Exploring models of time processing

Effects of training and modality, and the relationship with cognition in rhythmic motor

tasks

Olympia Karampela

Department of Psychology

Umeå 2017

Responsible publisher under swedish law: the Dean of the Social Sciences

This work is protected by the Swedish Copyright Legislation (Act 1960:729)

ISBN: 978-91-7601-668-8

Cover design by George Andreou

Elektronisk version tillgänglig på http://umu.diva-portal.org/

Tryck/Printed by: UmU Print Service, Umeå University

Umeå, Sweden 2017

i

Ithaca

When you start on your journey to Ithaca,

then pray that the road is long,

full of adventure, full of knowledge.

Do not fear the Lestrygonians

and the Cyclopes and the angry Poseidon.

You will never meet such as these on your path,

if your thoughts remain lofty, if a fine

emotion touches your body and your spirit.

You will never meet the Lestrygonians,

the Cyclopes and the fierce Poseidon,

if you do not carry them within your soul,

if your soul does not raise them up before you.

Then pray that the road is long.

That the summer mornings are many,

that you will enter ports seen for the first time

with such pleasure, with such joy!

Stop at Phoenician markets,

and purchase fine merchandise,

mother-of-pearl and corals, amber and ebony,

and pleasurable perfumes of all kinds,

buy as many pleasurable perfumes as you can;

visit hosts of Egyptian cities,

to learn and learn from those who have knowledge.

Always keep Ithaca fixed in your mind.

To arrive there is your ultimate goal.

But do not hurry the voyage at all.

It is better to let it last for long years;

and even to anchor at the isle when you are old,

rich with all that you have gained on the way,

not expecting that Ithaca will offer you riches.

Ithaca has given you the beautiful voyage.

Without her you would never have taken the road.

But she has nothing more to give you.

And if you find her poor, Ithaca has not defrauded you.

With the great wisdom you have gained, with so much experience,

you must surely have understood by then what Ithacas mean.

-K. P. Kavafis (C. P. Cavafy), translation by Rae Dalven

ii

iii

To my son and my husband

iv

v

TABLE OF CONTENTS

ABSTRACT vii

LIST OF PAPERS ix

SAMMANFATTNING PÅ SVENSKA x

ABBREVIATIONS xi

INTRODUCTION 1 Aims of the thesis 2 Time perception and the production of timed intervals 4 Dedicated models of timing 6 One clock? 9 Timing at different timescales 12 Task differences 14 The effect of feedback 15 The intrinsic models of timing 16 Cognitive capacities in timing performance 20 Impaired timing in clinical populations with an attention disorder 24 Attentional models of timing 25 Concluding remarks 28 The Main objectives of the Thesis 28

METHODS 29 Participants 29 Instruments-Procedure 31 Cognitive measurements (Study II) 32 Eye tracking (Study III) 33 Dependent measurements computation 33 Saccade onsets (Study III) 34 Statistical Analysis 34

Results 35 Study I 35 Study II 40 Study III 44

GENERAL DISCUSSION 47 Effects of training. Motor and cognitive gains 47 Modality differences – dedicated or intrinsic mechanisms? 51 Implications 54 Limitations 55

vi

Future 56

REFERENCES 58

ACKNOWLEDGEMENTS 76

vii

ABSTRACT

Timing can be defined as the ability to perceive temporal sequences and

regulate timed behaviors. As in other animals, our ability to make accurate

time estimations is crucial in order to accomplish several activities.

Organisms can process time over a wide range of durations ranging from

microseconds to days. In the middle of these extremes is the hundreds of

milliseconds to seconds range which is important for many everyday

behaviors, such as walking, speaking and dancing. Yet, how this is managed

remains poorly understood. Some central issues with regard time processing

in this particular time range are whether timing is governed by one, or by

several different mechanisms, possibly invoked by different effectors used to

perform the timing task, as well as, if cognitive capacities are also involved in

rhythmic motor timing.

This thesis includes three studies. Study I investigated the effects of short-

term practice on a motor timing task. Analyses of the timing variability

indicated that a substantial amount of learning occurred in the first hour of

practice and declined afterwards, exhibiting no trend for further decrease

across the remaining 60 or 210 minutes. This effect was similar across

effector, amount of feedback, and interval duration. Our results suggested

that training effects influenced mainly motor precision and raised the

question of whether motor timing training influenced also cognitive

capacities.

Study II investigated the relationship between motor timing and cognition.

Specifically, participants had to train a sensorimotor synchronization task

(SMS) over several days, and the question was whether this training would

improve cognitive performance. A near transfer effect was found between the

sensorimotor synchronization task and the sustained attention task,

indicating that sustained attention is involved in motor timing.

Study III compared the timing variability between the eyes and the hands, as

a function of four different intervals, in order to examine whether these

systems are temporally controlled by the same or different mechanism(s).

The results showed several positive correlations in variability, between the

eye and the finger movements, which, however, were significant only for the

longer intervals. In addition, they were differences in variability between the

eye and the hand, for the different interval durations.

In general, the pattern of results from these studies suggested that voluntary

motor timing is managed by overlapping distributed mechanisms and that

viii

these mechanisms are related to systems that manage cognitive processes,

such as attention. The results partially explain the well-known relationships

between cognitive ability and timing.

ix

LIST OF PAPERS

I. Madison, G., Karampela, O., Ullén, F., & Holm, L. (2013).

Effects of practice on variability in an isochronous serial

interval production task: Asymptotical levels of tapping

variability after training are similar to those of musicians.

Acta Psychologica, 1, 119-128.

II. Karampela, O., Madison, G., & Holm, L. (manuscript). Motor

timing training improves sustained attention performance.

III. Karampela, O., Holm, L., & Madison G. (2015). Shared timing

variability in eye and finger movements increases with interval

duration: Support for a distributed timing system below and

above 1 second. The Quarterly Journal of Experimental

Psychology, 68, 1965-1980.

Papers I and III have been reproduced with the permission of the

copyright holders.

x

SAMMANFATTNING PÅ SVENSKA

Timing innebär att uppfatta tidsmässiga förlopp och tidsmässigt reglera

beteende. Exakt timing är nödvändig för att koordinera vårt eget beteende

med andra personer och processer i vår miljö, inom ett spann, från

millisekunder till dagar eller år. För många vardagliga beteenden, som att

tala, promenera, dansa och köra bil är det kritiska spannet från några 10-tals

millisekunder till några sekunder. Hur detta går till är relativt outforskat.

Några centrala frågor är huruvida sådan timing styrs av en eller flera olika

mekanismer, om olika delar av kroppen regleras av samma eller olika

mekanismer, och i vilken mån kognitiva processer och förmågor är

involverade i timing av repetitiva motoriska förlopp.

Denna avhandling omfattar tre empiriska studier. Studie 1 undersökte

effekterna av träning på variabiliteten i timing av en motorisk rörelseuppgift.

Den största delen av minskningen i tidsvariabilitet inträffade under den

första timmen. Ytterligare träning i upp till 3.5 timmar verkade inte minska

tidsvariabiliteten. Detta gällde alla intervaller, från 500 ms till 1.6 s, oavsett

deltagarna använde ett finger eller en trumstock, och oavsett de fick taktil

eller auditiv återkoppling från sina egna responser. Resultaten tyder på att

träning främst påverkade motorisk precision, och väckte frågan om motorisk

timing påverkas av kognitiv förmåga.

Studie II undersökte sambandet mellan motorisk timing och kognition.

Deltagarna fick träna en sensomotorisk synkroniseringsuppgift (SMS) över

flera dagar, och frågan var om det skulle förbättra kognitiv prestation. Sådan

”near transfer” kunde ses från timing-träningen och prestationen på ett

uppmärksamhetstest, men inte på två spatiala intelligenstester.

Studie III jämförde timing-variabilitet mellan ögon och händer som en

funktion av fyra olika intervaller mellan 500 ms och 1.4 s, för att undersöka

om dessa system tidsmässigt styrs av samma eller olika mekanismer. Det

fanns positiva samband i variabilitet mellan ögon- och fingerrörelser, som

dock var signifikanta endast för längre intervall. Det fanns också stora

skillnader i relationen mellan intervall och variabilitet mellan ögon- och

fingerrörelser.

Sammantaget indikerar de tre studierna att motorisk timing hanteras av

delvis överlappande, distribuerade mekanismer och att dessa mekanismer är

relaterade till system som hanterar kognitiva processer som

uppmärksamhet. Resultaten förklarar delvis de välkända sambanden mellan

kognitiv förmåga och timing.

xi

ABBREVIATIONS

ADHD (Attentional Deficit Hyperactivity Disorder)

AGM (Attentional Gate Model)

ANOVA (Analysis of Variance)

CPT II (Continuous Performance Task II)

CTBG (Cortico-Thalamic-Basal Gnglia)

CV (Coefficient of Variation)

DAMP (Deficits in Attention, Motor control and Perception)

DNA (Deoxyribonucleic Acid)

f-MRI (functional Magnetic Resonance Imaging)

IOI (Inter-Onset-Interval)

IQ (Intelligence Quotient)

IRI (Inter-Response-Interval)

ISI (Inter-Stimulus-Interval)

M (Mean)

PET (Positron Emission Tomography)

PFC (Parietal Frontal Cortex)

RT (Reaction Time)

SBF (Striatal Beat Frequency)

SD (Standard Deviation)

SDN (State-Dependent Network model)

SET (Scalar Expectancy Theory)

xii

SMA (Supplementary motor area)

SMS (Sensorimotor Synchronization Task)

TBI (Traumatic Brain Injury)

5-HT (5-Hydroxy-Tryptamine or serotonin)

1

INTRODUCTION

Timing is fundamental for our existence. As in other animals, our ability to

regulate and estimate timed behaviors with high precision is crucial in order

to accomplish several activities. Organisms can process time over a wide

range of durations ranging from microsecond to days (Fig. 1). In this extreme

range of time processes, the hundreds of milliseconds to seconds range is

important for everyday behavior, including motor control and cognitive

operations, such as directing attention and language processing (Justus &

Ivry, 2001; Meck & Benson, 2002; Schirmer, 2004). For example, walking,

and dancing require precise temporal control of body movements.

Furthermore, to ensure accurate speech production and comprehension,

speakers must successfully time the production of individual sounds

precisely. Timing in the hundreds of milliseconds to seconds range is

essential for learning and for representing sequential relationships between

stimuli (Bangert, Reuter-Lorenz, & Seidler, 2011). Yet, there is no dedicated

sensory system, such as an organ or brain area for time, such as there is for

sight, hearing, and taste. There has been a surge in research dedicated to

timing over the past decades, but it has still not led to any consensus about

the nature of the mechanism(s) for timing in the order of the hundreds of

milliseconds and seconds.

Fig. 1. Scales of temporal processing. Adapted from Buonomano &

Karmarkar, (2002).

Two general approaches have been proposed to describe how the passage of

time is perceived in the hundreds of milliseconds to seconds range. The first

is the dedicated approach, where a specialized mechanism that represents

2

the temporal relationships between events is assumed (Droit-Volet, Meck, &

Penney, 2007; Woodrow, 1930). The second approach includes models that

assume intrinsic explanations, and describe time as a general and inherent

property of neural dynamics (Dragoi, Staddon, Palmer & Buhusi, 2003;

Karmarkar & Buonomano, 2007). Specifically, models in this framework

propose that the representation of duration could be an omnipresent

manifestation of intrinsic capacities of neural mechanisms, without any

apparent dedicated function to the treatment of time.

Aims of the thesis

The aims of this thesis were to narrow down these possibilities and provide

further understanding of the phenomena, so that these frameworks could be

evaluated. To these ends, three areas of specific interest have been identified.

They are the role of practice, the role of cognition, and the role of the

modality used to perform a motor timing task, in relation to timing

performance. This led to three more specific aims for the present thesis. The

first was to investigate the impact of practice on a motor timing task, in

order to explore the differential effects of training on two different

measurements of variability, namely the cognitive and the motor component

of timing variability. The second research aim concerned the role of

cognition in motor timing and specifically, the role of sustained attention

and fluid intelligence in motor timing performance, in order to explore

further the relationship found between timing variability and measurements

of sustained attention and fluid intelligence. The third research aim was to

study whether motor timing in the hundreds of milliseconds to seconds

range is managed by a dedicated single timing mechanism or by multiple,

effector-dependent timing mechanisms. Below, the rationale of choosing to

study these particular scientific questions will be described.

In study I we investigated the effect of practice on motor timing variability.

Relatively few studies within timing research have considered learning as an

important research question to assess. The study of learning and

generalization to other behaviors and modalities can provide important

insights into the nature of temporal processing in the hundreds of

milliseconds to seconds range. Specifically, we would expect that an

improvement in motor timing followed by training would generalize to

untrained conditions, and would display no differences with regard to the

interval, the effector used or the feedback provided if time is processed by a

dedicated mechanism. The rationale behind this is that the existence of a

dedicated timing mechanism implies that the “internal clock” should be

independent of modality, task, and context (Buonomano & Karmarkar 2002;

3

Zelaznik, Spencer, & Ivry, 2002). Previous studies have shown that timing

training generalized to different duration intervals (Karmarkar &

Buonomano, 2003), sensory modalities, as well as, from perceptual to

motor-timing tasks (Meegan, Aslin, & Jacobs, 2000). Here, we expand on

the previous research by using a motor timing task and by testing how these

training effects depend on a number of key variables; the duration of the

interval to be timed, the presence of auditory feedback, and the particular

way used to produce the intervals.

