skills
(Sugden & Wade, 2013). Critically, these performance measures point towards an increase in
motor skill performance between younger (late childhood) and older (pre-adolescent) children.
At the onset of adolescence, performance measures are more similar to those of adults and
continued change in motor skill development occurs at a slower rate.
Similarly, neurophysiological research has demonstrated that maturation of the cortical
spinal tract occurs at comparable time points. First, Koh and Eyre (1988) used electromagnetic
stimulation to establish age-related differences in the conductivity and integrity of the
corticospinal tract (CST). They found that conduction velocity of the CST increased with age
until about the age of eleven, where adult values were obtained. Studies employing TMS have
yielded similar results, suggesting maturation of the CST occurs just prior to or at the onset of
early adolescence (Fietzek et al., 2000; Nezu et al., 1997). Most recently, research utilizing
diffuse tensor imaging (DTI) has allowed researchers to estimate the integrity of white matter in
the CST as a function of age. Specifically, Yeo, Jang and Son (2014) noted that fibre volume in
the CST increased rapidly until the age of twelve, after which the rate of change was much
slower. The results of these studies clearly supports the results of performance measures,
demonstrating that there is continued rapid motor skill development in late childhood until the
onset of early adolescence (at approximately twelve years of age) where this development slows.
1.2.2 Motor Imagery Development
In recent years, researchers have also wanted to understand how the imagination of
movement develops over the lifespan. This question is of particular interest to researchers as MI
is thought to be a window into the cognitive representation of action (Gabbard, 2012). An
understanding of how these representations are formed in TD children may shed light on the
underlying mechanisms of neurodevelopmental disorders whose origins may be linked to action
representation (developmental coordination disorder (DCD), for example). Accordingly, research
in this area has studied MI in a variety of ways. These paradigms include mental chronometry
and measures of MI vividness as well as more implicit paradigms utilizing the mental rotation of
13
bodies and objects (Gabbard, 2009). This review, however, will focus on the use of explicit MI
and mental chronometry paradigms, specifically.
Similar to the study of MI in adults, mental chronometry paradigms have been used
extensively in the investigation of MI in children. Earlier studies involved simple comparisons of
the time necessary to complete an action and the time the child took to imagine the action.
Molina, Tijus and Joen (2008) demonstrated that a group of seven year old children were much
more accurate in imagining the amount of time necessary to move a doll from one location to
another than a group of five year old children. Furthermore, Hoyek and colleagues (2009) had a
group of elementary school children (mean age of 7.8 years) and a group of middle school
children (mean age of 11.4 years) both execute and imagine running through a set obstacle
course. The researchers found that congruence between real and imagined times was lowest in
the elementary school children. Similar results were found by Sachet and colleagues (2015), who
demonstrated that although there was a relationship between real and imagined movement times
in six year old children, there was a higher congruence between these measures in eight year old
children. These results demonstrate that a temporal congruence between real and imagined
movements emerges around approximately the age of seven and continues to develop through
late childhood.
To build on these results, researchers have gone beyond simple congruence between real
and imagined movement times and have demonstrated the presence of Fitts’ law, as described
previously, in the imagined movements of TD children. Interestingly, while it is clear that this
law emerges in the imagined movements of children, it has been demonstrated that the
relationship between speed and accuracy emerges to a lesser degree in younger children (six to
seven years of age). Additionally, there appear to be greater differences in the slopes of the
speed-accuracy trade-off between imagined and executed movements in younger children than in
older children (Caeyenberghs, Tsoupas, Wilson, & Smits-Engelsman, 2009; Caeyenberghs,
Wilson, van Roon, Swinnen, & Smits-Engelsman, 2009). Furthermore, although Fitts’ speed-
accuracy trade-off and the congruence between real and imagined movements appears relatively
stable in adolescence (Smits-Engelsman & Wilson, 2013), there may be some additional
development in this respect from adolescence to adulthood (Choudhury, Charman, Bird, &
14
Blakemore, 2007a, 2007b). However, these developmental differences seem to be more subtle
than those seen from early to late childhood.
1.2.3 Representing the Actions of Others
In contrast to the motor imagery literature, much less is known about the developmental
use of cognitive action representation in action perception and action possibility judgements. In
fact, much of the available research has focused on motor representation and imitation in infancy
(Sommerville & Decety, 2006). Although it has been demonstrated that motor simulation plays a
central role in imitation and observational learning in infancy (Paulus, Hunnius, Vissers, &
Bekkering, 2011a; 2011b), it remains largely unknown how use of this mechanism changes as a
child and their own motor repertoire develops and how it may be used to make decisions about
the possibility of observed actions as well as other motor cognitive processes.
There is, however, some evidence to suggest that young children perceive the movements
of others through a process of neural simulation of action. Two relatively recent experiments
have used motor contagion paradigms to investigate how young children represent the actions of
others. Motor contagion paradigms are constructed to investigate how an individual’s
movements change as a result of simultaneously observing the movement of another that is
incongruent with their own movement (e.g., Kilner, Paulignan, & Blakemore, 2003). Typically,
participants execute either horizontal or vertical arm movements while observing the
directionally mismatched or incongruent movement of an actor. It has been commonly observed
that participants’ movements are more variable in the direction of the observed movement. For
example, Kilner and colleagues (2003) had subjects observe a human executing vertical
movements or horizontal movements while they executed movements that were either in the
same or opposite direction. The researchers found the subjects’ movement trajectories were more
variable in the direction of the observed movement in incongruent trials compared to congruent
trials. This effect was not found when subjects observed the movements of a robot. Such findings
have been taken as evidence that participants are representing the observed movement using
some component of their motor system. Specifically, it is thought that when observing the action
of another, this action is mirrored in one’s own motor system (via a simulation process) which
then interferes with the current movement being executed (Blakemore & Frith, 2005).
15
In one study of children, Marshall and colleagues (Marshall, Bouquet, Thomas, &
Shipley, 2010) clearly demonstrated a motor contagion effect when a group of four year olds
moved a stylus on a tablet in one direction while observing a video of an individual moving in
the opposite direction. In a follow up study using a similar paradigm, it was shown that this
effect is subject to experience and observation of the observed actor’s motor capabilities (Saby,
Marshall, Smythe, Bouquet, & Comalli, 2011). The results of these studies demonstrate that
children represent the actions of others in a way that involves the motor system. In other words,
it is clear that neural simulation of action is intact at this age. However, it remains unclear how
these simulations change over time, as a function of age.
1.3 Summary
The literature presented here provides a comprehensive overview of common coding
theory and the neural simulation of action in action imagination and perception. Previous work
has demonstrated the activation of overlapping cortical motor networks in action imagination and
perception. This conclusion is supported by patient studies and behavioural work demonstrating
the presence of the laws that describe actual execution of movement in both the imagination and
perception of movement. Overall, this body of work implicates the simulation of the neural
components of movement in the processes of motor imagery, perception, and cognition.
Developmentally, it has been demonstrated that while there is rapid change in motor skill
acquisition and development in infancy and toddlerhood, the ongoing motor development in late
childhood seems to slow just prior to early adolescence. This developmental trajectory is
mirrored by the development of motor imagery - an ability which is thought to be formed by the
age of seven, although there is continued development into early adolescence. Critically, while
there is evidence of motor simulation in action observation in infants and children, a
developmental trajectory of how these simulations affect motor cognition and, specifically,
action possibility judgements has yet to be unearthed. Additionally, it appears that motor
imagery has largely been studied in isolation from other motor cognitive processes, in this
respect. Therefore, there are two meaningful gaps in the literature: in the ongoing development
of action perception and in how the development of explicit motor simulation in motor imagery
is related to development of other forms of motor cognition.
16
1.4 The Current Project: Purpose and Specific Hypotheses
There were two main interrelated purposes of the project reported in the present thesis.
The first purpose was to assess the changes in the temporal accuracy of action possibility
judgements that occur from late childhood to early adolescence. The second was to evaluate if
this developmental trajectory is similar to or differs from that of motor imagery. The selected
experimental paradigm employed mental chronometry much like the experiments of Wong et al.
(2013) and Yoxon et al. (2015) as well as action possibility judgements similar to that of
Grosjean and colleagues (2007). Accordingly, the study involved children from ages seven to
sixteen completing three main tasks - an action execution task, an action imagination task, and an
action possibility judgement task. Each of the tasks involved continuous aiming movements
(executed, perceived or imagined) between two targets of varying width and amplitude. In the
execution task, the children executed continuous movements between two targets of varying
width and amplitude. In the imagination task, the children imagined themselves performing these
same movements. Finally, in the action possibility judgement task, the children were asked to
judge whether or not the observed movement was possible to complete at varying speeds
between the same six sets of targets. Based on current knowledge in the area, several overall
predictions and some specific predictions were formed.
First, it was predicted that executed and imagined movements as well as the action
possibility judgements would conform to the speed-accuracy trade-off characterized by Fitts’ law
(described in section 1.1.2; Fitts, 1954). This overall prediction was formed because previous
research has demonstrated that neural simulation of action should be developed by the age of
seven (e.g., Gabbard, 2009; Molina et al., 2008). Additionally, it is predicted that imagined MTs
will likely be greater than actual execution MTs, as demonstrated in previous work (Wong et al.,
2013; Young et al., 2009; Yoxon et al., 2015). Second, it was predicted that, consistent with
previous literature (e.g., Caeyenberghs, Wilson, et al., 2009; Smits-Engelsman & Wilson, 2013),
imagined movement times should be more similar to actual movement times as age increases.
Therefore, MTs should decrease, closer to actual MTs with age. Furthermore, because it is
proposed that action imagination and perception are represented under a common representation
domain and are thought to share similar mechanisms (neural simulation of action), selected MTs
in action possibility judgements should also approach actual movement times with age and this
17
measure should develop in a similar fashion to motor imagery. If, however, action imagination
and perception do not share a common representational domain, or vary in their neural
mechanisms, then the measures of motor imagery and action perception ability may develop
independently of each other. That is, MTs in the two tasks may approach actual MTs
independently of each other.
18
Chapter 2 The Independent Development of Action Imagination and
Perception
2.1 Abstract
According to neural simulation theory (Jeannerod, 2001), the neural simulation of actions
serves as a unifying mechanism for the processes of motor cognition. Support for this notion
comes from neurophysiological data indicating motor system activation in these processes as
well as recent behavioural data demonstrating the presence of Fitts’ law in imagined actions and
action possibility judgements. This speed-accuracy trade-off has also been demonstrated in the
imagined movements of children. Interestingly, trade-offs observed in imagined movements have
been shown to be more similar to actual execution movement time (MT) as age increases. The
relationship between execution, imagination and perception movement times, however, has yet
to be evaluated in children. The purpose of the current experiment was to assess how imagined
and perceived MTs relate to actual MTs across a range of ages from seven to sixteen years. It
was hypothesized that these processes should develop in a similar fashion, due to their shared
mechanism. It was found that imagined MTs were longer than execution MTs, and that this
difference remained stable across the age ranges. Action possibility judgements, on the other
hand, were lower than execution times for younger children but became more consistent with
execution MTs as age increased. These results reflect potential mechanistic differences in the use
of action simulation between action imagination and perception.
19
2.2 Introduction
When pilots are first learning to fly, they train for the various situations they will
encounter during real flying using flight simulators: machines that simulate all of the internal
components of the aircraft, but where the actions of the pilot in training do not actually produce
any real external outcome (i.e., flying). Similarly, it is thought that the human brain can also
simulate the internal components of actions without actually producing any external movements.
For instance, many athletes use visualization or mental practice as a training tool to mentally take
themselves through a routine or a game without the need for actual physical practice. It is
thought that such simulation and mental practice is an effective training tool because it has been
demonstrated that when an individual imagines themselves performing an action, they activate a
neural network that is similar to the one that is involved in actual movement execution (Hétu et
al., 2013; Jeannerod, 2001; Munzert et al., 2009). In this sense, it is thought that the brain is
effectively simulating the internal components of movement by activating the neural codes that
actually generate action offline without leading to overt movement.
This concept of action code based simulation has been extended beyond the explicit
imagination of action and has been implicated as a more implicit mechanism that unifies the
processes that help us perceive and understand our own actions and the actions of others
(Jeannerod, 2001). In action observation and possibility judgements, for example, it is thought
that the neural simulation of action allows an individual to map an observed movement onto their
own motor system and capabilities, allowing the individual to make accurate judgements about a
given movement (Chandrasekharan et al., 2012; Grosjean et al., 2007; Welsh et al., 2013).
Furthermore, because of the important role that motor simulations play in perceiving and
understanding actions, there has been growing research in how simulation processes emerge in
childhood. Importantly, the mirroring of others’ actions via motor simulations is thought to play
an important role in how young children learn to link gestures and movements with their
meanings and effects (Paulus, Hunnius, & Bekkering, 2013; Paulus et al., 2011a, 2011b).
20
2.2.1 Neural simulation of action
Evidence for neural motor simulation in action imagination and perception has been
broadly drawn from both neurophysiological and behavioural studies that have reported that the
motor system is active in and constrains action imagination and perception. First, neuroimaging
studies have clearly demonstrated that when participants imagine themselves performing
movements, there is activation of a neural network that overlaps with that of actual execution
(e.g., Stippich, Ochmann, & Sartor, 2002; see Hétu et al., 2013 for a meta-analysis and
comprehensive review). Additionally, many studies have employed transcranial magnetic
stimulation (TMS) to study changes in the excitability of the corticospinal tract while individuals
imagine themselves performing actions. These studies find that when individuals imagine
themselves performing an action, there is a quantifiable increase in excitability of the
corticospinal tract, which suggests that the motor system is active in the imagination of action
(e.g., Clark et al., 2004; see Munzert, et al., 2009 for review). Similar neurophysiological results
have been found for action perception, wherein previous research has demonstrated that various
aspects of the motor system are active when an individual is making judgements about a
movement (Eskenazi et al., 2012). Importantly, it has also been demonstrated that there is
considerable overlap in the cortical areas shown to be active in action imagination, perception
and execution, suggesting that these processes are in fact part of a singular representational
domain (Grèzes & Decety, 2001).
Behaviourally, evidence for common neural substrates underlying action execution,
imagination and perception (and therefore motor simulation) has been drawn from
demonstrations of the temporal similarities between these three processes. Specifically, many
previous studies have demonstrated the presence of Fitts’ law in imagined and perceived
movement times (MTs). Fitts’ law is a mathematical equation that describes the relationship
between MTs and the difficulty of manual aiming movements. It can be described by the formal
equation: MT = a + b(ID), where a and b are constants relating to an individual’s base MT and
their unit increase in MT as a function of the index of difficulty(ID), respectively. The ID
component of this equation can be further broken down as ID = log2(2A/W), where A is the
movement amplitude between a pair of targets (centre-to-centre distance between the targets) and
W is the width of the targets. Essentially, in a task that requires an individual to touch back and
21
forth between a target pair, the MT will increase as a function of both movement amplitude and
target width (the index of movement difficulty).
Decety and Jeannerod (1995) first demonstrated this relationship between execution and
imagination in a study of the walking paths of their participants. The researchers had individuals
imagine walking towards openings that varied in their distance to the participant and in the width
of the opening. Critically, participants demonstrated Fitts’ law in both their actual and imagined
walking paths. Specifically, participants’ walking paths were slower when the door opening was
narrower and the walking path was longer, as was mirrored in the imagined movement time.
Since this original experiment, this result has been corroborated and extended to manual
movements similar to Fitts’ original aiming task (Sirigu et al., 1995; Wong et al., 2013; Young et
al., 2009; Yoxon et al., 2015).
This principle of speed-accuracy trade-offs has also been demonstrated in action
possibility judgements. In these studies, individuals are asked to decide on the possibility of an
aiming movement being executed accurately when it is shown at a given speed. It has been
repeatedly shown that participants choose MTs that are consistent with Fitts’ law
(Chandrasekharan et al., 2012; Grosjean et al., 2007; Welsh et al., 2013; Wong et al., 2013).
Together, the results of these studies demonstrate that there is congruency in the temporal
aspects of executed, imagined, and perceived action which suggests again that these processes
share a common representational network and involve similar neural mechanisms.
2.2.2 Action simulation in children
In recent years, there has also been interest in how action simulation emerges in
childhood. Developmental motor imagery research is similar to research that has been done in
adults in that it has focused on the increasing congruence of actual and imagined MTs across
childhood development (Gabbard, 2009). Research suggests that simulation processes are likely
intact in early childhood, evidenced by the presence of Fitts’ law in the imagined movements of
young children aged approximately seven years, although this relationship is more evident in
older children (Caeyenberghs, Tsoupas, et al., 2009; Caeyenberghs, Wilson, et al., 2009). It is
also clear that imagined MTs become closer to actual execution MTs as a child ages, becoming
closer to the congruence between these measures seen in adults in adolescence, suggesting that
22
there is ongoing refinement of action simulation processes in childhood (approximately six to
eleven years)(Caeyenberghs, Wilson, et al., 2009; Smits-Engelsman & Wilson, 2013). Some
studies have also demonstrated developmental differences in the temporal consistency of action
imagination between adolescents and adults (Choudhury et al., 2007a, 2007b) although studies
with larger age spans of six to nineteen years (Caeyenberghs, Wilson, et al., 2009; Smits-
Engelsman & Wilson, 2013) suggest that these changes are more subtle compared to the changes
seen from early to late childhood and just prior to adolescence.