In study II we explored the transfer gains from a sensorimotor

synchronization task (SMS) to fluid intelligence and sustained attention

tasks, by comparing a treatment group with an untrained control group.

Despite the vast amount of studies that have shown a relationship between

timing and cognition, the central question about the role of cognitive

functions in timing processing remains largely unanswered. To draw

stronger inferences about this relationship, one possibility would be to study

transfer effects between timing learning and other cognitive functions. If

there is a causal link between motor timing and cognition, it should be

possible to train one and see effects also on the other.

In study III we investigated the effects of modality on performance on a

motor timing task. Specifically, we compared timed finger and eye

movements. Several correlations have been found between different

modalities used to perform a timing task. These correlations imply that

participants who perform well in one behavioral context will also perform

well in another, and they have been used in support of the dedicated timing

models. Consistent with this idea, a dedicated timing mechanism should be

independent of modality, task, and context (Buonomano & Karmarkar 2002;

Zelaznik, Spencer, & Ivry, 2002). So far, the correlations reported have been

between effectors displaying similar response requirements, such as the

hand and the foot. It is possible that similarity between tasks or effectors

involves recruitment of the same or similar timing mechanisms. Timing

behavior and correlations between less similar effector systems would

provide a stronger test and evidence of the dedicated mechanism hypothesis.

4

Time perception and the production of timed intervals

An important issue is whether the processing of time intervals is based on

the same timing process in perception (discrimination) and motor timing

(tapping) tasks (Keele, Pokorny, Corcos, & Ivry, 1985; Ivry & Hazeltine,

1995; Repp, 1998) (Fig. 2). Several researchers have posed the question of

whether tasks with different response requirements depend on the same

dedicated timing mechanism(s) (Ivry & Hazeltine, 1995; Keele et al.,

1985; Merchant, Zarco, & Prado, 2008). A general consensus among these

studies is that being able to translate durations into motor programs for

production (motor timing), is rather different than comparing two durations

represented in memory (perceptual timing). Therefore, cognitive capacities

might be differently involved in perceptual and motor timing.

Time perception refers to the subjective experience of time, for example

someone's own perception of a specific duration. Four main tasks are

commonly used to investigate the mechanisms involved in time perception.

These are (1) time discrimination, (2) time production, (3) verbal estimation,

and (4) time reproduction (Grondin, 2010). In time discrimination tasks,

subjects first compare interval durations and then indicate whether these

durations were different or not, or whether the second interval duration was

shorter or longer than the first. In the time production method, a participant

produces an interval that the experimenter has specified. Two finger “taps”

marking the beginning and end of the interval are usually involved in time

production. Another way is by pushing a button for a duration that the

participant decides was equivalent to the target interval. In the verbal

estimation method, the participant has to provide a verbal estimation of a

target interval in seconds or minutes. Finally, in the reproduction method

the participant has to reproduce the length of an interval that an

experimenter has presented, usually with a continuous auditory or visual

stimuli (Grondin, 2010).

Motor timing refers to the motoric implementation of a timed action. One of

the most common experimental paradigms when investigating motor timing

is the synchronization-continuation tapping. In this task, subjects have to

produce a continuous series of intervals with their taps. Specifically, subjects

listen to sounds which they synchronize their tap with and then continue

tapping at the same pace without listening to the external sounds (Repp,

2005). Two dependent variables are extracted in this procedure, the mean

(M) produced intervals, and variability expressed as the standard deviation

(SD) of the distribution of the intervals. From the SD it is possible to extract

the coefficient of variation (CV) which is the ratio of SD to M, as well as other

5

metrics (Madison, 2001). Based on assumptions of underlying functional

models of both synchronization and production, higher-order parameters of

the behavioral processes can also be estimated (e.g. Wing & Kristofferson,

1973; Semjen, Vorberg, & Schulze, 1998; Vorberg & Wing, 1996). For

example, Wing and Kristofferson (1973) proposed that variability during the

continuation phase arises from two sources: an internal clock, which triggers

the time for a movement, and a motor implementation system, that

translates this signal into a movement. The clock and the motor processes

are considered to be independent. However, this independence is violated if

tempo tends to drift, meaning that there is a slow change in the tempo.

Fig. 2. Motor timing and perceptual timing. Adapted from Breska and Ivry,

2016.

Following the introduction, dedicated models of timing as well as behavioral

and neuroimaging data that support their existence will be presented. The

Scalar Expectancy Theory (SET), which is the most influential model within

this category will be described in more detail. Other dedicated models, as

well as the limitations concerning these particular models, will also be

introduced. After that, the second approach which includes intrinsic models

that describe time as a general and inherent property of neural dynamics will

be presented and discussed, in comparison to the dedicated models. Finally,

the correlations found between timing variability and several cognitive traits,

6

such as working memory, attention and intelligence will be highlighted and

various cognitive models of timing that can accommodate and provide

explanations for these results will be presented. The role of sustained

attention in motor timing will be particularly addressed, given that most

timing models address the role of sustained attention in timing. A more

recent attempt to combine the existing traditional theories and models for

temporal processing in the hundreds of milliseconds to seconds range will

also be presented and discussed. Finally, a number of unresolved issues will

be discussed, based on their crucial role in increasing our understanding of

the nature of temporal processing in the hundreds of milliseconds to seconds

range.

Dedicated models of timing

The dedicated approach assumes dedicated neural mechanisms that are

specialized at representing the temporal relationships between events (Bueti,

2011; Ivry & Schlerf, 2008). This category includes models that are clock-

based, where a pacemaker-accumulator system is proposed. According to

these models, a pacemaker emits ‘ticks’ that are collected by an accumulator

which stores the number of ticks (Droit-Volet et al., 2007; Gibbon, Church &

Meck, 1984; Woodrow, 1930). The more ticks that are accumulated, the

greater the duration perceived. In other words, a 300 ms and 400 ms

auditory interval would involve the same pacemaker but the longer interval

would achieve a larger count in the accumulator (Spencer, Karmakar & Ivry,

2009). Creelman (1962) and Treisman (1963) formulated two of the earliest

models in this category. They both proposed a pacemaker-accumulator

system but disagreed on whether external variables might affect the

frequency with which the pacemaker emits pulses. Greelman assumed that

this frequency is fixed, whereas Treisman argued that various external

factors can affect the frequency of the pacemaker (Creelman, 1962; Treisman

1963).

The most influential model in the dedicated timer category is the Scalar

Expectancy Theory (SET) proposed by Gibbon et al. (1984) (Fig 3). This

model posits an internal clock, memory and decision processes. Specifically,

SET proposes a clock system that consists of a pacemaker that generates or

emits neural ticks which are sent to the accumulator. This accumulation of

ticks represents the elapsed time. The representation of the duration is then

transferred to working memory and then compared to a reference standard

in the long-term memory. Finally, in the decision stage, a comparison is

made between the contents of the long term and the working memory

(Gibbon, et al., 1984). The rate of the pacemaker could be influenced by the

7

stimulus intensity (Zelkind, 1973) and switch latencies (Wearden, Edwards,

Fakhri, & Percival, 1998). The relationship between timing variability and

the target duration should follow Weber’s law; that is the standard deviation

of the represented duration increases proportionally with increasing

duration. In other words, the coefficient of variation (CV), which is the

standard deviation divided by the mean target interval, should be constant

across durations (Gibbon, et al., 1984).

Fig 3. Outline of the scalar expectancy theory (SET). The pacemaker-

accumulator clock is shown to the left, the long term reference memory and

the short-term working memory in the middle, and the decision mechanism

to the right. Adapted from Gibbon et al. (1984).

Neural underpinnings of dedicated timing models

Dedicated models of timing are by their nature modular. It is therefore not

surprising that much research has been aimed at identifying specific brain

areas that constitute basic elements of a dedicated system. The basal ganglia

have been widely considered as a core element of the timing system (e.g.

Buhusi & Meck, 2009; Meck & Benson, 2002) This finding is based on

evidence which shows that dopamine manipulations known to affect the

basal ganglia also affect time performance (Rammsayer, 1993; Meck &

Benson 2002; Ivry & Spencer 2004), as well as, on evidence which has

8

shown that Parkinson patients, who suffer from problems in the basal

ganglia, exhibit deficits in timing in the hundreds of milliseconds to seconds

range (Artieda, Pastor, Lacruz, & Obeso, 1992; Harrington, Haaland &

Hermanowicz, 1998).

The cerebellum has also been also implicated in dedicated timing, by several

studies indicating that the cerebellum has a central role in tasks that require

precise timing (Ackermann, Graber, Hertrich & Daum, 1999; Ivry, Keele, &

Diener, 1988). An extensively used example here is the eyeblink conditioning

method, which involves the relative simple procedure of pairing an auditory

or visual stimulus, with an eyeblink-unconditioned stimulus (air puff). It is

well established that people with damage in the cerebellum exhibit poor

performance on this task. In general, it has been reported that an intact

cerebellum and interpositus nucleus are required for this form of classical

conditioning. Taken together, different research findings indicate that the

cerebellum has a computational role in different timing tasks (Ivry &

Hazeltine, 1995). In addition, prefrontal and parietal cortices seem to be

among the main cortical regions involved in timing and time perception

(Koch, Oliveri, Carlesimo, & Caltagirone 2002; Lewis & Miall, 2006) while

the basal ganglia has been proposed to be the main subcortical cerebral

structure involved in the processing of temporal information (Pouthas et al.,

2005; Jahanshahi, Jones, Dirnberger & Frith, 2006).

Other dedicated models do not suggest a specific area in the brain but

instead, they propose that the successful representation of time intervals is

the result of activity deriving from a distributed network of regions. For

example, the operation of some areas could be specific to timing, such as a

pacemaker function, whereas other areas might be related to more general

functions, such as working memory which stores temporal information.

Deficits in any part of this network would disrupt performance on timing

tasks (Ivry & Schlerf, 2008).

Another model of dedicated systems suggests that timing is managed by a

frontal-striatal circuitry (Meck & Benson, 2002; Matell & Meck, 2004;

Matell, Meck, & Nicolelis, 2003). In this model, the striatum plays a critical

computational role by receiving millions of impulses from the cortical cells.

These cortical cells, which have firing rates from 10 to 40 cycles per second,

begin firing simultaneously for a moment, creating a specific pattern of

neural activity. Then, the substantia nigra sends a message to the striatum

when the timekeeping activity must stop. Through this dopaminergic

activation, identification of a specific interval length is achieved

(Rammsayer, 2008).

9

The idea of a dedicated mechanism has been widely used and validated by

many behavioral studies, in both animals and humans. Below, several

present behavioral and neuroimaging results which support the existence

and the highly adaptive nature of the dedicated models in several different

settings will be described.

One clock?

Timing variability in the hundreds of milliseconds to seconds range has been

found to increase monotonically as a function of the interval to be timed, in

accordance with Weber’s law. This scalar property compliance is true for a

wide range of duration from 200 ms to 2 s (for a review see Grondin, 2001).

Weber’s law implies the constancy of the coefficient of variation, across

different intervals (Allan & Kristofferson, 1974; Gibbon, 1977; Killeen &

Weiss, 1987). Weber’s law is given as SD (T) = kT, where k is a constant

corresponding to the Weber fraction. Furthermore, the Weber fractions

display similar values in the range of hundreds of milliseconds in tasks with

different response requirements and different modalities (Merchant, Zarco &

Prado, 2008). These findings are suggestive of a common dedicated

mechanism.

The overall variability in a motor timing task can be dissociated into both

time-dependent and time-independent sources of variation (Wing &

Kristofferson 1973). Slope analysis is one of the methods used to dissociate

the total performance variability. Assuming a linear relationship between

timing variability (SD) and the mean of the inter- response intervals (IRIs),

the slope of the regression line represents the duration-dependent timer SD,

and the intercept represents the duration-independent component (e.g.,

motor). It has been demonstrated that the slopes of interval discrimination

and motor timing tasks are similar for a range of intervals ranging from 325

to 550 ms (Ivry & Hazeltine, 1995), supporting the view of one dedicated

clock in different timing contexts. In addition, it has been shown that

temporal variability is correlated between different tasks, such as time

perception and time production tasks (Keele et al. 1985). This implies that

individuals who perform well in time perception tasks also perform well in

other behavioral contexts, such as the time. The fact that individuals can be

equally good in many different behavioral contexts with regard to their

timing performance provides additional evidence for the dedicated

perspective of timing.

Support for the dedicated models of timing stems from neuroimaging studies

that propose that specific brain areas contribute to the successful processing

of time. The cerebellar timing hypothesis, for example, suggests that

10

cerebellum contributes in many different types of tasks that require precise

timing. Damage to the cerebellum can disrupt timing performance on a

number of tasks, such as duration discrimination, the organization of skilled

movements and in eyeblink conditioning. Ivry and colleagues (2002)

proposed that while the cerebellum contributes in many different skilled

tasks, its contribution might be more important when the tasks involve event

based timing (Breska & Ivry, 2016).

Finally, results from training studies have demonstrated that training in one

timing context can be generalized to other timing contexts. For example,

Meegan et al (2000), showed that training to discriminate time intervals

caused an improvement in temporal performance in a motor timing task,

favoring the dedicated mechanism hypothesis (Buonomano & Karmarkar,

2002).