Interestingly, while developmental aspects of motor imagery have been relatively well
characterized, there has been little focus in developmental changes in the perception of human
movement. There is some evidence that the motor system is likely engaged in the imitative
processes that are central to motor learning in infancy and toddlerhood (Paulus et al., 2011b) as
well as evidence of action co-representation (the representation of an observed action in one’s
own motor system) in early childhood (Marshall et al., 2010; Saby et al., 2011). Although the
results of this previous research again demonstrate that motor simulation processes are likely
intact at a young age, it remains largely unknown how perceptions of action change as a child’s
own motor repertoire changes.
Therefore, the purpose of the current experiment was to evaluate changes in the temporal
accuracy of action possibility judgements as a function of age and to assess differences or
similarities in the developmental trajectories of action possibility judgements and motor imagery.
To address this purpose, an experimental paradigm employing action possibility judgements
similar to that of Grosjean, Shiffrar and Knoblich (2007) and a mental chronometry paradigm
similar to those of Wong et al. (2013) and Yoxon et al. (2015) was used. Children between the
ages of seven and sixteen, as well as a control group of adults executed aiming movements to
target pairs with varying accuracy demands. Imagination and perception tasks employed the
same target pairs. It was broadly hypothesized that MTs in action perception, imagination and
execution would conform to Fitts’ law, as action simulation processes should be initially
developed before the age of seven. Of greater theoretical relevance were predictions that
involved comparisons across the different tasks. If action perception and imagination share a
common representational domain and action simulation is an underlying mechanism for both
action possibility judgements and imagination, then these processes should develop in similar
23
ways. Specifically, MTs in both action imagination and action possibility judgements should
approach actual execution MTs with age. If this is not the case, it is possible that these processes
may have different underlying mechanisms that develop independently of each other.
2.3 Methods
2.3.1 Participants
Thirty-four children between the ages of seven and sixteen (24 Male, 10 Female) and 11
adults (2 Male, 9 Female, Mean Age = 22.4) were recruited for the study. Two male participants
were removed because it was disclosed to the experimenter subsequent to testing that they were
diagnosed with unspecified developmental and learning disabilities that may have impacted
motor skills and comprehension of task instructions. All other participants were reported to be
typically developing and had normal or corrected to normal vision. All participants, with the
exception of two children, were right handed. One of these participants was left-handed and the
other reported no distinct preference for activities of daily living. Handedness was collected
using a modified version of the Edinburgh Handedness Questionnaire1. All participants provided
informed assent and their parents or guardians provided informed consent prior to testing. All
procedures were approved by the University of Toronto Research Ethics Board.
2.3.2 Study Design and Tasks
The design of the present study was based on previous studies (Chandrasekharan et al.,
2012; Grosjean et al., 2007; Wong et al., 2013). Participants all completed action execution,
imagination and perception tasks. The execution task was always performed first, followed by
the remaining two tasks. Previous research has demonstrated the effect of experience on
perceived and imagined MTs (Chandrasekharan et al., 2012; Wong et al., 2013; Yoxon et al.,
2015). Specifically, it has been found that perceived and imagined MTs more accurately reflect
actual execution MTs after task-specific experience. This increased consistency may reflect
experience-based refinement of internal action simulations and occurs because participants are
able to link the perceptual effects of an action with the actual motor experience during practice.
1 See Appendix A
24
These more closely linked representations of action and effect can then facilitate imaginations
that are more temporally similar to actual execution MTs. Because of the observed benefit of
experience on perception and imagination, the execution task was performed first in order to
avoid any variance in the data due to differences in task experience. Essentially, the execution
task was performed prior to the other two tasks in order to equate the groups on task-specific
experience. Any developmental differences observed can therefore be more closely equated to
differences in age and not differences in previous experience with similar tasks. Additionally, the
imagination and perception tasks were counterbalanced to ensure there were no effects of
execution experience just prior to completing either of these tasks. To confirm that this was not
the case, the effect of order was assessed by carrying out a 2 (task: imagination, perception) by 2
(order: imagination first, perception first) mixed ANOVA with repeated measures on the task
variable on imagined and perceived MTs. Although there was a main effect of task, F(1,41) =
96.097, p < .01, there was no significant effect of order, F(1,41) = .048, p = .827, or task by
order interaction, F(1,41) = .551, p = .462. These results indicate that there was no significant
effect of task order in relation to execution on imagined or perceived movement times.
Participants were tested individually, under the direct supervision of the experimenter. In
each of the tasks, data collection only proceeded when the experimenter had confirmed that the
participant understood the task. The overall time in testing was between 30 and 45 minutes.
All tasks were performed using a touch screen monitor (3M™ MicroTouch™ Display,
473.8mm (W) x 296.1 mm (H)). In all of the tasks, six sets of two targets varying in target width
and movement amplitude were used. The targets were one of two widths: 2.5 cm or 3.5 cm. The
centre-to-centre measurement (movement amplitude) for a given movement context was one of
7.5 cm, 15 cm or 30 cm for the 2.5 cm target. For the 3.5 cm target, the centre-to-centre
measurement was one of 10.5 cm, 21 cm or 42 cm. These combinations generated two target
pairs for each of the IDs: 2.6, 3.6 and 4.6. The combination of target width and movement
amplitude remained consistent throughout a specific trial.
2.3.2.1 Execution Task
Participants were seated comfortably in front of a table upon which the touch screen
monitor rested. In a given trial, they were presented with one of the six target pairs. Beginning
25
with the index finger of their dominant hand on the right side target, participants were asked to
perform ten continuous pointing movements as quickly and accurately as possible between the
two targets. One movement was from the right to the left target and the next from the left to the
right target. They were told that they must move as quickly as possible but that they must also try
to always land “on the line” (i.e., within the target). This sequence was repeated three
consecutive times for each target pair for a total of 30 movements per target condition. Prior to
the experimental trials, the experimenter gave the task instructions and participants experienced
three practice trials, during which they had the opportunity to ask the experimenter questions
about the task and the experimenter could confirm their understanding of the task demands.
The order of the target combinations was randomized. The accuracy (spatial coordinates
of screen contact) and the time to complete the movements were recorded, by the custom
program which also displayed the stimuli for analysis offline. A single mean movement time was
calculated for each of the six combinations of target width and movement amplitude. Erroneous
trials where there was clearly either computer or human error (a touch recorded on the same side
of space two times in a row or where the touch screen failed to record a touch) were removed
prior to calculation of the mean movement time for each target pair. Additionally, touches that
fell beyond 35 px (approximately 1cm) of the target were considered errors and were also
removed at this point. These procedures resulted in the removal of approximately 4.3%
(erroneous trials) and 0.7% (error) of the individual touches in the execution data overall.
2.3.2.2 Imagination Task
The experimental set up was consistent with that of the execution task. In a given trial,
participants were presented with one of the six target pairs. They began with the index finger of
their dominant hand on the right side target and imagined executing ten pointing movements
between the targets as quickly and accurately as they could execute the movements in real time.
Similar to the execution task, participants were asked to imagine their finger moving as quickly
as it did in “real life” and to imagine themselves always landing “on the line” (i.e. on the target).
Participants were asked to perform the imagination from an internal (first person) perspective.
They were instructed to lift their finger off the monitor (a maximum of about 5 cm) for the time
required to imagine the movement and then place their finger back onto the monitor after
26
imagining the movements. They were also instructed that the finger lift should occur when they
imagine their finger lifting off for the first time and the placing of the finger back onto the
monitor should occur when they imagine their finger returning to the right side target on the
tenth movement. This sequence was repeated three times for each target pair for a total of 30
imagined movements.
Prior to experimental trials, participants experienced three practice trials during which
they had the opportunity to ask the experimenter questions about the task and the experimenter
could confirm the participant’s understanding. The order of the target pairs was randomized. The
total time required to complete the imagination (from finger lift to contact with the touch screen)
was recorded for analysis. A mean movement time was calculated for each of the target pairs and
this mean time was derived by dividing the total imagination time by ten to provide a mean
movement time for each target pair in a given trial. Trials where there was either computer or
human error (where the touch screen failed to register a touch or the participant admitted to
improperly performing the task) were flagged throughout testing and were removed prior to
analysis. This resulted in the removal of approximately 2% of the imagination data.
2.3.2.3 Perception Task
Participants were seated as in the other two tasks. In the perception task, the touch screen
monitor presented two digital photographs of a young adult woman performing the execution
task from an internal or first person perspective: the first picture was of a person with their right
index finger on the right side target and the second picture was of the finger on the left side
target (see Figure 1). These photos were presented alternately to create the apparent motion of
the model in the photographs moving between the two targets. In a given trial, the photos were
presented at one of eleven different stimulus onset asynchronies (SOAs) ranging between 120 ms
and 520 ms. The SOA remained constant within a trial and the trial ended when the participant’s
response was recorded. Participants were asked to judge if it is possible for them to maintain
accuracy while moving at the shown speed. Specifically, they were told that the task was to
decide if it was possible or impossible to move as fast as the hand is moving and still be able to
land “on the line” (i.e., to land accurately on the targets). They were told to pick the best answer
(possible or impossible) for them. Participants would verbally tell the experimenter, yes or no, if
27
they thought the movement was possible or impossible to move at the given movement time,
respectively.