The reasons why the concept of a dedicated mechanism is favored are its

conceptual simplicity and reasonable account of many empirical

observations. However, the concept of a dedicated, specialized mechanism is

not unanimously supported. The assumptions of a dedicated mechanism

have been violated several times in many different ways. Therefore, a better

and a more neurally plausible explanation for the processing of time in the

hundreds of milliseconds to seconds range than the centralized, supramodal

timer is called for. Below, various different contexts where these violations

occur and where other possibilities for processing time seem more

reasonable will be presented.

The existence of a dedicated mechanism implies that the timing behavior

should be independent of modality, task, and context (Buonomano &

Karmarkar, 2002; Zelaznik et al., 2002). Indeed, several correlations have

been found between different modalities, but violations have also been

reported. For example, it has been shown that performance in timing tasks is

more accurate for auditory rather than visual stimuli and for multiple

isochronous intervals rather than single intervals (Ivry & Hazeltine, 1995).

The question arising here is how several non temporal factors, such as the

nature of the tasks used and the different modalities, can affect timing

performance. Do all these external factors directly affect the dedicated clock

or should we start thinking about different clocks? (Grondin, 2001).

One clock or several clocks?

Transitions and non- linearities in timescales provide additional evidence

against the dedicated clock. Temporal variability does not always conform to

11

Weber’s law, and several breakpoints (violations of Weber’s law) have been

found at 1s (Allan & Gibbon 1991; Jingu, 1989, Madison 2001), and between

1 and 2 sec intervals (Getty, 1975; Matthews & Grondin, 2012; Madison

2006;) suggesting multiple clock timing mechanisms across those time

frames. This resonates with Michon’s (1985) view, who argued that there is a

distinction between automatic (< 500 ms) versus cognitive (> 500 ms)

temporal processes, as well as with Karmarkar and Buonomano (2007), who

suggested that this duration is the transition between a state-dependent

(< 500 ms) and a scalar timer (> 500 ms). Finally, several neuroimaging

experiments implicate a shift between motor and cognitive timing systems in

the region above 1 s (Lewis & Miall, 2003a; 2003b; 2006).

Additional evidence against the dedicated clock mechanism comes from

research findings showing differences in duration discrimination across

sensory modalities. For example, a well-established finding is that interval

discrimination is more accurate for auditory compared to visual stimuli

(Grondin & Rousseau, 1991; Jaskowski, Purszewicz, & Swidzinski, 1990).

There is also evidence that these differences can be specific to a single

sensory system. It has been found, for example, that saccadic eye movements

cause compression of the perceived duration of visual but not auditory

stimuli (Burr, Cicchini, Arrighi, & Morrone, 2011; Morrone, Ross, & Burr,

2005). The general notion of a single clock has also been challenged by the

finding that temporal perception in any given spatial region of the visual

field seems to be independent of perception in other regions of the visual

field (Burr, Tozzi, & Morrone, 2007). Moreover, significant differences in

timing accuracy have been reported when participants perform a timing task

with the hand and the feet, as well as with the dominant and the non-

dominant side (Kauranen & Vanharanta, 1996). Failure to observe a cross

modality transfer adds to the non-dedicated clock mechanism hypothesis.

Several studies failed to find an improvement in visual duration

discrimination after temporal training in the auditory mode (Grondin

Gamache, Tobin, Bisson, & Hawke 2008; Grondin, Bisson, Gagnon,

Gamache, & Matteau, 2009; Lapid, Ulrich & Rammsayer, 2009). These

unsuccessful attempts for improving visual duration discrimination with

audition have been interpreted as an indirect support against the idea of the

one dedicated clock and suggest that processing of timing events is modality-

specific.

Although one dedicated mechanism has been proposed also for the

perception and motor timing tasks, several research findings challenge that

notion, by pointing out differences in timing accuracy for motor timing and

discrimination timing tasks. Bangert, Reuter-Lorenz, & Seidler (2011)

compared timing variability between a motor timing and a discrimination

12

task and found that the coefficients of variation, (CV) values, and Weber

fractions were larger for discrimination than for motor timing tasks. Also,

tap latency increased with increasing duration for motor timing, but

remained constant for discrimination taks. Finally, no between-task

correlations for CV were found. All these results are suggestive of a non-

dedicated timing mechanism across the two tasks. Given the different nature

of the discrimination and reproduction tasks, it seems reasonable that the

patterns of performance variability are rather different between these tasks.

Specifically, in a motor timing task, individuals first encode an interval

duration and then via movement they express the target duration. In a

temporal discrimination task, individuals compare only two or more abstract

representations of durations in memory and then produce a non-timed

response to indicate whether or not they match. While it may not seem

surprising that tasks with different response requirements produce different

patterns of timing variability, that finding is in contradiction with the

dedicated models where correlations would have been expected between the

tasks, regardless the differences in sensorimotor processing and response

requirements (Bangert et al., 2011).

Interestingly, recent work has found between-task correlation accompanied

by task-specific differences in timing variability (Merchant, Zarco & Prado,

2008) for durations of 1 s and less (350 ms to 1000 ms). This latter finding is

again suggestive of a non-dedicated timing system. More evidence against

the dedicated models of interval timing comes from studies which have

shown differences in timing performance between different timescales

(Buhusi & Meck, 2005).

Timing at different timescales

Several differences found in the existing models of time processing have also

been attributed to differences between different timescales (Buhusi & Meck,

2005). Indeed, several differences in timing performance have been found

between different timescales. Overall, the research findings agree with the

notion that duration intervals in the hundreds of milliseconds to seconds

range are more dependent on sensory processes, whereas temporal intervals

above the seconds range are more related to cognitive capacities, and are

therefore processed differently from the shorter ones (Buhusi & Μeck, 2005;

Rammsayer, 1999). Studies manipulating cognitive load and attention provide further evidence

for the role of cognitive functions in timing performance above the 1 second

range (Brown, 1997; Fortin & Breton, 1995; Fortin & Rousseau, 1998;

Rammsayer, 1992; 1997; 1999; 2006). On the other hand, the involvement of

13

cognitive capacities is less clear for timing durations in the hundreds of

milliseconds to second range (Grondin & Rammsayer, 2003; Macar,

Grondin, & Casini 1994; Rammsayer & Lima, 1991, Rammsayer & Ulrich,

2005). While some studies point to common mechanism(s) across such tasks

for timing in the hundreds of milliseconds range (Keele et al., 1985; Ivry &

Hazeltine, 1995), other studies have pointed towards different mechanisms

for different timescales (e.g. Rammsayer, 1999).

In addition, evidence from pharmacological studies adds to the distinction

between the neural systems of the hundreds of milliseconds and seconds

range. It has been shown, for example, that benzodiazepines and

remoxipride impair the discrimination ability for intervals above the second

range, while the ability to discriminate intervals in the hundreds of

milliseconds to seconds range remained unaffected. In addition, it has been

shown that ethanol impaired performance in the hundreds of milliseconds to

second range but it dit not affect timing performance above the second range

(Rammsayer, 1997; 1999). According to these results, any pharmacological

treatment that affects cognitive capacities, such as working memory capacity,

interferes with temporal processing of intervals above the second range,

whereas does not affect the intervals below the second scale.

Several neuroimaging studies also provide evidence for the duration-specific

timing mechanisms. Functional magnetic resonance imaging (fMRI) and

positron emission tomography (PET) studies have revealed areas in the brain

that were selectively activated during the hundreds of milliseconds and the

seconds range. For instance, Lewis and Miall (2003a) reported specific brain

areas that were more strongly activated during a 0.6 s temporal

discrimination task, compared to a 3 s discrimination task. Rao et al (1997),

found that the selective brain areas were more activated when participants

were performing a motor timing task in the hundreds of milliseconds to

seconds range (300 and 600 ms) compared to when they performed a motor

timing task above the 1 sec range. Specifically, Rao et al (1997) showed that

time estimation for the hundreds of milliseconds to seconds range recruited

the primary sensorimotor and supplementary motor areas and the

cerebellum, whereas time estimation above the second range recruited the

posterior dorsolateral parietal frontal cortex (Lewis & Miall, 2003a; 2003b;

2006). Lewis and Miall (2006), showed that when the time interval was

increasing, there was more activation in the dorsolateral prefrontal cortex, a

brain region that is associated with executive functions and working

memory. Consistent with that neuroimaging finding, Brown (1997) reported

greater variability in timing performance when participants had to

simultaneously solve a difficult math problem when the target interval was at

5 seconds, whereas there was no effect at 2 seconds. These results indicate

14

that tasks demanding higher attentional and executive resources might

engage different mechanisms, support for which has been presented by

Miyake, Onishi & Poppel (2004).

Behavioral data might point to the same direction but results are rather

inconclusive. For example, Merchant, Zarco, and Prado (2008), used a wide

range of intervals from 350 to 1000 ms, and found a complex pattern of

correlations between tasks with different response requirements. These tasks

differed in sensorimotor processing (perception–production), in modality

(visual–auditory), and in the number of intervals produced (one–four). The

results appeared somehow mixed. There was a linear increase in timing

variability in accordance to Weber's law, but there was also a strong effect of

sensorimotor processing, the modality used, and the number of intervals, on

participants’ temporal performance. Some temporal tasks were correlated

with each other and others were not, suggesting the existence of a possible

context-specific timing mechanism based on the different requirements of

the tasks.

Task differences

One fundamental question with regard timing in the hundreds of

milliseconds to second range is whether the same mechanism(s) control

timing in different modalities such as the fingers, feet, and arms or whether

different mechanisms are engaged by different modalities. The most

common approach which has been used to answer this question is by testing

the correlations between diferent effectors. If there is some common timing

mechanism used by different effectors, then subjects who perform precisely

with one effector should also perform precise with another. Indeed, several

correlations have been found across motor timing tapping tasks, duration

judgement tasks as well as eye blink classical conditioning tasks. In addition,

significant correlations have been reported between temporal variability in

finger tapping and in tapping with the forearm and the foot (Keele &

Hawkins, 1982). However, the motor demands of these tasks are relatively

small. It seems reasonable that in such tasks, an explicit representation of

time would be important. These results are suggestive of one dedicated clock

involved in probably all the aforementioned tapping tasks. However,

different results have been reported for different kind of tasks that require

for example smooth movements and different movement trajectories such as

continuous drawing. For example, research findings have shown that

variability in a tapping task was not significantly correlated with variability

in a circle-drawing task (Zelaznik, Spencer, & Doffin, 2000; Zelaznik, et al.,

2002). An explanation for these results has been provided by pointing to two

classes of timing processes, the explicit and the implicit timing.

15

An explicit representation of the interval to be timed is used in explicit

timing tasks, such as the tapping tasks, whereas implicit timing is thought to

be an emergent property of the trajectory produced (Turvey, 1977). Explicit

timing is assumed to be a shared process across tasks and relies on a central

dedicated timer (Lorås, Sigmundsson, Talcott, Öhberg, & Stensdotter, 2012).

In contrast, intrinsic timing tasks such as circle drawing are considered to be

continuous and timing is assumed to be regulated on the basis of a different

process, such as the control of movement dynamics. The physical guidance

provided during different tasks can be one way to explain individual

differences in performance variability. For example, individual differences in

the accuracy of synchronizing repetitive discrete movements to an external

pace stimulus can be attributed to the relative efficiency of the matching

process, of participants’ owns beats from the contact to the drum pad. In

general, a different kind of sensory feedback has been considered to be an

important factor with regard performance in motor timing tasks.

The effect of feedback

Motor learning has been proposed to involve three different phases. The first

phase is assumed to be an attention-demanding phase, where learning

progresses really fast and the motor program to be learned is formed. In the

second phase, the motor presentations are further refined, and error

detection/correction mechanisms emerge and sensory afferences of the

ongoing movement are compared with the intended motor output. In a third

phase, movements have become highly automatic (Fitts & Posner, 1967;

Schmidt & Wrisber, 2008). The informational role of feedback (e.g., how the

participants use augmented feedback to correct errors, how error estimation

influences learning), as well as, the motivational aspects of feedback, have

been considered as an important factor for successful motor timing learning

(Schmidt & Wrisber, 2008; Wulf, 2010).

In the dedicated timing models, such as the Wing and Kristofferson’s

(1973) model, it is assumed that timing is performed in an open-loop

manner which assumes a negligible role of feedback and error correction

clearly present during synchronization (Studenka & Zelaznik, 2011).

However, several findings have shown that time keeping might not be

attributable to an open-loop, central clock-like timing process. For example,

there are some observations showing that sensory feedback can affect

accuracy as well as variability in timing tasks (Billon et al. 1996; Chen, Repp,

& Patel, 2002; Drewing, Hennings & Aschersleben, 2002; Kolers &

Brewster, 1985). These findings contradict the dedicated, open-loop

mechanisms described above, suggesting that a closed loop mechanism

16

(where each interval is made with reference to the previous one) seems more

plausible (Madison, & Delignières, 2009).

The intrinsic models of timing

Intrinsic models suggest that temporal information may be encoded in a

more flexible way, depending on the task demands (Buonomano &

Merzenich 1995; Buonomano 2000). In contrast to the dedicated models of

timing, no specialized mechanism is needed and time estimation relies on

local and general multimodal neural activity.

In a motor timing task, for example, as the one we have used in our studies,

the rhythmic responses produced by the individual are assumed to emerge

not from a dedicated clock, but from the continuous activation and

deactivation of the signals that control the consecutive actions. Since the

neural firing is distributed in the brain, it means that timing of an auditory

stimulus, for example, would be represented by changes in the neural

dynamics of the auditory cortex, whereas timing a visual stimulus would be

represented by changes in the neural dynamics in the visual cortex (Spencer

& Ivry, 2009). These models imply no explicit or linear measure of time, like

the ticks, as in the dedicated approach. Instead, a general-purpose network

of neurons can process time and time co-exists with the processing of other

sensory properties of that stimuli.

In some models of intrinsic timing, the representations of timing are

inherently interdependent. For example, one form of intrinsic timing

proposes that duration is encoded in the magnitude of neural activity. In

other words, the amount of energy spent over time (cognitive and emotional

effort) would be the basis for the subjective experience of time. As such,

there will be a bias to perceive a stimulus of a fixed duration as longer, if it is

brighter or louder (Eagleman, 2008; Pariyadath & Eagleman, 2007).

Other models suggest that timing of isochronous sounds is represented in

neuromagnetic beta oscillations (Fujioka, Trainor, Large & Ross, 2012). This

hypothesis proposes that beta power oscillations serve as a transporter for

temporal information during motor timing tasks. Higher neural oscillation

beta power is hypothesized to be associated with longer temporal durations,

whereas lower neural oscillation power is hypothesized to be associated with

shorter temporal durations. Through this change in dynamics of neural

oscillations, it becomes possible to match the temporal scales of perceptual

phenomena (Buzsáki & Draguhn, 2004). One mechanism which has been

proposed to underlie these changes in dynamics of neural oscillations is the

synaptic plasticity in cortico-striatum connections. That is an interesting

17

finding given that the striato-thalamo-cortical circuit has been suggested to

be involved in coincidence detection of oscillatory activities for temporal

coding (Matell & Meck, 2004).

An alternative form of intrinsic models suggests that the precise

representation of timing arises as a result of patterns of activity throughout

the brain (Buonomano & Merzenich 1995). A well known intrinsic timing

model of this type is the state-dependent network model (SDN), in which

time is implicitly represented in the synaptic properties or state of a neural

network (Buonomano & Mauk, 1994; Yamazaki & Tanaka, 2005; Karmarkar

& Buonomano, 2007). SDN models propose that any cortical network could

potentially process temporal information, as populations within each region

are capable of processing duration (Fig 4). In other words, time can be

inferred through changes in the pattern of activity over the network. In order

to better understand the functional properties of an SDN model, Karmakar &

Buonomano (2007) provided a simple example: imagine that we have two

auditory tones presented 100 ms apart from each other. When the first tone

arrives, it will initiate a series of synaptic processes resulting in the activation

of various neuron groups. The response to the second one will produce a

different pattern of neural activity, which changed from S0 to S100. The final

state of the network can be informative in that the previous tone was

presented 100 ms before (Karmakar & Buonomano, 2007).

18

Fig 4. Neural models for temporal representation. The figure was adapted

from Ivry and Schlerf (2008, p 2). (a) Cerebellum as a specialized neural

structure to represent temporal information. Some models postulate a

specialized role for the basal ganglia, supplementary motor area or right

prefrontal cortex. (b) A dedicated system could involve activity across a

distributed network of neural regions. (c) In a state-dependent network,

temporal patterns are represented as spatial patterns of activity across a

neural network.

It has been proposed that the intrinsic models of timing may be especially

suited for timing at the hundreds of milliseconds to seconds range scale,

while dedicated timing models may be more suitable for timing behaviors in

the above the seconds range (Ivry & Schlerf, 2008; Mauk & Buonomano,

2004). However, other researchers have raised several concerns with regard

to the feasibility of a state-dependent network over the hundred of

milliseconds timescale (Spencer, Karmarkar & Ivry, 2009). For example,

human studies testing the state-dependent network model of time

19

perception revealed that specific experimental manipulations influenced

only the perception of intervals lasting up to 300 ms, reducing the

applicability of this model to very short temporal intervals (Buonomano,

Bramen & Khodadadifar, 2009). This raises the question of how an intrinsic

model could account for transfer between different modalities. For example,

how could an intrinsic model of timing explain how training on an auditory

duration discrimination task would facilitate performance for judging the

duration of a visual stimulus?

An attempt to combine the dedicated and the intrinsic models is provided by

the Striatal Beat Frequency (SBF) model (Matell & Meck, 2004; Meck,

Penney, & Pouthas, 2008). The SBF model tries to exemplify the neural

mechanisms of timing by the coincidence detection of oscillatory processes

(Matell & Meck, 2004; Lustig, Matell, & Meck, 2005; Harrington

Zimbelman, Hinton, & Rao, 2010). The SBF model involves a set of cortical

neurons (timekeepers) that oscillate at regular, but distinct frequencies. This

vacillatory activity of brain cells in the upper cortex allows a pattern of

activation to occur at each point in time. A specific rhythm characterizes the

activity of each oscillator cell. These activations are assisted by striatal

integrators that combine their information with feedback and form the basis

of interval timing. The prefrontal cortex, striatum, and thalamus (Allman &

Meck, 2012; Buhusi & Meck, 2005; Hinton & Meck, 2004; Meck, 2006a;

2006b) constitutes the hypothesized frontal–striatal system corresponding

to the functional components of the SBF model (Meck, 1996; Meck &

Benson, 2002; Meck et al., 2008; Matell, Meck, & Nicolelis, 2003; Matell &

Meck, 2004; .

Taken together, both the dedicated and the intrinsic models of timing are

well supported but both have limitations that cannot be disregarded. The

main limitation of the dedicated models concerns their inability to explain

modality specific differences in time estimation. On the other hand, intrinsic

models have limited processing capacity which makes them inappropriate

for time estimation in more complex and real life tasks (Maniadakis &

Trahanias, 2014).

A hybrid model has been also proposed as an explanation of how humans

process timing. That model suggests that a core timing system, the cortico-

thalamic-basal ganglia (CTBG), interact with areas that are selectively

engaged based on the task requirements (Buhusi & Meck 2005; Merchant et

al., 2013). Other researchers do not refer to internal clock models or intrinsic

models but they offer a purely cognitive explanation for time perception and

motor timing (Block, 2003; Block, Hancock & Zakay, 2010).

20

In conclusion, the neural basis of time processing in the hundreds of

milliseconds to seconds range remains poorly understood, despite the vast

amount of research within the field. One reason for this could be that the

existing and highly influential timing models do not take into consideration

other factors that are independent of the internal mechanisms. These factors

can be strategies and previous experiences, as well as attention, memory,

and decision processes that have been found to be correlated with timing

performance. Exploration of these factors, and the precise way in which they

are involved in timing, seems to be required in order to advance the existing

timing models.

In general, many different areas of the frontoparietal cortex, such as the

insular cortex, the striatum and the cerebellum, have been implicated to be

involved in temporal processing, as has been shown in neuroimaging studies

(Lewis & Miall, 2003a; 2003b; Wiener, Turkeltaub, & Coslett, 2010).

Moreover, deficits in different brain areas can result in problems with

temporal processing. These findings indicate that in addition to core timing

mechanisms, the whole brain is needed for the accurate processing of time.

Specifically, distributed brain networks that are responsible for cognitive

control also seem to be essential for the successful processing of time. Next,

several findings that provide evidence for the relationship between timing

performance and cognitive capacities will be described.

Cognitive capacities in timing performance

Several studies have shown a relationship between timing variability and

cognitive traits (Deary et al. 2000; Rammsayer, 2011; Ullén et al. 2012;

2015). Growing evidence from dual-task studies suggests that executive

resources are recruited during timing tasks, above 1 second (Brown, 2006;

Ogden et al., 2011). Brown (1997), in a meta-analytic review article, showed

that several timing tasks exhibit an interference effect when non-temporal

tasks disrupted timing performance. Specifically, the time responses were

shorter, more variable, or more inaccurate when the individuals had to

perform a timing task simultaneously with a non-temporal task.

Furthermore, behavioral data have indicated a strong negative relationship

between reaction time (RT) and various elementary cognitive tasks, as well

as, psychometric intelligence (Jensen, 2006).

As it was described earlier in the introduction, the involvement of cognitive

capacities in timing has been shown to vary with the timed duration. For

example, Miyake, Onishi, & Poppel (2004) showed that participants’ ability

21

to synchronize to a metronome while they performed a memory task was

more impaired at longer intervals, specifically at intervals above two

seconds, compared to shorter intervals. A general finding is that in the range

from approximately 350 ms to 1 s, time perception and motor timing

variability has been found to be linear with time (Getty 1975; Zarco et al.,

2009), whereas above 1 second, there are several non-linearities in the

relationship between interval duration and performance (Madison, 2001).

Madison’s findings are in line with previous research findings (Karmarkar &

Buonomano, 2007; Lewis & Miall, 2006; Maes, Wanderley & Palmer, 2015;

Rammsayer, 1999; Rammsayer & Troche, 2014). In general, it has been

proposed that durations between 0.45 and 1.5 s are automatically processed,

whereas processing of intervals in the range from 1.8 to 3.6 s is affected by

attention and working memory processes (Miyake et al., 2004).

Other researchers, have also shown the role of attention and executive

control in motor timing tasks by using dual-task paradigms. The results from

a recent study suggested that attention and executive control was involved

when participants performed a rhythmic timing task in a slow tempo

compared to when the tempo was faster. The tempo of the sequences ranged

from 600 ms to 3000 ms between each beat. The distractor task was a novel

covert n-back task. A fast tempo was defined as an interstimulus interval

(ISI) shorter than 1500 ms and a slow tempo as an ISI longer than 1500 ms.

The largest increase in interference occurred in two different time ranges,

specifically, between an ISI of 897 and 1342 ms and between an ISI of 200

and 300 ms. This is consistent with models of time perception that assume

that different timing mechanisms are recruited at different time scales

(Bååth, Tjøstheim & Lingonblad, 2016).

Although the effect of cognitive control has been evident for the longer

durations, it remains unclear whether the same holds for voluntary timing in

the hundreds of milliseconds to seconds range. Some studies using dual

tasks paradigms provided support for cognitive control also in that range

(Maes, Wanderley & Palmer, 2015; McFarland & Ashton, 1978; Sergent,

Hellige & Cherry, 1993) In healthy individuals, disruptions of attention have

been found to affect perception of very brief durations in the hundreds of

milliseconds range, as well as estimation of longer durations, in the seconds

range (Casini & Macar, 1997; Grondin & Macar, 1992). Kee et al. (1986)

reported that solving anagrams and thinking out loud simultaneously led to

more timing variability compared to when subjects performed a single task.

Another study found that variability in finger tapping increased when

participants performed a recognition memory task simultaneously

(McFarland & Ashton, 1978). On the other hand, other researchers have

reported limited effects on timing variability as a result of dual task

22

interference on finger tapping (Michon, 1966; Nagasaki, 1990) in the

hundreds of milliseconds to second range. In addition, another study found

no reliable evidence for cognitive control on repetitive motor timing

variability in the range from 0.5 to 2.0 s. (Holm et al. 2013).

In general, it has been shown that the ability to accurately estimate and

produce the duration of a stimulus depends upon processes of sustained

attention and working memory capacity, and that timing and other cognitive

processes are known to share brain networks (Gómez, Molero, Atakan, &

Ortuño, 2014). Therefore, many researchers attribute the interference effect

in terms of a competition for attentional resources (e.g., Brown, 1985; Brown

& Bennette, 2002; Thomas & Weaver, 1975; Zakay, 1989) which leads to a

mutual deterioration of timing performance.

Furthermore, psychometric intelligence has been found to be correlated with

reaction time tasks (Rammsayer & Brandler 2002; 2007) as well as with

performance in time discrimination and production tasks (Deary, 2000;

Jensen, 2006, Madison et al., 2009; Ullén et al., 2008). In addition,

intelligence and time production performance seem to share a common

neural basis indicated by correlations found between self-paced tapping

performance and regional white matter volume in the prefrontal lobes (Ullén

et al., 2008). So far, the source of the associations between intelligence and

timing accuracy have been attributed to bottom-up mechanisms, meaning

that timing accuracy is under the influence of the same low-level processes

that also provide the foundation for the cognitive processes involved in

intelligence (Ullén et al., 2008; Ullén & Madison, 2009). Another source of

associations could be due to individual differences in cognitive capacities, for

example, sustained attention. Indeed, this explanation seems plausible given

that several research findings have shown that reaction time tasks (RT)

depend on attention, which in turn is correlated with intelligence (Schweizer,

Moosbrugger, & Goldhammer, 2005; Schweizer & Moosbrugge, 2004).

Below, several research findings that support the role of attention in timing

performance will be presented.

The role of attention in timing performance

Attention has been found to be highly related with our subjective and

objective experience of timing (Brown 1997; Zakay & Block 2004). For

example, time intervals are judged to be longer when we pay more attention

to time or when we are involved in pleasant activities, but time seems

prolonged during periods of boredom. Attention has also been defined as an

important concept in many models of time perception and time production,

23

including also the cognitive ones in which no reference is made to a

dedicated timing mechanism (Block, 2003; Block et al., 2010).