Figure 2.1. An example of the pictures that were displayed in the perception task. Images 1 (hand
on the right side target) and 2 (hand on the left side target) were alternated at a range of SOAs to
create the apparent motion of the hand.
Prior to testing, participants were shown 3 practice trials, two of which represented the
extremes of the SOA/target pair combinations (a high difficulty with the fastest possible
movement and a low difficulty with the slowest possible movement) and the other was a target
pair with ID = 2.6 at SOA = 200 in order to represent a trial that was at neither of the previously
presented extremes. During these practice trials, the experimenter asked the participants
questions to confirm their understanding. For example, the experimenter would ask the
participant why they thought a particular movement was possible or impossible. During this time
the experimenter answered any questions the participants had. After confirming their
understanding of the task, participants completed one block of perceptual judgments consisting
of 66 trials (6 target combinations x 11 SOAs). For each of the 6 target pairs, The point at which
participants changed their responses from impossible to possible (this point was the point along
the spectrum of SOAs where the participant answered “yes” twice in a row) was determined and
the SOA at this point was considered to be the minimum MT perceived to be possible for a given
combination of target width and amplitude (or ID). This process generated one data point for
each of the 6 target pairs. A similar type of task has been used several times in adult studies to
28
assess how individuals perceive an observed action (Chandrasekharan et al., 2012; Eskenazi et
al., 2012; Grosjean et al., 2007; Welsh et al., 2013). Therefore, the current method is consistent
with previous work, but was modified for use with children. Specifically, a single block of trials
(as opposed to multiple blocks of trials) was performed to maintain a relatively short time in
testing, in order to balance task demands and fatigue, particularly in younger children. Although
keeping the time in testing shorter has the potential of decreasing the effects of boredom and
fatigue, it does raise challenges to the reliability of the data given that a single observation is
recorded for each trial type. This approach was deemed appropriate given the overall aims of the
work and that this study was the first to examine action perception in children and adolescents in
this manner. Future work will need to expand on and refine this initial research.
2.4 Results2
Prior to statistical analysis, child participants were initially divided into two experimental
groups to facilitate analysis of differences between younger (seven to eleven, n = 18) and older
(12 to 16, n = 14) children. This division was chosen as this age has been shown as a critical
point for corticospinal maturation (Yeo et al., 2014), motor skill acquisition (Sugden & Wade,
2013) as well as motor imagery development (Caeyenberghs, Wilson, et al., 2009; Smits-
Engelsman & Wilson, 2013). Specifically, this age seems to delineate all three of these processes
in that there is more rapid development up until approximately eleven or twelve years, followed
by more subtle developmental changes in corticospinal maturation, motor skill acquisition and
motor imagery ability from this age onward.
2.4.1 Fitts’ Law
To determine if movement times in each group and task conformed to Fitts’ speed
accuracy trade-off, a linear regression was calculated between group mean MT for each of the
six combinations of target width and amplitude and ID. For all groups and all tasks, MT was
significantly correlated with ID, confirming the presence of Fitts’ law in all groups and tasks
(Figure 2.2). For equations and statistics, see Table 2.1.
2 For additional analyses and results not included in the to-be-submitted paper, but might be of interest for the thesis
see Appendix B.
29
Figure 2.2 Linear regressions between index of difficulty (ID) and movement time for each of
the three tasks, for each of the three experimental groups.
30
Table 2.1. Fitts’ law equations and statistical analysis for the linear regressions calculated
between MT and ID for each of the tasks and groups.
Children Aged 7-11 MT = R2 p
Execution 116.1 + 81.13(ID) .98 .0002
Imagination 324.0 + 59.99(ID) .95 .0009
Perception 25.78 + 75.56(ID) .99 <.0001
Children Aged 12-16
Execution 26.18 + 85.54(ID) .97 .0004
Imagination 207.6 + 66.30(ID) .97 .0004
Perception 53.14 + 102.9(ID) .99 <.0001
Adults
Execution 44.98 + 78.95(ID) .99 <.0001
Imagination 187.6 + 68.68(ID) .96 .0007
Perception -81.33 + 102.7 (ID) .98 <.0001
2.4.2 Group Differences
An analysis of variance was conducted to further examine within- and between-group
differences between execution, imagination and perception task MTs. Because MTs were found
to conform to Fitts’ speed-accuracy trade-off, a single mean MT was calculated per participant
for each task. These mean MTs were submitted to a 3 (task: execution, imagination, perception)
by 3 (group: younger, older, adult) mixed ANOVA with repeated measures on the first factor.
Mauchly’s test indicated that the assumption of sphericity had been violated (X2(2) = 12.42, p <
.01). Therefore, degrees of freedom were corrected using Greenhouse-Geisser estimates of
sphericity (ε = .79). There was a significant effect of task, F(1.57, 62.85) = 71.30, p < .001,
where imagined MTs were significantly higher than those for perception and execution, and
perception MTs were significantly lower than those for execution and imagination (Tukey’s
HSD, p < .05, CV = 53.10). There was also a significant effect of group, F(2,40) = 5.07, p < .05,
where MTs for the younger child group were significantly higher than those of the adult group
31
(Tukey’s HSD, p < .05, CV = 34.57). There was no significant difference in overall MT between
the two child groups, although there was a trend towards higher overall MTs for the younger
group. Finally, there was a significant group by task interaction, F(3.14, 62.85) = 3.73, p < .05.
Post-hoc analysis (Tukey’s HSD, p < .05, CV = 79.95) of the interaction showed that in the
youngest group, there were significant differences between the mean MTs of all tasks, with the
perception task being the lowest and the imagination task the highest (see Figure 3). In the older
child group and adult groups, the mean MT for imagination was significantly higher than the
mean MT for both perception and execution, but there were no significant differences between
perception and execution. Between the groups, there were no significant differences between
MTs for the execution or perception tasks, although the difference in execution times between
the younger child group and the other two groups approached significance. In the imagination
task the younger group had significantly higher MTs than the older group and the adult group.
2.4.3 The relationship between age and simulation congruency
To assess the relationship between congruency of action simulation processes (the degree
to which perception and imagination MTs reflect actual execution MTs) and age (i.e., how these
measures change with age), difference scores between mean imagination and execution and
mean perception and execution MTs were calculated for each participant. This analysis only
included the child participants. Pearson’s correlation coefficient and linear regressions were
calculated for the correlation between difference scores for imagination and age as well as the
correlation between difference scores for perception and age. It was found that difference scores
for perception were significantly and positively correlated with age (r = .71, p < .001, y = 20.16x
– 299.40) such that perception MTs approach actual MTs as age increased (Figure 4-A). In
contrast, there was no significant correlation between the difference scores for imagination and
age (r = .005, p = .98, y = 0.19x + 121.60), suggesting that these differences were more stable
across the age range (Figure 4-B).
32
Figure 2.3. Mean imagined MTs for each of the execution, perception and imagination tasks.
Asterisks indicate significant (Tukey’s HSD, p < .05, CV = 79.9) within group differences
between the tasks.
Figure 2.4. Difference scores between imagination and execution (A) and perception and
execution (B) as a function of age. Note: This analysis includes only child participants.
33
2.5 Discussion
The aim of the current study was to quantify the relationship between action execution,
imagination and perception in children and adolescents and to describe how these relationships
change as a function of age. Typically developing (TD) children between the ages of seven and
sixteen participated in three tasks: execution, imagination and perception (action possibility
judgement), which involved continuous aiming movements to the same six target pairs. Overall,
the movement times (MTs) for each of these tasks increased as a function of the index of
difficulty (ID) of the aiming movement and therefore conformed to Fitts’ law, regardless of the
age of the participants. The critical finding, however, was that MTs selected in the action
possibility judgements were more congruent with actual execution MTs as age increased,
whereas imagined movement times did not exhibit a similar developmental change. These results
are discussed over the following sections as they relate to neural simulation in action imagination
and perception.
2.5.1 Fitts’ law in imagination and perception
Overall, the presence of Fitts’ law in action perception and imagination across all age
groups demonstrates that action simulation processes are intact and at least partially developed
by the age of seven. This result is in line with the findings of previous motor imagery (MI)
research, where the consensus is that the ability to effectively engage in MI is present by
approximately seven years of age (Gabbard, 2009). This finding is also consistent with previous
action observation and imitation research in infancy which has implicated a motor simulation
mechanism for imitation and observational learning (Paulus et al., 2013, 2011a, 2011b). It is also
consistent with the work of Marshall et al. (2010) and Saby et al. (2011) who demonstrated that
children’s movement trajectories change when observing the incongruent movement trajectory of
another person (i.e., a motor contagion effect). In sum, the result that both imagined and
perceived MTs followed Fitts’ speed-accuracy trade-off is additional evidence for motor system
activation in the imagination and perception of movement in children.