An extensive literature provides evidence for the hypothesis that attention

plays a crucial role in the estimation of time durations (James, 1950; Mattes

& Ulrich, 1998). Temporal productions have been found to be substantially

more variable under dual-task conditions compared with single-task

(timing-only) conditions. Other studies have also shown that temporal

productions were longer for dual task compared to single task conditions

(Brown, 1997; Miyake et al., 2004). Specifically, longer intervals

corresponded to an underestimation of time. According to an attentional

resource account, these effects occur because a concurrent distractor task

disrupts the accumulation of temporal cues.

Overall, a central idea in research on time performance is that when more

attention is directed towards a nontemporal task, attention towards the

timing task is reduced. It has been shown, for instance, that timing accuracy

was decreased when participants performed a timing task with the presence

of a concurrent task compared with a single task timing condition (Brown,

2006; 2008; Brown & Merchant, 2007). In the same vein, the results of

Chaston and Kingstone (2004) suggested that as attentional demands for a

search task increased, the estimated time duration spent searching

decreased. Lamotte, Izaute, and Droit-Volet (2012) suggested that the source

of underestimation of time durations was the higher attentional resources

demanded for the dual tasks. In line with this, Woehrle and Magliano (2012)

found differences in timing variability between participants with high

working-memory capacities and low working memory capacities.

Specifically, individuals with greater working memory capacity estimated the

length of the timing task as shorter than those with low working-memory

capacities. These differences could be due to those with higher working-

memory capacities directing more attention to the cognitive task, which

resulted in less accurate time estimations (Woehrle & Magliano, 2012).

Overall, the results of this study suggest that the amount of attention

directed toward a stimulus is reciprocally related to the accuracy of time

estimation.

Zélanti and Droit-Volet (2012) found that adults were more accurate in

perceiving time durations compared to children. The same researchers

reported that time perception was less accurate for visual than for auditory

stimuli for both adults and children, while the magnitude of this difference

was larger for the children. An explanation for these observations could be

that attention is involved in the accuracy of time processing. Children, for

whom attentional abilities are considered to be not fully developed,

24

performed poorly in the time perception of visual stimuli as a result of visual

signals requiring more attentional resources than the ones needed for the

processing of auditory signalas.

In line with this behavioral evidence, neuroimaging studies have shown that

attention can modulate time estimation accuracy and brain activity for time

estimation Increasing attention to time selectively increased activity in the

corticostriatal network, including pre–supplementary motor area and right

frontal areas, specifically when the task required a greater degree of

attentional resources, defining this as the core neuroanatomical substrates of

timing behavior (Coull, Vidal, Nazarian & Macar, 2004). Evidence for the

critical role of attention in time perception and motor timing comes from

various clinical populations where a deficit in time accuracy co exists with

distortions in attentional capacity. Below, some of the clinical cases will be

presented.

Impaired timing in clinical populations with an attention disorder

Evidence from clinical populations also supports the role of attention in time

perception and motor timing. It has been shown, for example, that autistic

children or children diagnosed with Attentional Deficit Hyperactivity

Disorder (ADHD) exhibit mixed problems in both motor control and

attentional capacity. These children face problems not only in the execution

and control of cognitive tasks, but also in motor coordination (Kaplan,

Wilson, Dewey & Crawford, 1998; Piek, Pitcher & Hay, 1999). Indeed, several

investigators have shown strong relationships between attention and aspects

of motor regulation, including inhibition, speed, and rhythmicity (Barkley &

Biederman, 1997; Denckla et al., 1985). Valera and colleagues (2010)

performed a neuroimaging study where they compared 20 unmedicated

adults with ADHD and 19 control subjects comparable on age, sex, and

estimated IQ, by using a finger-tapping paradigm. What they found was an

atypical pattern of neural activity in the cerebellum and basal ganglia areas

that have been considered important for motor timing. A common finding

was that individuals with ADHD were more inaccurate in temporal duration

judgments across the hundreds of milliseconds to seconds range, for both

visual and auditory stimuli. Specifically, ADHD individuals tended to

underestimate durations, and their judgements were more variable than

those of typically developing children (Allman & Meck, 2012).

Variable or inaccurate time judgments have also been reported in patients

with damage in the frontal lobes. These timing problems have been related

to the involvement of the frontal lobes in working memory and attention

25

processes. Nichelli, Always, & Grafman, (1996) found that patients with

frontal lesions, and specifically patients who suffered from traumatic brain

injury (TBI), judged short intervals as more similar to the long standard, at

two time ranges (100–900 ms and 8–32 s), a pattern related to problems

with attention, working memory, and processing speed mechanisms.

In another study, depressed patients had difficulties to estimate time

durations (Gualtieri, Johnson & Benedict, 2006). It is well known that

individuals suffering from depression face difficulties in terms of vigilance,

as well as sustained and selective attention, which affects the processing of

durations and especially the longer ones. Therefore, these deficits in the

processing of interval durations had been attributed to their limited

attentional resources or working memory capacity, which led to more

variability in the timing tasks (Gualtieri, Johnson & Benedict, 2006).

All these findings have challenged the traditional clock models by pointing

out their weakness in explaining the effect of cognitive functions in the

estimation of time. Thus, other researchers have proposed models for timing

which have scalar-timing properties but also consider the role of cognitive

factors, such as attention and memory.

Attentional models of timing

Attentional models for time perception and estimation propose that the

existence of non–temporal stimuli can affect attentional processes.

Specifically, they suggest that when more attention is given to non-temporal

information processing, less attentional resources are allocated to temporal

processing. In turn, less attention to timing tasks results in more timing

variability. Specifically, they propose that estimating time intervals that

involve more non-temporal information processing result in greater

problems estimating time, compared to estimating intervals when there is no

attention directed to a task.

The first attentional model of time proposed by Thomas and Weaver (1975)

was for very short (less than 100 ms) durations. It was later extended it to

longer durations by Zakay (1989). The first assumption of the model is that

organisms draw attentional resources from the same pool and distribute

them between all tasks. Consequently, when more attention is needed to

process a non- temporal task, fewer resources are available for the temporal

one, and temporal performance is impaired. Moreover, considering that

subjective duration depends on the number of accumulated pulses, as

proposed in the model of Gibbon et al. (1984), the second assumption is that

fewer pulses are accumulated and the duration is perceived as shorter when

26

the temporal processor cannot function optimally. This hypothesis explains

why durations seem shorter when more attentional resources are allocated to

non-temporal tasks. In this model, attention is supposed to influence the

gate opening, but other hypotheses have also been proposed. Casini and

Macar (1997) suggested that attentional disruption during a target interval

might have two different effects, either stopping pulse accumulation by

opening the switch or affecting the rate of the pacemaker.

A more advanced type of this kind of model is the Attentional Gate Model

(Zakay & Block, 1995). The gate is considered to be a cognitive mechanism

regulated by the distribution of attention to time. If more attention is

dispensed, the gate opens wider and more pulses can pass through and be

transferred to the accumulator.

Fig 4. The Attentional Gate Model (AGM). Adapted from Zakay & Block,

1995.

The attentional gate model by Zakay and Block (1995) used properties from

different models of time perception to explain the relationship between

attentional allocation and time estimation. This model combines most

internal clock theories by suggesting that the mind counts pulses, but also

adds that an "attentional gate" controls how many pulses can enter the

cognitive counter. The proposed model has been widely used in explaining

human timing behavior. Nevertheless, the model must be validated

27

empirically by testing both scalar properties of human-timing behavior and

the necessary properties of the attentional gate.

Lejeune (1998) introduced a similar model, also based on attention. In this

model a dynamic switch is controlled by the attentional resources allocated

for timing. The switch is opened and closed at a frequency dependent on the

amount of attentional resources allocated for timing. When there are more

attentional resources, the frequency and the number of pulses that can enter

and be accumulated in the cognitive counter increase (Zakay, 2000). The

attentional gate, as well as, the dynamic-switch described in the previous

models serve as mechanisms for the control of attentional resources between

timing- and non-temporal tasks. A common ground in both models is that

duration intervals will be judged as longer when non-temporal tasks are

simple and non-demanding compared to when non-temporal tasks are

complex and demanding. A difference between the two models is that in the

Attentional Gate Model, a gate determines the flow of pulses when attending

to time, while in Lejeune’s model the switch is associated with attending to a

duration-onset signal.

The attentional models of time perception are also in symphony with Mihaly

Csíkszentmihályi's theory of flow (1990). This theory considers what

happens when an individual is fully absorbed in the activity at hand. These

activities are usually internally motivated, meaning that the activity itself

rather than the result of it offers satisfaction (e.g. playing a sport or

painting). This intense concentration in activities results in a distortion of

time perception, leading to underestimating the passage of time (Nakamura

& Csikszentmihalyi, 2002). The flow theory suggests that the most important

factor that makes people underestimate time perception is the focused

attention toward an activity that also involves non-temporal information.

Specifically, they suggest that processing 0f non -temporal information can

take away attentional resources from estimating time accurately

(Csíkszentmihályi, 1990).

A similar procedure described in the models above seem to be also involved

in a motor timing task such as the one we have also employed in our studies.

Specifically, subjects must accumulate and store pulses while accumulating

new ones and then make the comparison of these pulses in memory (Bangert

et al., 2011). It seems reasonable that variability in duration reproductions

could be due to variable representations of the target duration in memory

and attention (Harrington & Haaland, & Hermanowicz, 1998). That could

account for the greater timing variability obtained by several clinical

populations with a clear deficit in memory or in attention, such as patients

28

with traumatic brain injury. A failure to fully attend to temporal information

will lead to inaccurate timing performance.

Concluding remarks

Despite the explosion of research on time perception and motor timing,

there are still critical questions that need to be resolved. Firstly, it seems

interesting how reasonable it is to believe that a timing mechanism applies to

a large interval durations when so many violations have been observed. Now,

that we also know that other external factors, such as previous experience,

the context, and the modality can affect timing performance, it seems more

important than ever to test the relative importance of these factors and how

they may vary with interval duration. Secondly, given the fact that several

studies have shown a relationship between timing accuracy and different

cognitive capacities it seems crucial to study the role of cognition in timing

performance.

The Main objectives of the Thesis

The overall aim of this thesis was to explore the nature of temporal

processing in motor timing in the hundreds of milliseconds to seconds range.

Is timing managed by one dedicated system or is it generated intrinsically

based at the different demands of the different timing tasks? This general

aim was broken down into three specific aims. The first aim was to study the

effects of short-term training in a motor timing task. The second aim was to

study the relationship between motor timing and cognition. Specifically, the

transfer gains and differences between a control and a treatment group, in a

sustained attention task, measurements of fluid intelligence and a motor

timing task were tested. The third aim was to study the effects of using

different means to express timed behavior. Specifically, variations across

distinctly different effectors such as the hand and ocular motor system were

tested.

The specific research questions were:

1. Can timing variability in a simple motor timing task be improved

with practice? How do these training effects depend on a number of

key variables, such as the duration of the interval to be timed, the

auditory feedback and the effector used to produce the intervals?

2. Does motor timing training improve performance in cognitive tasks,

such as sustained attention and measurements of fluid intelligence?

29

3. Is motor timing governed by one dedicated mechanism or by several

different mechanisms, possibly invoked by different effectors used to

perform the timing task?

METHODS

A schematic overview of the three different studies is presented in Table 1. In

all studies the synchronization-continuation timing task was employed. In

Study II, we also used a sensorimotor synchronization timing task, as well as

a sustained attention task and two measurements of fluid intelligence.

Common to all studies was that written informed consent was obtained in

accordance with the Declaration of Helsinki, and all studies were approved

by the Regional Ethics Committee at Umeå University. Participation was

voluntary.

Participants

All participants in three studies were students from Umeå University. They

were recruited via flyers at the Umeå University campus and through an

internet Facebook group. Exclusion criteria were report of neurological

disorders as well as being a professional musician. All participants gave their

informed consent to participate.

In Study I, in the first experiment, eight right-handed participants (5 men, 3

women) between 25 to 36 years participated. In the second experiment ten

right-handed participants between 22 and 41 years participated, none of

whom had participated in experiment 1. All participants were paid the

equivalent of 8 Euro per session. In study II, forty right-handed students (16

men, 24 women) between 18-35 years participated. In study III, sixty

participants (31 men, 29 women) between 20-35 years participated. Fifty-

nine participants were right-handed according to a self-report and fifty-nine

participants were right-eye dominant as determined by the Convergence

near-point test (Chen et al., 2000).

30

Table 1. A schematic overview of the three studies.

Study 1 Study 2 Study 3

Purpose To study the effects of training on timing variability in a synchronization-continuation tapping task. Of additional interest was if the training effects depended on the interval duration, the effector used to perform the task and the feedback.

To study transfer effects between a

motor timing task, a sustained

attention task, as well as,

measurements of fluid intelligence. To

this end, we compared timing

performance between a control group,

which received no training, and a

training group, which received motor

timing training

To test whether voluntary milliseconds

to seconds timing is managed by a

single or by multiple effector-dependent

timing mechanisms. To this end we

compared timing variability produced

by two different effectors, the eyes and

the fingers.

Participants Students from Umeå University.

N=(18)

Students from Umeå University.