34
2.5.2 The relationship between age and action imagination
In contrast to past work, the current experiment found no age-related differences in the
congruency between imagined and executed MTs. Specifically, although action imagination
times were longest in the youngest children, the difference between real and imagined MTs did
not change as a function of age. Instead, imagined MTs were consistently greater than actual
execution MTs. It should be noted that the overestimation of imagined movement times is not
uncommon for this type of aiming task (see Wong et al., 2013; Young et al., 2009; Yoxon et al.,
2015). Additionally, because the younger children’s actual execution MTs were numerically but
not significantly longer than those of the other groups, the small decrease in imagined MT from
the younger to the older group does not represent a developmental change in the temporal
congruency of action imagination. Moreover, the correlational analysis presented here
demonstrates that there is not a consistent change in the way imagined MTs approached actual
MTs as a function of age. Therefore, the ability to imagine the temporal aspect of imagined
movements (i.e. the congruency between real and imagined movements) does not seem to
change from later childhood into adolescence.
This result may be related to the nature of developmental changes in motor imagery in
late childhood. Previous studies (Caeyenberghs, Wilson, et al., 2009; Smits-Engelsman &
Wilson, 2013) included a large number of children in early childhood. It is this early age range
(five to eight years) that seem to have the largest discrepancies between real and imagined MT
(Caeyenberghs, Wilson, et al., 2009; Molina et al., 2008; Skoura, Vinter, & Papaxanthis, 2009;
Smits-Engelsman & Wilson, 2013). It is possible, therefore, that these differences in younger
children under the age of seven (where it is thought that motor imagery processes are largely
developed, see Gabbard, 2009; Molina et al., 2008) contributed to the age-related differences
seen in previous studies. Relatedly, a recent study that evaluated age-related imagery ability
using a self-report questionnaire developed for use with children found no age-related
differences in self-reported ease of imagery in children aged seven to twelve years (Martini,
Carter, Yoxon, Cumming, & Ste-Marie, in press). Therefore, it is likely that age-related changes
in the temporal congruency of motor imagery did not emerge in the current experiment because
developmental differences in action imagination are more subtle at the age range used in the
current study.
35
Recent task experience may have also played an important role in the consistency of
motor imagery ability. In the current study, all children executed the pointing task before
imagination and perception tasks. This experience, however, was not provided in past studies.
For example, the study of Caeyenberghs and colleagues (2009) included many tasks that were
counterbalanced across participants, meaning that execution experience may have occurred at
various times relative to the imagination task. Smits-Engelsman and Wilson (2013) only used
imagination and execution tasks but these were also counterbalanced so that half the participants
began with imagined movements. Recent work has demonstrated the effect of task experience on
the accuracy of imagination and perception of continuous aiming movements (Chandrasekharan
et al., 2012; Wong et al., 2013; Yoxon et al., 2015), suggesting that recent task experience may
lead to more accurate imagined and perceived MTs. Essentially, it is thought that experience
with a given task can generate a more accurate representation of an action and its associated
perceptual effects – leading to more accurate imagination and perception of movement.
Therefore, it is also possible that the very recent task experience in the current study may have
led to more accurate cognitive representations of action, particularly in the younger children who
may lack in general motor experience compared to older children. However, if this were the case,
similar effects would be expected in action perception, because this process would also be
affected by experience (as in Chandrasekharan et al., 2012). Therefore, although the lack of age-
related changes in action imagination may be related to recent task experience, it is more likely
that this effect is due to more subtle developmental differences in this task, from late childhood
to adolescence.
2.5.3 The relationship between age and action perception
In contrast to the results of action imagination, the difference between actual MTs and
MTs selected as possible in the perception task decreased as a function of age. Specifically,
younger children were shown to have a larger disparity between actual execution MTs and
perceived MTs than older children, demonstrating that action possibility judgements became
more congruent with actual MTs as a function of the children’s age. Developmental research of
risk taking and injury proneness in children would suggest that this overestimation of abilities in
younger children is not uncommon (Sandseter & Kennair, 2011). It is known, for instance, that
children regularly engage in “risky play” or situations that provide a thrilling experience (such as
36
jumping from heights and moving at high speeds) as part of normal cognitive and motor
development. These behaviours are likely related to the abilities of younger children to estimate
their abilities and the consequences of their actions (Sandseter & Kennair, 2011). Specifically,
children are more likely to overestimate their abilities in tasks beyond their abilities than adults
(Plumert, 1995). Although these abilities may also be influenced by social and individual factors,
there is also an evident developmental component as older children seem to be able to more
accurately characterize their abilities than younger children (Plumert & Schwebel, 1997). It
should be noted here that the body of literature on children’s ability to estimate what they are
capable of performing goes beyond what is very briefly discussed here. For example, authors
have examined the individual and developmental differences that impact a child’s risk perception
for a given task (e.g., Schwebel & Bounds, 2003; Schwebel & Plumert, 1999, see Sandseter &
Kennair, 2011 for review). Critically, however, whereas the body of literature on children’s
assessment of ability focuses mainly on how children perceive their own abilities when they are
presented with a given task, the current study asks children to make a judgement on a movement
that they are actively observing. To the authors’ knowledge, this is the first time that action
perception and possibility judgments have been studied in this way. For this reason, although
developmental differences in ability estimation are likely related to the results of the current
study, these initial results should be interpreted with some caution. Future more dedicated and
expansive work is needed to assess the interpretation of the results.
That said, the current study’s results could also be accounted for by typical perceptual-
motor development. From late childhood to adolescence (the age range used in the current
study), children begin to engage in more complex activities that present new challenges (e.g.
engaging in more open motor skills). These new challenges, coupled with neurophysiological
development, are thought to preclude improvements in perception-action coupling (Sugden &
Wade, 2013). Essentially, more experience and a larger motor repertoire afford older children
stronger associations between actions and their effects, allowing them to better predict the
outcomes of their actions. This effect is evidenced by experimental work demonstrating the
increased ability of older children to intercept objects (Chohan et al., 2008) and to plan safe
movement trajectories (Chihak et al., 2010; Plumert et al., 2007). In the context of the current
experiment, weaker perception-action coupling in younger children could have led to less
37
accurate action possibility judgements in that an inability to link the perceptual effects of an
action (observed by the participants) with a specific motor pattern could hinder a child’s ability
to make effective predictions about the action’s possibility. Although the presence of Fitts' law
suggests that action simulation occurs, underdeveloped perception-action coupling may have
interfered with the transformation of perception-to-action. This underdevelopment may have also
interfered with the ability to relate the motor simulation to an accurate choice of possible MT.
Critically, action possibility judgements rely on the comparison between what is observed and
what is simulated and determination of a threshold for possibility. Children, therefore, may be
choosing faster or “riskier” perceived MTs, as they often do in more ecologically valid
situations, because they cannot adequately predict the true consequences of the observed action
or are not making appropriate comparisons between the observed action and their own internal
representation of the action. This conclusion should be explored in future studies.
2.5.4 Conclusions
The critical finding of the current study was that the developmental trajectories of
imagination and perception from late childhood to adolescence are dissimilar. Between the ages
of seven and sixteen years, the temporal congruency between real and imagined MTs remained
relatively stable, whereas the congruency between MTs selected in the action possibility
judgement task and real MTs increased as children aged. This result stands in contrast to the
hypothesis that the temporal congruencies in these two tasks should increase in a similar way due
to their shared mechanisms. Therefore, it is likely that there are fundamental differences in the
underlying neural mechanisms of action imagination and perception. The presence of Fitts’ law
in both tasks and the stability of action imagination suggest that neural motor simulation is
developed by late childhood. However, the underestimation of movement times in the action
possibility judgements may be due to differences in an additional self-other or perception-action
matching process necessary to form accurate judgements. Specifically, action possibility
judgements involve the comparing or relating of observed action effects to one’s own
representation of the task. Therefore, while children may be able to neurally simulate actions,
their ability to effectively use these simulations to predict movement outcomes likely continues
to develop into adolescence.
38
Chapter 3 Summary and Conclusions
3.1 Summary
Previous literature has demonstrated that when individuals imagine themselves
performing actions and observe the actions of others, there is activation of the neural circuits
associated with actual movement execution. Although the activated neural networks for action
imagination and observation vary between processes and even within experiments, there is
consensus that activation of an overlapping neural network, involving specific motor areas
occurs in action imagination and perception (Grèzes & Decety, 2001). Neurophysiological and
behavioural experiments have supported this notion, indicating that both imagination and
perception involve the internal simulation of actual action (Jeannerod, 2001).