N= (40)

Students from Umeå University

N=(60)

Tasks Synchronization-Continuation paradigm with 4 diferent responsemodes (Finger, Lightbeam, Drumstick, Cuff) with and without auditory feedback

Synchronization-Continuation

paradigm with a drumstick

Sensorimotor-Synchronization Task

with a drumstick

Corners’ Continuous Performance Task

II (CPT II)

Block Design

Figure Weights

Synchronization-Continuation paradigm

on a Synthesizer

Eye tracking

Analysis Repeated measurements of Anova Repeated measurements of Anova

Pearson’s correlation coifficients

Dependent sample t- test

Repeated measurements of Anova

Pearson’s Correlation coifficients

Slope Analysis

Exploratory Factor Analysis

31

Instruments-Procedure

Synchronization-Continuation paradigm (Study I, II, III)

In this task subjects had to synchronize their beats with external rhythmic

stimuli, and then continue producing those intervals without hearing the

external pacing signal. Specifically, each trial consisted of synchronization to

40 stimuli followed immediately by the production of 200 unsupported

intervals. The stimuli interval (ISI) in a trial were: 524, 733, 1,024, or 1,431

ms with two replications of each ISI. The tapping task lasted about 35

minutes.

An alesis DM5 sound module produced the stimulus sounds which were

presented through a pair of Peltor HTB7A headphones. The rhythmic stimuli

the participants synchronized with, consisted of a repeated sound sampled

from a pair of claves. Sounds were presented at 84 dBA.

Each participant was tested individually, sitting upright on a chair with the

feet on the floor. A computer was at the left side of the participant and in

front was the drum pad. The computer generated a warning signal if the

mean response inter-onset-interval (IOI) across the last 6 responses was

shorter than 66% of the stimulus IOI or if the stimulus IOI had elapsed since

the last response more than three times

In Study I, each participant made responses using four response modes

(Finger, Lightbeam, Drumstick and Cuff), both with and without auditory

feedback. In Study II, the responses were made with a drumstick. In Study

III, the Synchronization- Continutation paradigm was employed on a

synthesizer. The participants’ task was to synchronize their finger presses on

each of three keys on the keyboard, to the sounds presented from a

loudspeaker for 16 times. Then they continued making 39 key presses at the

same frequency and pattern as during the synchronization phase. Responses

were given by pressing keys on a synthesizer keyboard. In study III, an eye

tracking system was also employed. In those kind of tasks the accuracy

(mean distance of taps from the beat) and precision (variability of tap

position relative to the beat) of the motor movement can provide valid

information about participants’ timing performance. This type of task is

considered to be an automatic rather than a cognitive timing task according

to the classification criteria of automatic and cognitive temporal processing

provided by Lewis and Miall (2003a). One reason is that, in this type of type

of task there are no explicit instructions to involve working memory, as in

the time discrimination and time judgement tasks for example.

32

Sensorimotor synchronization task (Study II)

In the sensorimotor synchronization task, participants had to synchronize

drumstick beats to external rhythmic stimuli. Participants were seated

upright on a chair in front of a PC running the Windows 7 operating system

and an LCD screen with a resolution of 1920x1080 pixels. A drum pad was

positioned at the right of the participant, at a convenient position for beating

with the right hand. The metronome sounds were produced by an Arduino

Uno 1.6.3 microcomputer and consisted of a sine tone lasting 50 ms. The

inter-stimulus-interval (ISI) of the metronome was 1024 ms. The beats were

recorded and time stamped by the Arduino in real time and sent for further

processing to an in-house developed matlab code running in the Matlab

2014b environment. Participants received feedback simultaneously with

their beats via the matlab code on the PC. The basic data that can be

extracted from this task are the asynchronies which can be defined as the

difference between the time of a tap and the time of the time onset in the

external rhythm. The standard deviation of asynchronies (SDasy) is an index

of stability (Repp, 2013).

Cognitive measurements (Study II)

Block design

In the Block Design task, participants were shown a pattern and they had to

reproduce the presented pattern using red and white colored blocks. Scores

were calculated based on speed and accuracy, with a maximum score of 51

(WAIS- IV).

Figure Weights

In the Figure Weights, the participant viewed scales with missing weights

and they had to select the weights out of several options to balance the

scales. Scores were calculated based on the number of correct answers with a

maximum score of 27 (WAIS -IV).

Sustained Attention Task (CPT II)

Conners’ Continuous Performance Test II (CPT II) was used for the

measurement of sustained attention. The CPT II task was computerized on a

33

PC running the Windows 7 operating system. Participants had to press the

space bar or click the mouse button when any letter except the target letter

‘’X’’ was presented on a screen. The inter-stimulus-intervals (ISIs) between

letters were 1, 2, and 4 seconds in the same condition. There were 6 blocks,

with 60 trials each. There were 54 targets ‘nonX’ (90%) and 6 non-targets ‘X’

in one block (10%). Target order was randomly permuted in each block. The

CPT II took fourteen minutes to complete.

Eye tracking (Study III)

An Eyelink 1 eye tracker (SR research) recorded participants’ eye movements

at 250 Hz. The visual stimulus material was a mid grey image consisting of

three black dot forming an equilateral triangle with its base rotated

approximately 15 degrees clockwise.

Participants’ task was to synchronize their eye movements to sound pulses

presented from a loud speaker. Furthermore, they had to move their eyes in

a clockwise pattern between the dots at the same rate as the sound pulses.

The sound was switched off after 16 sound pulses, and the participants had

to continue making eye movements at the same frequency and pattern as

they did when they synchronized to the sounds. The trial was ended after 40

times. The same sound stimulus order and the same intervals 524 ms,

733ms, 1024 ms, 1431 ms were used in the key pressing and eye movement

conditions, which were alternated across participant (For a more detailed

description of the design see Article 2).

Dependent measurements computation

SD, CV, Local and Drift Variability (Study I, II, III).

Apart from the standard deviation (SD), and the coefficient of variation (CV)

two additional measurements were used in order to calculate the total

variability: Local, a measure of interval-to interval variability that was

calculated for each trial as follows:

22

1

2

1

2

2

N

xx

xLocal

N

ii

where xi is the duration of the temporal interval between response i and

response i + 1, x is the mean of all intervals of the trial and N is the number

of intervals in a trial. The expression inside the square root is a variance

34

measure based on lag 2 local differences between data points, which is

further divided by the mean duration of all intervals in the trial (to make

Local comparable to a coefficient of variation and inter-onset-intervals), and

was computed as the difference between total variability and Local:

22

1

2

1

2

22

N

xx

xDrift

N

ii

where variables are notated as for Local;

σ2 refers to the total variance in a trial.

Detectability (Study II)

As a measure of sustained attention Discriminability (d') was used, which

was calculated as the difference between the omissions (non X) and

commissions (X) distributions. d' = Z(hit rate) − Z(false alarm rate). A hit

corresponded to not responding in the presence of the X target and a false

alarm corresponds to responding in the absence of a target.

Saccade onsets (Study III)

For the eye task, saccade onsets were used to define the timing of the eye

movements. A saccade onset was defined as a change in position exceeding

15° across a four-sample window (16 ms), or the time at which the eye

velocity became greater than 35° s–1, or reached an acceleration greater than

9500° s–2. The nearest vector angle to the angle between the different points

in the triangle determined the target of each saccade. Since the triangle was

equidistant, the angle between each point of the triangle was 60°.

Statistical Analysis

Different statistical tools were used based on the specific research design in

the three studies. However, a common statistical method that was employed

in all threes studies was the repeated measurements of ANOVA. In study I,

in the first experiment, a 6 Practice × 4 Response Mode × 2 Feedback × 3

Inter-onset-interval (IOI) repeated measurements of ANOVA was employed

to investigate the effects of practice on timing performance. This analysis

resulted in a full factorial design with 144 conditions. Conditions were

blocked with respect to Response Mode and the order was rotated within

each session. The order of IOI and Feedback conditions were randomly

varied within each Response Mode block. In the second experiment of study

35

I, the number of intervals increased to 18. This resulted in an 18 IOI × 20

Practice design with 360 conditions repeated measurements of ANOVA.

In study II, a repeated measurements of ANOVA was employed to test the

effects of training in the control and in the treatment group, using training

group as a between factor and session order as a within factor. A dependent

sample t-test was also employed to access the performance difference in the

sensorimotor synchronization task between the first and last session.

Pearson’s product–moment correlation coefficients was computed to access

the rate of improvement between the motor timing task, the sustained

attention task as well as the tasks of fluid intelligence. In study III, a 2

Effector × 4 Interval duration repeated measurements (ANOVA) was

employed with CV as dependent variable to test if there were any mains

effect of interval duration, as well as, any interactions between the eye and

the finger. Pearson’s product–moment correlation coefficients were

computed to assess the relationship between the timing variability across the

two effectors, using the mean standard deviation across the two repetitions

for all eight trials. Furthermore, an exploratory factor analysis across the

eight conditions (correlations) was employed, in order to further indicate the

source of the variances that underlied the results reported.

A slope analysis was also conducted in study III, to partition the time

dependent and the time independent variability for each effector. For the

slope analysis, the two equations proposed by Ivry and Hazeltine (1995) were

used: σ2Total = kD+ c, where σ2 is overall variance, k is the proportionality

constant, D is the duration-dependent variability, and c is a constant

representing timing-independent variability. Furthermore, the squared

version of Equation 1 was also employed, producing the following equation:

σ2 Total = k2 D2 + c where k is the slope, which approximates the Weber

fraction, D2 is the mean subjective duration squared, and the intercept c is a

constant representing the time-independent component of variability. To

this end a linear regression was also performed on the timing variance and

the squared interval duration according to Equations 1 and 2 (Ivry &

Hazeltine, 1995).

Results

Study I

Madison, G., Karampela, O., Ullen, F., & Holm, L. (2013). Effects of

practice on variability in an isochronous serial interval production

task: Asymptotical levels of tapping variability after training are

similar to those of musicians. Acta Psychologica, 1, 119-128.

36

Aim

The overall aim of the study was to explore to what extent temporal accuracy

in a simple motor timing task can be improved with practice. Moreover, we

investigated how training effects depend on a number of key variables such

as the duration of the interval to be timed, the auditory feedback and the

effector used to produce the intervals.

Results

Experiment 1

Local Variability

Local variability decreased in the first 90 min of practice, across all response

modes and reached a stable level at about 4% thereafter. That was

demonstated by employing a 6 Practice × 2 fFedback × 4 Response mode × 3

IOI repeated measurements of ANOVA. There was a main effect of practice

(F5, 35 = 10.09, p < .00001), response mode, (F3, 21 = 23.89, p < 0.0001),

and feedback, (F1, 7 = 10.91, p = .010). There was a trend for a main effect of

IOI (F2, 14 = 3. 63, p=0.54) suggestive of higher rates of variability at the

longer durations. No significant interaction was found between practice and

response mode as well as between response mode and feedback F < 1,

Ηowever, there was a significant interaction between feedback and response

mode, (F7, 3 = 10.3, p = 0.14) indicating that variability in the Lightbeam

condition reduced more compared to the other response modes.

There was also a significant interaction between practice and IOI (F7,

10 = 2.39, p = .017) suggesting that longer IOIs benefited more from practice.

Furthermore, there was a significant interaction between IOI and response

mode, (F7, 6 = 3.24, p = .010), indicating that longer IOIs caused more

variability in the Lightbeam condition compared to the other response

modes. No more significant interactions were found.

Drift variability

Drift Variability decreased across the first 90 min of practice and reached a

stable level at about 0.5% for the Drumstick and Finger. There was a

significant main effect of practice (F3, 35 = 5.19, p = .001) response mode (F3,

21 = 28.16, p = 0.0001), as well as a main effect of feedback (F1,

7 = 51.03, p = .010). There was a trend for a main effect of IOI (F2,

7 = 3.23, p = .070) indicating higher variability at longer IOIs. Finally, a

37

significant interaction between feedback and response mode was found (F7,

3 = 3.81, p = .025).

Experiment 2

Local Variability

There was no effect of practice as it was demonstrated by a 20 Practice × 18

IOI repeated measures ANOVA. All main effects were for F < 1 and the

interaction effects was at F = 1.4.

Drift Variability

There was a trend towards a significant effect of practice, F19, 171 = 1.50, p =

.088, and a significant effect of IOI, F17, 153 = 18.97, p < .0001. The interaction

between practice and IOI was not significant, F = 1.

Given the results of Experiment 1, which showed that most of the learning

occurred over the first hour of practice, we considered it important to

analyze these data separately. Therefore, we analyzed data from the first two

sessions in a 4 Practice × 18 IOI repeated measures ANOVA.

Local and Drift variability across the first 60 min, in four 15 min

blocks

For Local variability, the main effect of practice was now statistically

significant, F3, 27 = 4.24, p < 0.014. There were no effects of IOI, nor any

interaction between practice and IOI, both F's < 1.

For Drift variability, there was no significant effect of Practice, F < 1, but a

significant effect of IOI, F17, 153 = 7.19, p < 0.0001. There was no significant

interaction, F = 1.

Combined results from both experiments

Here, we considered local and standard deviation, (SD) in order to compare

across different studies. Drift was excluded because the effect of training was

not consistent in Expt 1 and not significant in Expt 2 (Table 2).

Performance for the response mode Drumstick was really close in both

experiments indicating high reliability of the results. The smaller decrease in

variability exhibited in Expt 2 (19 vs. ~30%), was likely because the training

spread across a wide range of intervals from 500 to 1624 ms which was less

effective compared to the training on a few close IOIs. As one can notice in

38

Table 2, the decrease in variability was smaller for Local than for SD. That

might be attributable to the fact that SD is reflecting more sources of

variance (such as drift and other sources) therefore, has more room for

improvement. The initial SD, was larger for the Finger than for the

Drumstick, allowing more improvement overall (31.2%), while the

corresponding decrease in Local was limited by the finger production

method (13.4%) as compared to the Drumstick (27.9%). Only the two 500

and 574 ms were considered in Expt 2, to render the results comparable with

500, 536, and 574 ms used in Expt 1.