In children, it appears that this mechanism may be present from infancy, facilitating
learning through imitation and observation (Paulus et al., 2013). To further probe the
development of cognitive action representations in children and adolescents, many researchers
have examined how motor imagery ability changes as a function of age. The results of this work
suggest that the ability to form a motor image or to effectively imagine oneself moving is formed
by the age of seven, although this process may continue to be refined into adolescence
(Caeyenberghs, Tsoupas, et al., 2009; Caeyenberghs, Wilson, et al., 2009; Choudhury et al.,
2007a, 2007b; Gabbard, 2009; Molina et al., 2008; Smits-Engelsman & Wilson, 2013). This
developmental trajectory, however, has yet to be linked to developmental differences seen in
other tasks or processes involving motor simulation such as action possibility judgements.
The research reported in the present thesis was designed to analyse developmental
differences in action imagination and perception from late childhood to adolescence. Because it
is thought that action execution, perception and imagination share a representational domain and
the neural simulation of action is a potential mechanism for both action imagination and
perception, it was hypothesized that the temporal congruency of both measures with actual
execution movement times should develop in similar ways. The critical finding of this study,
however, was that there were distinct differences in the developmental changes in congruency of
39
imagination and perception. It was found that there was no consistent effect of age on the
difference between imagined and actual movement times, whereas the difference between
perceived and actual movement times decreased for older children. Specifically, younger
children tended to select movement times that were much shorter than their own, and improved
in this respect as age increased.
3.2 Conclusions
The results of the current study suggest that there are fundamental age-related differences
in the mechanisms for action imagination and perception. One of the more apparent differences
in the tasks is that action imagination is a self- or internally-generated process, where individuals
simulate their own actions in the absence of any visual or terminal feedback. Action possibility
judgements, however, also involve externally-generated stimuli and therefore require that the
perceived action be associated with the appropriate motor pattern, in order to accurately predict
the consequences of the observed action. As such, the action possibility judgement may first
require a kind of self-other matching process, where the observed action is matched to the
participant’s own motor system, and then simulated to make an informed decision. In addition,
the judgement of what is possible or not requires one to set a threshold of possibility. In action
imagination, this matching or transformative process should be more or less absent, as the
participant is already imagining his or her actions.
This difference in initial self-other matching between the tasks may shed light on the
differences in the developmental trajectories noted in this experiment. Previous developmental
literature states that between late childhood and adolescence, there are notable changes in the
temporal and spatial accuracy of movements (Sugden & Wade, 2013) as well as in decision
making and evaluation of risk (Chihak et al., 2010; Plumert, 1995). These changes are thought to
occur because of experience-based refinement of perception-action coupling (Plumert et al.,
2007; Sugden & Wade, 2013) and may speak to developmental differences in the threshold for
possibility. Additionally, although some studies have demonstrated the concept of motor
resonance in young children (Marshall et al., 2010; Saby et al., 2011), the extent to which
children internally represent observed actions may also depend on how similar the observed
agent is to the self (Liuzza, Setti, & Borghi, 2012; Marshall et al., 2010). Therefore, it seems
40
likely that a developmental change in one’s ability to adequately match the observed action onto
one’s own motor system, could lead to the underestimation of MTs in the perceptual task for
younger children. Specifically, difficulty mapping the observed movement onto one’s own motor
system could have led younger children to use alternative, perhaps more cognitive mechanisms
to make appropriate although not entirely congruent judgements. This result highlights that while
both perceptual judgements and motor imagery involve the cognitive representation or
simulation of action, perceptual judgements also involve important comparative or
transformative processes to interpret and predict the consequences of an observed movement.
Additionally, these results highlight that although execution, imagination and perception likely
share common substrates (as proposed in common coding theory), these three processes can also
be limited by other neural mechanisms.
3.3 Limitations and Future Directions
Although the results of the current thesis are informative, certain limitations to the current
experiment remain to be addressed in future studies. These limitations include the relatively
small sample size and constraints in the experimental setup. First, thirty-two participants were
included in this study. Although this number of participants, generally speaking, provides
sufficient statistical power for the current types of analyses in neurotypical adults, future work
should attempt to collect a larger sample size both to be able to afford more complex analyses
with a larger amount of age groups and to account for the larger inherent variability of young
children’s behaviours and development.
Second, the number of trials in each task was limited due to time constraints when
working with children. In particular, because the action possibility judgements were done across
eleven different stimulus onset asynchronies, only one set of judgements was possible to obtain
at each stimulus onset asynchrony (the proxy for MT). Additionally, although the range of SOAs
selected for the task represents typical execution MTs for this type of pointing task, a larger
range of SOAs could be used to ensure that individuals are choosing MTs that they actually
consider possible and that are not simply the mean of presented SOAs. A second set of
judgements and/or a larger range of SOAs may improve the reliability of the perception data and
afford further statistical analyses. However, in the current study, inclusion of a second set of
41
judgements and a larger range of SOAs would have greatly elongated the experimental protocol
which would have led to fatigue and boredom, and likely unreliable data. In the future, this task
could be studied in isolation, which may allow for a larger number of experimental trials. Future
studies may also consider using alternative protocols to measure the threshold MT for possibility
in these judgements. For instance, a staircase procedure (Cornsweet, 1962) may more accurately
capture where children perceive a threshold for possibility while observing an action and
whether this threshold represents a sharp boundary or a progression from perceived impossibility
to possibility. Again, this type of procedure would have been beyond the scope and time limits of
this project but, in isolation, could yield influential results in future studies.
Another limitation might be that the perception task involved the arm of an adult. This
arrangement might have hindered self-other matching and, as such, the reliability of the
judgements in the perception task. A future direction, to understand further how self-other
matching processes may contribute to action perception and prediction, would be to manipulate
the agent used in the perceptual stimuli. Because past research has demonstrated that children
may more readily represent the actions and hands of younger children (Liuzza et al., 2012;
Marshall et al., 2010), it is possible that children in the current study were unable to adequately
represent the actions of the model in the perceptual task. Therefore, manipulation of the hand
model (adult vs child) may provide insight into how self-other matching processes impact action
possibility judgements. For instance, if younger children continue to underestimate perceptual
movement times, despite differences in model type, then developmental differences in this
respect may be related to differences in simulation processes as opposed to matching or
transformation of perception to action, which cannot be differentiated in the current experiment.
Finally, a further future direction would be to expand the current experiment to include
special populations such as children with autism. Autism spectrum disorders (ASD) are
characterized by challenges in social communication and interaction and restricted or repetitive
patterns of behaviour and interests. Importantly, many researchers have suggested that the social
and communicative differences seen in ASD could be related to how children and adults with
ASD are able to represent or mirror the actions and emotions of others (Williams, Whiten,
Suddendorf, & Perrett, 2001). Because the understanding of one’s own and others’ actions
(motor cognition) may rely heavily on action simulation, it would be worthwhile to investigate
42
how children with ASD differentiate from typically developing children in this respect. It would
be hypothesized, that children with autism may show lesser congruency in the imagination and
perception of action than typically developing children, because this process may underlie motor
cognitive processes, as previously discussed. Such a finding would suggest that children with
autism differ in their understanding of others’ actions (such as the gestures used in social
interactions) because of differences in the internal simulation of action. The results of this work
could highlight important areas for new therapies and accessibility services.
43
References
Blakemore, S.-J., & Decety, J. (2001). From the perception of action to the understanding of
intention. Nature Reviews. Neuroscience, 2(8), 561–7. doi:10.1038/35086023
Blakemore, S.-J., & Frith, C. (2005). The role of motor contagion in the prediction of action.
Neuropsychologia, 43(2), 260–7. doi:10.1016/j.neuropsychologia.2004.11.012
Caeyenberghs, K., Tsoupas, J., Wilson, P. H., & Smits-Engelsman, B. C. M. (2009). Motor
imagery development in primary school children. Developmental Neuropsychology, 34(1),
103–21. doi:10.1080/87565640802499183
Caeyenberghs, K., Wilson, P. H., van Roon, D., Swinnen, S. P., & Smits-Engelsman, B. C. M.
(2009). Increasing convergence between imagined and executed movement across
development: evidence for the emergence of movement representations. Developmental
Science, 12(3), 474–83. doi:10.1111/j.1467-7687.2008.00803.x
Chandrasekharan, S., Binsted, G., Ayres, F., Higgins, L., & Welsh, T. N. (2012). Factors that
affect action possibility judgements: recent experience with the action and the current body
state. Quarterly Journal of Experimental Psychology, 65(5), 976–93.
doi:10.1080/17470218.2011.638720
Chihak, B. J., Plumert, J. M., Ziemer, C. J., Babu, S., Grechkin, T., Cremer, J. F., & Kearney, J.
K. (2010). Synchronizing self and object movement: how child and adult cyclists intercept
moving gaps in a virtual environment. Journal of Experimental Psychology. Human
Perception and Performance, 36(6), 1535–1552. doi:10.1037/a0020560
Chohan, A., Verheul, M. H. G., Van Kampen, P. M., Wind, M., & Savelsbergh, G. J. P. (2008).
Children’s use of the bearing angle in interceptive actions. Journal of Motor Behavior,
40(1), 18–28. doi:10.3200/JMBR.40.1.18-28
Choudhury, S., Charman, T., Bird, V., & Blakemore, S.-J. (2007a). Adolescent development of
motor imagery in a visually guided pointing task. Consciousness and Cognition, 16(4),
886–96. doi:10.1016/j.concog.2006.11.001
44
Choudhury, S., Charman, T., Bird, V., & Blakemore, S.-J. (2007b). Development of action
representation during adolescence. Neuropsychologia, 45(2), 255–62.
doi:10.1016/j.neuropsychologia.2006.07.010
Clark, S., Tremblay, F., & Ste-Marie, D. (2004). Differential modulation of corticospinal
excitability during observation, mental imagery and imitation of hand actions.
Neuropsychologia, 42(1), 105–112. doi:10.1016/S0028-3932(03)00144-1
Cornsweet, T. N. (1962). The Staircase-Method in Psychophysics. The American Journal of
Psychology, 75(3), 485–491. doi:10.2307/1419876
Decety, J. (1996). The neurophysiological basis of motor imagery. Behavioural Brain Research,
77, 45–52.
Decety, J., & Jeannerod, M. (1995). Mentally simulated movements in virtual reality: does
Fitts’s law hold in motor imagery? Behavioural Brain Research, 72(1-2), 127.
Eskenazi, T., Grosjean, M., Humphreys, G. W., & Knoblich, G. (2009). The role of motor
simulation in action perception: a neuropsychological case study. Psychological Research,
73(4), 477–85. doi:10.1007/s00426-009-0231-5
Eskenazi, T., Rotshtein, P., Grosjean, M., & Knoblich, G. (2012). The neural correlates of Fitts’s
law in action observation: An fMRI study. Social Neuroscience, 7(1), 30–41.
doi:10.1080/17470919.2011.576871
Fadiga, L., Buccino, G., Craighero, L., Fogassi, L., Gallese, V., & Pavesi, G. (1998).
Corticospinal excitability is specifically modulated by motor imagery: A magnetic
stimulation study. Neuropsychologia, 37(2), 147–158. doi:10.1016/S0028-3932(98)00089-
X
Feurra, M., Bianco, G., Polizzotto, N. R., Innocenti, I., Rossi, A., & Rossi, S. (2011). Cortico-
Cortical Connectivity between Right Parietal and Bilateral Primary Motor Cortices during
Imagined and Observed Actions: A Combined TMS/tDCS Study. Frontiers in Neural
Circuits, 5(August), 10. doi:10.3389/fncir.2011.00010
45
Fietzek, U. M., Heinen, F., Berweck, S., Maute, S., Hufschmidt, A., Schulte-Mönting, J., …
Korinthenberg, R. (2000). Development of the corticospinal system and hand motor
function: central conduction times and motor performance tests. Developmental Medicine
and Child Neurology, 42(4), 220–227. doi:10.1111/j.1469-8749.2000.tb00076.x
Flach, R., Knoblich, G., & Prinz, W. (2004). The two-thirds power law in motion perception.
Visual Cognition, 11(4), 461–481. doi:10.1080/13506280344000392
Gabbard, C. (2009). Studying action representation in children via motor imagery. Brain and
Cognition, 71(3), 234–9. doi:10.1016/j.bandc.2009.08.011
Gabbard, C. (2012). The role of mental simulation in embodied cognition. Early Child
Development and Care, (June 2015), 1–8. doi:10.1080/03004430.2012.681649
Gallese, V., Fadiga, L., Fogassi, L., & Rizzolatti, G. (1996). Action Recognition in the Premotor
Cortex. Brain, 119(2), 593–609. doi:10.1093/brain/119.2.593
Grèzes, J., & Decety, J. (2001). Functional anatomy of execution, mental simulation,
observation, and verb generation of actions: a meta-analysis. Human Brain Mapping, 12(1),
1–19.
Grosjean, M., Shiffrar, M., & Knoblich, G. (2007). Fitts’s Law Holds for Action Perception.
Psychological Science, 18(2), 95–99.
Hétu, S., Grégoire, M., Saimpont, A., Coll, M.-P., Eugène, F., Michon, P.-E., & Jackson, P. L.
(2013). The neural network of motor imagery: an ALE meta-analysis. Neuroscience and
Biobehavioral Reviews, 37(5), 930–49. doi:10.1016/j.neubiorev.2013.03.017
Hommel, B. (2009). Action control according to TEC (theory of event coding). Psychological
Research, 73(4), 512–526.
Hommel, B., Müsseler, J., Aschersleben, G., & Prinz, W. (2001). The Theory of Event Coding
(TEC): a framework for perception and action planning. The Behavioral and Brain
Sciences, 24(5), 849–878. doi:10.1017/S0140525X01000103
46
Hoyek, N., Champely, S., Collet, C., Fargier, P., & Guillot, A. (2009). Age and gender-related
differences in the temporal congruence development between motor imagery and motor
performance. Learning and Individual Differences, 19(4), 555–560.
doi:10.1016/j.lindif.2009.07.003
Jeannerod, M. (1995). Mental Imagery in the Motor Context. Neuropsychologia, 33(11), 1419–
1432.
Jeannerod, M. (2001). Neural Simulation of Action: A Unifying Mechanism for Motor
Cognition. NeuroImage, 14(1), S103–S109. doi:10.1006/nimg.2001.0832
Kilner, J. ., Paulignan, Y., & Blakemore, S.-J. (2003). An Interference Effect of Observed
Biological Movement on Action. Current Biology, 13(6), 522–525. doi:10.1016/S0960-
9822(03)00165-9
Knoblich, G. (2008). Bodily and motor contributions to action perception. Embodiment, Ego-
Space, and Action, (1945), 43–76. doi:10.4324/9780203809891
Koh, T. H., & Eyre, J. a. (1988). Maturation of corticospinal tracts assessed by electromagnetic
stimulation of the motor cortex. Archives of Disease in Childhood, 63(11), 1347–1352.
doi:10.1136/adc.63.11.1347
Lacquaniti, F., Terzuolo, C., & Viviani, P. (1983). The law relating the kinematic and figural
aspects of drawing movements. Acta Psychologica, 54(1-3), 115–130. doi:10.1016/0001-
6918(83)90027-6
Liuzza, M. T., Setti, A., & Borghi, A. M. (2012). Kids observing other kids’ hands: Visuomotor
priming in children. Consciousness and Cognition, 21(1), 383–392.
doi:10.1016/j.concog.2011.09.015
Manson, G. a., Sayenko, D. G., Masani, K., Goodman, R., Wong, L., Popovic, M. R., … Welsh,
T. N. (2014). Action Possibility Judgments of People with Varying Motor Abilities Due to
Spinal Cord Injury. PLoS ONE, 9(10), e110250. doi:10.1371/journal.pone.0110250
47
Marshall, P. J., Bouquet, C. a., Thomas, A. L., & Shipley, T. F. (2010). Motor contagion in
young children: Exploring social influences on perception-action coupling. Neural
Networks, 23(8-9), 1017–1025. doi:10.1016/j.neunet.2010.07.007
Martini, R., Carter, M. J., Yoxon, E., Cumming, J., & Ste-Marie, D. (2015). Development and
validation of the Movement Imagery Questionnaire for Children (MIQ-C). Sport and
Exercise Psychology, Submitted.
Molina, M., Tijus, C., & Jouen, F. (2008). The emergence of motor imagery in children. Journal
of Experimental Child Psychology, 99(3), 196–209. doi:10.1016/j.jecp.2007.10.001
Munzert, J., Lorey, B., & Zentgraf, K. (2009). Cognitive motor processes: the role of motor
imagery in the study of motor representations. Brain Research Reviews, 60(2), 306–326.
Nezu, A., Kimura, S., Uehara, S., Kobayashi, T., Tanaka, M., & Saito, K. (1997). Magnetic
stimulation of motor cortex in children: Maturity of corticospinal pathway and problem of
clinical application. Brain and Development, 19(3), 176–180. doi:10.1016/S0387-
7604(96)00552-9
Paulus, M., Hunnius, S., & Bekkering, H. (2013). Neurocognitive mechanisms underlying social
learning in infancy: Infants’ neural processing of the effects of others' actions. Social
Cognitive and Affective Neuroscience, 8(7), 774–779. doi:10.1093/scan/nss065
Paulus, M., Hunnius, S., Vissers, M., & Bekkering, H. (2011a). Bridging the gap between the
other and me: the functional role of motor resonance and action effects in infants’ imitation.