Table 2.

Summary of comparable practice effects across the two experiments.

Pre

training

(Local

ms)

Post-

training

(Local

ms)

Decrease

(Local %)

Pre-

training

(SD ms)

Post-

training

(SD ms)

Decrease

(SD %)

Expt 1

Finger 26.1 22.6 13.4 36.1 25.2 31.2

Expt 1

Drumstick 18.3 13.2 27.9 18.9 13.5 29.6

Expt 2

Drumstick

500 and

574 ms

IOI

18.0 15.4 14.5 19.5 15.8 19.0

Conclusions

In general, both Local and Drift variability decreased after a short amount of

training and regardless of the amount of feedback provided and the effector

used. Taken together, the response mode Drumstick came closest to

providing a measure of the individual’s lower limit of timing variability. In

addition, the finger and drumstick conditions yielded similar proportional

decrease in variability. Earlier studies on training effects on timing

performance are not so many, as it was mentioned in the introduction. These

studies covered a range of different timing tasks, but only a few have

considered synchronization-continuation tapping. Furthermore, most of

them have compared experts and novices such as musicians and non-

musicians (Billon & Semjen, 1995; Keele et al., 1985). To our knowledge,

39

only one synchronization-continuation tapping study compared the same

participants pre- and post training (Nagasaki, 1990). By employing a direct

comparison of both the absolute standard deviations and the differences

between pre- and post practice, we have gained some interesting results. The

first important result was that that the pre- and post practice differences

were all around a narrow range of about 30 percent, despite the pronounced

differences between the studies.

The substantial learning was found in the first hour of practice and exhibited

no trend for further decrease thereafter. This learning effect, which is likely

to occur within one single experimental session has many implications for

many timing studies that have not taken into account training. If not

carefully monitored, the effects of experimentally induced manipulations

might be confounded with learning.

Given the relationship between cognition and motor timing (e.g, Deary et

al. 2000) an interesting question with regard to the results of this study

would be to understand if training reflected improvements only in motor

precision or in cognitive performance also. Our results could not provide

direct evidence to this question, as we did not find any support for

differences between the shorter and the longer IOIs. Consistent with the

automatic and cognitive distinction of timing proposed by several

researchers (Lewis & Miall, 2003a; 2003b; Rammsayer, 2008), there were

also no differences in variability produced by the Local and the Drift

component.

In general, two approaches have been proposed with regard to the timing-

cognition relationship. The first is a top-down approach which proposes that

cognitive functions, such as memory or sustained attention, can affect timing

performance. The second approach is a bottom-up one, which suggests that

timing performance is independent of cognitive control. For drawing

stronger inferences about the relationship between timing and cognition, it

would be useful to study transfer effects between timing learning and other

cognitive functions. For instance, if training in a timing task also improves

verbal short term memory span, it would suggest that the timing training

effects reflects the improvement of intrinsic timing capacity and not just the

motor pattern involved in the timing task.

Here, it is important to mention that the differences found between different

response modes might be attributable to differences in motor control. For

example, the inertia found to a great extent in the forearm which is involved

in the drumstick movements may have helped to produce more stable

consecutive intervals. On the other hand, finger tapping, may be more

40

susceptible to small errors from muscle tremor. Other factors that have

might influenced timing performance and learning are genetic influences

(Madison et al., 2009b; Ullén et al., 2008), which might have a large effect

given the small sample of these studies.

Study II

Karampela, O., Madison, G., & Holm, L. (manuscript). Motor timing training

improves sustained attention performance.

Aim

The aim of this study was to test the transfer effects between a motor timing

task, a sustained attention task, and measurements of fluid intelligence.

Specifically, we tested if there were differences in transfer gains between a

training group, that received motor training, and a control group that

received any training.

Results

As you can see in Fig. 9, participants in the training group performed better

at the last training session compared to the first one. That was demonstrated

by employing a dependent samples t-test where a significant difference in

the scores for training session one (M = 45.05, SD = 4) and training session

five (M = 37.1, SD = 5), t (18) = 6.27, p < 0.05 was found.

Overall, the training and the control group performed rather equally in all

the cognitive tasks except for the sustained attention task, where the training

group improved more after the motor timing training (Fig 10). That was

employed by a repeated measurement of ANOVA, where a significant main

effect of session was found for the sustained attention task, F1, 38 = 24.910, p

< 0.05, ηp2 = .396, as well as a significant interaction was found between the

session and group, F1, 38 =5.865, p = .0.02, ηp2 = .134. No other main effects

and interactions were found.

With regard to the synchronization-continuation tapping task, there was no

main effect of group F1, 38 = 0.33 p = 0.56, neither a significant effect of

session, F1, 38 = 0.11 2.6 p = 0.56 for the SD measurement. Furthermore, the

results indicated a non-significant interaction between session and group, F1,

38 = 3.63, p = .0.06. With regard the Local measurement there was a

significant main effect of group, F1, 38 = 2.7, p = .010. However, there was

also a significant main effect of session, F1, 38 = 4.06, p = .0.05, as well as a

41

significant interaction between the group and the session, F1, 38 = 5.92, p =

.0.019.

Fig 9. Period synchronization variability as a function of training. Error

bars indicate the SEM.

42

Fig 10. Performance on Sustained Attention (CPT II), as a function of test session for control and training group. Error bars express the SEM.

Conclusions

The aim of this study was to test if cognitive capacities are involved in motor

timing, by assessing transfer effects from a sensorimotor timing task to a

sustained attention task, as well as, measures of fluid intelligence.

Furthermore, we were interested to test if training in the sensorimotor

timing task would enhance performance in a continuation tapping task. The

hypothesis, that the timing trained group would improve more in the

sustained attention task compared to the control group was confirmed.

Furthermore, we found no evidence for timing training improvements on

measures of other types of cognitive ability that we used in the study.

Specifically, the training and the control group improved equally in the Block

design and the Figure Weight task. In addition, we found an improvement in

43

the continuation tapping task after the sensorimotor synchronization

training.

Overall, the results of this study provide support that motor timing in the

milliseconds to seconds range involves cognitive capacities, and specifically

sustained attention. The close match in execution requirements between the

sensorimotor timing task and the sustained attention task could be one

reason for the transfer found between those two tasks. Specifically, the

sensorimotor timing task seems to be a task that involves regulation of

attention over time. The CPT II, also involves regulation of behavior over

time. In other words, it seems very likely that we might have trained

sustained attention and not motor timing ability per se. Therefore, we view

our results as near transfer rather than far transfer.

The present results also suggest that top-down control might be involved in

the relationship between intelligence and motor timing. Specifically, people

who perform better in intelligence tasks could also perform better in the

timing tasks because of better cognitive control capacity. For example, better

focus capacity and less lapses of attention could lead to enhanced

performance in timing tasks (Ullén et al., 2008). That is in line with several

studies which have shown an association between intelligence and sustained

attention (Schweizer, Moosbrugger, & Goldhammer, 2005).

With regard to the transfer effect from the sensorimotor timing task to the

continuation tapping task, one explanation could be that training in the

synchronization task reduced the motor implementation error expressed as

reduced timing variability, both in the sensorimotor synchronization and the

continuation tapping task. Another explanation could be that training in the

sensorimotor synchronization task increased predictive control. The ability

to coordinate with a motoric action an external predictable event, as happens

in the sensorimotor timing task, increased during the sensorimotor

synchronization, affecting at the same time the internal, tap to tap

variability, during the continuation task.

Taken together, after the relationship found between sustained attention and

the motor timing, it seems reasonable to expect an improvement in

sustained attention after intense training in a sensorimotor timing task, at

least, under the experimental conditions we used in our study. That would be

beneficial for several clinical populations, such as, people with ADHD and

people with autism who exhibit deficits both in timing and attentional

capacity.

.

44

Study III

Karampela, O., Holm, L., & Madison G. (2015). Shared timing variability in

eye and finger movements increases with interval duration: Support for a

distributed timing system below and above 1 second. The Quarterly Journal

of Experimental Psychology, 68, 1965-1980.

Aim

The aim of this study was to test differences in timing variability produced by

the eye and the finger so as to check whether motor timing in the hundreds

of milliseconds to seconds range is managed by a single or multiple timing

mechanisms. To test these possibilities, we compared the timing variability

between and within the eye and the finger as a function of the interval to be

produced.

Results

First we computed the accuracy of the sequence reproduction by calculating

the proportion of hits for the fingers and the eyes. Six participants were

excluded as they had really low response hit rates for the eye conditions.

The variability (SD) was increasing monotonically for the eyes and the

fingers for each IOI and the coefficients of variation was flat in accordance

with Weber’s law. There was no significant main effect of interval duration as

it was indicated by a 2 (effector) x 4 (interval duration) repeated measures

ANOVA with CV as dependent variable duration, F3, 159 = 1.48, p = .22. A

significant difference was found for the two effectors of the eye and the

finger, F1, 53 = .859, p < .001, as well as a significant interaction F (3,128)

=11.2, p<0.01, residing in differences between effectors across the interval

durations of 1024 and 1431 ms. Specifically, the CV for the fingers increased,

whereas the CV for the eyes decreased.

Pearson’s product–moment correlation coefficients were computed to test

the correlations between the two different effectors. Table 3 shows all eye

finger correlations for each IOI combination. All four between-effector

correlations were positive, but only those for 1024 and 1431 ms were

statistically significant with Bonferroni correction. Seven different-interval

comparisons were also significant. That was suggestive of a relatively strong

relationship between fingers and eyes when fingers produce the two longer

intervals (1024 and 1431 ms) and relatively low relationships for the lower

intervals. The Equations 1 and 2 (see Statistical Analysis) for the slope

45

analysis were submitted to a 2 (effector) × 4 (interval duration) repeated

measures ANOVA to test whether the slopes of eye and finger variability

differed. There were significant main effects for interval duration, effector

and their interaction for both equations, p<.001. The significant interaction

in both analyses resided in a steeper slope of eye than finger variability

across interval duration.

Table 2. Correlations between the two effectors and all IOIs.

Intervals 524 ms

hand

733 ms

hand

1024 ms

hand

1431 ms

hand

524 ms eye 0.09

0.18 0.46* 0.47*

733 ms eye 0.16

0.16 0.39* 0.34*

1024 ms eye 0.18

0.27 0.36* 0.35*

1431 ms eye 0.32* 0.30* 0.31* 0.32*

Conclusions

The overall aim of this study was to test whether temporal processing in the

hundreds of milliseconds to seconds range is governed by one or by multiple

timing mechanisms. The most important result found in this study were the

several positive correlations found between the two different effectors.

Specifically, individuals who varied more in the finger movements also

varied more in the eye movements and vice versa. That result was consistent

with the idea of a single timing mechanism where no differences between

different effectors should be expected. Several other findings were also

consistent with the notion of a single timing mechanism. For example,

temporal variability in both fingers and eyes followed Weber’s law and the

CV tended to increase for the fingers, consistent with previous studies (e.g.

Madison, 2001).

46

On the other hand, several results were inconsistent with a single timing

mechanism. For example, between effector correlations were substantially

different across IOIs. Furthermore, these correlations were not statistically

significant for intervals below one second after correcting for reliability

finally, the slopes derived from the slope analysis were significantly different

for the eyes and the fingers. An interesting finding was that CV values were

decreasing for the eyes with increasing IOIs. However, that cannot invalidate

the single clock hypothesis as these differences can be due to larger motor

noise dominating eye variability at shorter intervals and hiding possible

correlations between the two effectors at the shorter intervals.

Overall, the pattern of results from this study could not support either the

single clock or the multiple clock hypothesis. Instead, in symphony with

other studies these results proposed that voluntary motor timing is managed

by a partially overlapping distributed timing system. That system could be

responsible for temporal processing for both in the eyes and the fingers but

changing according to the task demands. (Merchant et al., 2008).

47

GENERAL DISCUSSION

The overall aim of this thesis was to explore and increase our knowledge of

the nature of timing processing in the hundreds of milliseconds to seconds

range. To these ends, three areas of specific interest were identified.

Specifically, the studies in this thesis examined whether motor timing

training can influence timing performance and whether motor timing

training can improve performance on a variety of cognitive tasks.

Furthermore, the effect of using different means to express timed behavior

on timing performance was assessed. The discussion will initially be

organized according to these three themes. The effects of training, as well as,

the transfer gains and differences between motor timing and cognitive tasks

will be presented and critically discussed. Furthermore, the findings with

regard modality differences in relationship to timing performance will be

presented and discussed, in light of the existing models for time processing.

Following the discussion of the findings related to these themes, the

limitations of the studies, as well as, implications of the results found will be

presented. Finally, several ideas for future studies will be presented.