Developmental Science, 14(4), 901–910. doi:10.1111/j.1467-7687.2011.01040.x
Paulus, M., Hunnius, S., Vissers, M., & Bekkering, H. (2011b). Imitation in infancy: rational or
motor resonance? Child Development, 82(4), 1047–57. doi:10.1111/j.1467-
8624.2011.01610.x
Plumert, J. M. (1995). Relations between children’s overestimation of their physical abilities and
accident proneness. Developmental Psychology, 31(5), 866–876. doi:10.1037/0012-
1649.31.5.866
48
Plumert, J. M., Kearney, J. K., & Cremer, J. F. (2007). Children ’ s Road Crossing A Window
Into Perceptual – Motor Development, 16(5), 255–258.
Plumert, J. M., Kearney, J. K., Cremer, J. F., Recker, K. M., & Strutt, J. (2011). Changes in
children’s perception-action tuning over short time scales: Bicycling across traffic-filled
intersections in a virtual environment. Journal of Experimental Child Psychology, 108(2),
322–337. doi:10.1016/j.jecp.2010.07.005
Plumert, J. M., & Schwebel, D. C. (1997). Social and temperamental influences on children’s
overestimation of their physical abilities: links to accidental injuries. Journal of
Experimental Child Psychology, 67(3), 317–37. doi:10.1006/jecp.1997.2411
Prinz, W. (1997). Perception and Action Planning. European Journal of Perception and Action
Planning, 9(2), 129–154.
Rizzolatti, G., & Craighero, L. (2004). The mirror-neuron system. Annual Review of
Neuroscience, 27, 169–92. doi:10.1146/annurev.neuro.27.070203.144230
Rizzolatti, G., Fadiga, L., Gallese, V., & Fogassi, L. (1996). Premotor cortex and the recognition
of motor actions. Cognitive Brain Research, 3(2), 131–141. doi:10.1016/0926-
6410(95)00038-0
Saby, J. N., Marshall, P. J., Smythe, R., Bouquet, C. a., & Comalli, C. E. (2011). An
investigation of the determinants of motor contagion in preschool children. Acta
Psychologica, 138(1), 231–236. doi:10.1016/j.actpsy.2011.06.008
Sachet, A. B., Frey, S. H., Jacobs, S., & Taylor, M. (2015). Development of the Correspondence
Between Real and Imagined Fine and Gross Motor Actions. Journal of Cognition and
Development, (June), 150527104304009. doi:10.1080/15248372.2014.963585
Sandseter, E. B. H., & Kennair, L. E. O. (2011). Children’s risky play from an evolutionary
perspective: The Anti-phobic effects of thrilling experiences. Evolutionary Psychology,
9(2), 257–284.
49
Schwebel, D. C., & Bounds, M. L. (2003). The Role of Parents and Temperament on Children’s
Estimation of Physical Ability: Links to Unintentional Injury Prevention. Journal of
Pediatric Psychology, 28(7), 505–516. doi:10.1093/jpepsy/jsg041
Schwebel, D. C., & Plumert, J. M. (1999). Longitudinal and Concurrent Relations among
Temperament, Ability Estimation, and Injury Proneness. Child Development, 70(3), 700–
712. doi:10.1111/1467-8624.00050
Sebanz, N., & Knoblich, G. (2009). Prediction in Joint Action: What, When, and Where. Topics
in Cognitive Science, 1(2), 353–367. doi:10.1111/j.1756-8765.2009.01024.x
Serino, A., De Filippo, L., Casavecchia, C., Coccia, M., Shiffrar, M., & Làdavas, E. (2010).
Lesions to the motor system affect action perception. Journal of Cognitive Neuroscience,
22(3), 413–426. doi:10.1162/jocn.2009.21206
Sharma, N., & Baron, J.-C. (2013). Does motor imagery share neural networks with executed
movement: a multivariate fMRI analysis. Frontiers in Human Neuroscience, 7(September),
564. doi:10.3389/fnhum.2013.00564
Sirigu, A. ., Cohen, L., Duhamel, J. R., Pillon, B., Dubois, B., Agid, Y., & Pierrot-Deseilligny,
C. (1995). Congruent unilateral impairments for real and imagined hand movements.
NeuroReport, 6, 997–1001.
Skoura, X., Vinter, A., & Papaxanthis, C. (2009). Mentally simulated motor actions in children.
Developmental Neuropsychology, 34(3), 356–367. doi:10.1080/87565640902801874
Smits-Engelsman, B. C. M., & Wilson, P. H. (2013). Age-related changes in motor imagery from
early childhood to adulthood: probing the internal representation of speed-accuracy trade-
offs. Human Movement Science, 32(5), 1151–62. doi:10.1016/j.humov.2012.06.006
Sommerville, J. a., & Decety, J. (2006). Weaving the fabric of social interaction: Articulating
developmental psychology and cognitive neuroscience in the domain of motor cognition.
Psychonomic Bulletin & Review, 13(2), 179–200. doi:10.3758/BF03193831
50
Stadler, W., Ott, D. V. M., Springer, A., Schubotz, R. I., Schütz-Bosbach, S., & Prinz, W.
(2012). Repetitive TMS Suggests a Role of the Human Dorsal Premotor Cortex in Action
Prediction. Frontiers in Human Neuroscience. doi:10.3389/fnhum.2012.00020
Stippich, C., Ochmann, H., & Sartor, K. (2002). Somatotopic mapping of the human primary
sensorimotor cortex during motor imagery and motor execution by functional magnetic
resonance imaging. Neuroscience Letters, 331(1), 50–54. doi:10.1016/S0304-
3940(02)00826-1
Sugden, D., & Wade, M. G. (2013). Typical and atypical motor development. London: Mac
Keith Press.
Welsh, T. N., Wong, L., & Chandrasekharan, S. (2013). Factors that affect action possibility
judgments: The assumed abilities of other people. Acta Psychologica, 143(2), 235–244.
doi:10.1016/j.actpsy.2013.04.003
Williams, J. H. G., Whiten, A., Suddendorf, T., & Perrett, D. I. (2001). Imitation, mirror neurons
and autism. Neuroscience and Biobehavioral Reviews, 25, 287–95.
Wong, L., Manson, G. A., Tremblay, L., & Welsh, T. N. (2013). On the relationship between the
execution, perception, and imagination of action. Behavioural Brain Research, 257, 242–
252.
Yeo, S. S., Jang, S. H., & Son, S. M. (2014). The different maturation of the corticospinal tract
and corticoreticular pathway in normal brain development: diffusion tensor imaging study.
Frontiers in Human Neuroscience, 8(August), 1–6. doi:10.3389/fnhum.2014.00573
Young, S. J., Pratt, J., & Chau, T. (2009). Misperceiving the speed-accuracy tradeoff: Imagined
movements and perceptual decisions. Experimental Brain Research, 192(1), 121–132.
doi:10.1007/s00221-008-1563-x
Yoxon, E., Tremblay, L., & Welsh, T. N. (2015). Effect of task-specific execution on accuracy
of imagined aiming movements. Neuroscience Letters, 585, 72–6.
doi:10.1016/j.neulet.2014.11.021
51
Appendices
52
Appendix A: Edinburgh Handedness Questionnaire
53
Appendix B: Additional Analysis
One of the goals of this thesis was to investigate how the accuracy of simulation in action
perception and imagination changed as a function of age. Although the correlational analysis
presented in the above paper largely answers this question, a subsequent analysis sought to
characterize mean imagined and perceived MTs as ratios, in relation to actual execution MT. To
this end, a ratio was calculated for each participant, with perception or imagination as the
numerator and actual execution MT as the denominator. In order to ensure data that has an even
interval between values (and does not violate the assumptions of an ANOVA), these values were
again transformed by taking the natural logarithm of the ratios. These transformed values were
submitted to a 2 (Task: Perception, Imagination) by 3 (Group: Younger, Older, Adult) mixed
ANOVA with repeated measures on the first factor. The results of this analysis were similar to
that of the initial ANOVA on the absolute MTs, as there was a significant effect of task, F (1,40)
= 131.40, p < .001 and a significant task by group interaction, F (2, 40) = 3.74, p < .05. There
was, however, no significant effect of group, F (2, 40) = 3.06, p = .083 (Figure 4).
To further analyze the effect of age on imagined and perceived MT as a ratio of actual
execution MT, the transformed MT data was correlated with age for the child participants.
Similar to the correlational analysis using difference scores, it was found that the ratio values for
perception were significantly correlated with age, r = .68, p < .001 (Figure 5-A), whereas the
imagination ratio values were not significantly correlated with age, r = .17, p = .36 (Figure 5-B).
It was decided that because the results of this analysis mimicked those of the analysis
presented previously that it would not add any new information. Therefore, this analysis was not
included in the paper written for publication.
54
Figure 4.1. Ln transformation of ratios of imagination and perception MT as a function of actual
execution movement time for each of the groups.
Figure 4.2. Ln transformation of ratios of imagination and perception MT (as a function of actual
execution MT), correlated with age.
Related Documents