Effects of training. Motor and cognitive gains

The results of Study I indicated that participants’ performance reached a

stable plateau after 90 minutes of practice regardless of the response mode

employed. In symphony with study I the results of study II showed a

substantial effect of training after 5 training sessions, also around 90

minutes. The importance of these results lies in that repetitive motor timing

can be improved and reaches a stable plateau after a few training sessions,

even with different intervals. That is an important finding as it is indicative

of motor timing being a skill, which can be learned and improved, and not a

predetermined ability that cannot be changed. These results are on the one

hand in line with previous research, which has shown that practice can

enhance timing performance (Nagarajan et al., 1998; Wright et al., 1997),

while on the other hand, they raise methodological concerns about other

studies that have not controlled for practice and response modes. For

example, different levels of practice among participants might have affected

the validity of the timing performance data in previous studies and thus have

led to incorrect conclusions. Interestingly enough, there is only one study

which has used naïve participants at a pre and post session (Nagasaki, 1990).

All earlier studies testing the training effects on timing performance have

compared experts and novices such as musicians and non-musicians (Billon,

48

& Semjen, 1995; Keele et al., 1985) and they have usually used the finger as

the response mode.

Furthermore, these results raise the interesting question of what property do

we train when we receive training in a motor timing task. Performance

improvements may reside in either improved timing ability or task

adaptation, such as adapting the sensorimotor model to the response

requirements imposed by the task. That is a really interesting question for

researchers dealing with the nature of the relation between timing and

cognition (Helmbold & Rammsayer, 2006; Helmbold, Troche, &

Rammsayer, 2007). The bottom-up account, which suggests that basic

neural properties influence the amount of temporal variability, which is

reflected both in timing tasks and cognitive performance, is strongly

supported so far (Madison et al. 2009; Ullén et al, 2008;2012).

Local variability has been considered an automatic process which is

inaccessible to conscious control. On the other hand, Drift has been

considered to be dependent on short-term memory of previously produced

intervals and it has been considered to be more susceptible to top–down

influences (Forsman et al., 2009; Madison & Delignières, 2009; Madison et

al., 2009b). We hypothesized that if training affected Local and Drift

variability differently, this would be suggestive of which functional parts of

the timing mechanisms are more adoptive to learning. For example, a

decrease in Local variability would suggest that training affected motor

control more (Flanagan, Vetter, Johansson, & Wolpert, 2002), whereas a

decrease in Drift would suggest that training would affect processes related

to short-term memory during several intervals.

Study I could not provide a direct answer to this question. Specifically,

despite our hypotheses that Drift and Local variability would be affected

differently from the training, we found that Drift did not decrease more than

Local, according to the idea that training would mainly affect cognitive

processes. In addition, we used a wide range of durations in order to explore

the differences of the training effect as a possible function of IOI. Lewis and

Miall’s (2003a) classification criteria for the distinction between the

automatic and the cognitive based timing are (1) different psychophysical

characteristics at different durations (several breakpoints above 1 sec), (2)

differential responses to pharmacological manipulations (durations above 1

sec seem to be affected), (3) differential impairment of performance by dual

task (durations above 1 sec are affected) and, (4) different patterns of brain

activation during the measurement of sub-second and supra- second

intervals (the sub-second timing activates more motor circuits whereas the

supra seconds system activates more prefrontal and parietal regions). Study

49

I found no differences in training effects between the longer and shorter

intervals, indicating no dominance of cognitive effects of training. That is in

line with Holm, Ullén and Madison (2013), who did not find an effect on

cognitve load; nor did they find an interaction between cognitive load and

sequence.

The results of study II, however, provided evidence for a direct involvement

of cognitive capacities, especially sustained attention, in repetitive motor

timing. Previous studies dealing with the timing-cognition relationship have

mostly used dual task paradigms where participants perform a timing task

simultaneously with a cognitive task. In study II, we extended previous

research by employing a transfer study. The transfer hypothesis was based

on the condition that executive functions are involved in timing performance

found in the previous dual task studies. Furthermore, transfer enabled us to

draw stronger inferences about which specific cognitive capacity is involved

in the relationship between timing and cognition. The results of study II

could be attributed to the close similarity in the executive demands of the

two tasks we used. Motor timing behavior is a regulated process that involves

attentional control over time. The CPT II, the sustained attention task that

we used, is also a regulation process of sustained behavior over time.

Therefore, it seems reasonable that the close nature of the two tasks led to

near transfer between motor timing and sustained attention. Therefore, it

seems possible that through the motor timing training we have trained the

regulation of motor and cognitive control and specifically attention. Overall,

study II provided support that repetitive motor timing in the hundreds of

milliseconds to seconds range employs regulation and is subject to

controlled cognitive processes and specifically sustained attention. That

agrees with research findings which suggest that measurements of sustained

attention, as well as, working memory capacity seem to control the

sensorimotor and the cognitive demands of a timing task (Block, Hancock, &

Zakay, 2010).

While the involvement of cognitive capacities in longer durations has been

reported by several studies, the involvement of cognitive capacities in shorter

durations remains rather unclear. The presence of cognitive control also in

the hundreds of milliseconds to seconds range seems to suggest that it is the

predictive control demands that determine the involvement of cognitive

control rather than time itself (Holm, et al., 2016). Task difficulty and the

time frame of a planned action can affect anticipatory timing. For example,

when the time frame increases predictive uncertainty also increases, and

therefore more cognitive control is needed. So, the recent view of automatic

versus cognitive timing for intervals below and above 1 sec could be replaced

with the view of at least some cognitive control being present throughout the

milliseconds to seconds timing, possibly increasing with increased durations.

50

Since we did not employ a wide range of durations below 500 ms in our

studies we cannot give much support towards that direction but this is

evident from other studies. Furthermore, that is in line with timing models

which acknowledge the role of attention in successful time processing (Zakay

& Block, 1995). In support of the involvement of cognitive control in motor

timing, research findings suggest a partial overlap of brain regions engaged

in time processing tasks, with regions engaged in tasks requiring increased

cognitive effort. Specifically, timing and other cognitive processes were

found to be interlinked when there was an increase in the level of difficulty

or increased effort (Alustiza et al., 2016). In general cognitive control seems

to be recruited in response to the execution demands of the task employed,

such as cognitive effort demanded by the task difficulty.

The present results shed light on the established relationship between timing

and intelligence. In general, intelligence has been positively correlated with

accuracy of performance in a wide range of timing tasks (Jensen, 2006). This

relationship holds true also for the synchronization-continuation task used

in various studies (Holm, Ullén, & Madison, 2011; 2013). Mainly, this

relation has been attributed to bottom-up mechanisms that are independent

of cognitive control (Holm et al., 2011; Madison et al, 2009; Ullén et al.,

2008). Variability in the isochronous tapping had been considered to be

inaccessible to top-down control and unaffected under dual task conditions

(Holm, et al., 2013). In addition, relations with intelligence are stronger for

Local than for Drift (Ullén et al. 2008; Madison et al., 2009). Consistent with

this, children diagnosed with Attention Deficit Hyperactivity Disorder

(ADHD), who exhibit deficits in executive functions, exhibit larger deviations

in Drift than in Local (Jucaitė et al. 2008). Finally, manipulation of

motivation states did not affect the magnitude of relations between motor

timing and intelligence (Ullén & Madison, 2009) indicating a lack of top-

down control.

However, a general conclusion from study II was that there is also a top-

down relationship between cognitive capacities and motor timing in the

hundreds of milliseconds to seconds range. The timing-intelligence

relationship could therefore be attributed to individual differences in top-

down control and specifically sustained attention, which influence

performance in both timing tasks and intelligence. For example, more

intelligent people could perform better in timed tasks due to better cognitive

control mechanisms which might lead to less lapses of attention (Madison et

al., 2009).

51

That accords with converging results from several studies which have found

that sustained attention and psychometric intelligence are related (Burns,

Nettelbeck, & McPherson, 2009; Ren et al., 2013; Schweizer & Moosbrugger,

2004). For example, several research findings have shown that reaction time

tasks (RT) depend on attention, which in turn is correlated to intelligence

(Schweizer & Moosbrugger, 2004; Schweizer et al. 2005). Further support

derives from the Worst Performance Rule, which states that the worst trials

in reaction timing tasks, which are the trials most likely to be affected by

attentional lapses, are more strongly associated with intelligence, provides

further support for the role of attention on the relationship between

intelligence and timing tasks (Coyle, 2003). In summary, both top-down and

bottom-up processes are likely to be involved in associations between

intelligence and performance in timing tasks.

The involvement of cognitive control in repetitive motor timing is also

indicated by the results of study III. Several positive correlations were found

between the effectors of the eye and the hand of the order of .5 for the two

longer IOIs. That is supported from the increase in correlations between the

two different effectors at the longer IOIs. The lack of significant correlations

in the shorter interval could be due to independent noise dominating at the

hundreds of milliseconds range and covering this effect. On the other hand,

cognitive control might take over in the longer intervals, which may be why

we found stronger correlations for them. These results support that there

might at least be some cognitive control, below one second, which varies

with the task characteristics (Holm, Ullén & Madison, 2013; Holm et al.,

2016).

The behavioral results described in this thesis challenge the traditional

models of processing time in the hundreds of milliseconds and seconds

range, by pointing out their inability to fully accommodate the results of our

studies as well as the results of several others. This pattern of results opens

up to investigation of long-standing issues in complex motor behavior, like

the interplay between different effectors (fingers or hands) as well as the

manipulation of different modalities such as vision and audition and

between different target durations. Moreover, a long-standing question is

whether time is intrinsically encoded in the activity of general purpose

networks of neurons or time is processed through a time-dedicated neural

circuit.

Modality differences – dedicated or intrinsic mechanisms?

A fundamental question regarding timing processing is whether it relies on a

single centralized mechanism or is distributed throughout different areas.

52

The results of study III showed several positive correlations between the

effectors of the eye and the hand, on the order of .5 for the two longer

intervals. In addition, the variability of the eye and the finger movements

increased with the interval duration. Under the assumption of a centralized

clock, no differences between different modalities should be expected.

Therefore, these results could be viewed as a support of the single timing

mechanism hypothesis. In the same vein, others have shown that temporal

training generalized across audition and touch, across different skin

positions (Nagarajan et al., 1998), as well as, across different frequencies of

the stimulus tone (Wright et al., 1997) in support of the centralized clock

hypothesis.

Furthemore, the lack of differences in variability between short and long

intervals (> ~ 1 s,) (Lewis & Miall, 2003a) indicated in study I, was in

accoradance with models of timing that assume one dedicated clock for all

time durations. However, several other results of study I and study III were

inconsistent with a single timing mechanism. For example, cross-effector

correlations were different across duration intervals. Moreover, the slopes of

variability differed markedly for the eyes and the fingers. In addition, in

study I, there were differences in the variability produced by the different

effectors, as well as, differences in the effect of feedback, with the effect of

feedback being more pronounced in the Lightbeam condition. These findings

were suggestive of multiple timing mechanisms operating and they were in

accordance with research findings that have failed to find any transfer

between different modalities. For example, no improvement in performance

found in visual processing after temporal auditory training (Grondin et al.,

2008; Grondin & Ulrich, 2011). The lack of transfer or any training or

association found in these studies, does not simply imply that visual

processing cannot be enhaced by auditory processing. One explanation

accounted for these results could be the moderate amount of training

provided. It has been shown that for interval duration discrimination

extensive training is needed in order to exert a strong effect on performance

(Kristofersson, 1980). Another explanation could be that different timing

mechanisms are involved in different sensory modalities in the

discrimination of brief durations (Rammsayer & Lima, 1991).

The results of study II suggested that cognitive control expressed via

sustained attention directly affected timing performance, as well as, that

training in sensorimotor synchronization produced improvements in a

continuation tapping task. These results are also consistent with the

executive control hypothesis (Krampe, Kliegl & Mayr, 2005), which assumes

that timing and sequencing in paced movement production reflect two

distinct processes: a low level and a high-level component. According to

53

these researchers, during the production of timed sequences, executive

functions program the low-level timing mechanism such updating,

maintenance, and adjustment of the specific action plan. In the same vein,

evidence from recent studies suggest that modulation of attention resources

predicts the speed and accuracy of visual and visuomotor processing

(Klimesch et al. 2008; Serences & Yantis 2006), indicating that the nervous

system’s ability to modulate its representation of time is highly influenced by

the attentional state of the observer. These results act against a dedicated

approach which is unaffected from the computational demands of the tasks.

Taken together, the results of the three studies cannot support unanimously

the dedicated or the distributed timing system. The mixed pattern of results

might be suggestive of an intermediate hypothesis, namely a partially

distributed timing system. These results suggest that timing mechanisms

interact with a distributed brain network which is engaged in all tasks

requiring timing processing, whereas this activation can be altered

depending on task demands. That would be in agreement with Merchant and

colleagues (2008) who also found a mixed pattern of results when analyzing

timing variability produced by four timing tasks that differed in many ways,

such as the duration interval and the the modality used. Timing variability

increased linearly as a function of the interval to be timed in all tasks, in

accordance to the scalar property, but a strong effect of the nontemporal

variables was also found (Merchant et al., 2008).

As proposed by Merchant et al (2013), temporal estimation may depend on

the interaction of multiple areas including regions that conform to the main

core timing network, including the cortico-basal ganglia-thalamus circuit

(CBSTc) and areas that are activated in a context-dependent fashion

(Merchant et al., 2008; 2013). The main core timing network consists of the

supplementary motor area (SMA) and the basal ganglia. The pattern of

results found in study III could also be explained on the properties of that

circuitry. The common system proposed in the CBSTc circuit could be

dedicated to temporal processing for both the eyes and the fingers (given the

similarities found), but other areas could also be activated according to task

incident (given the differences found). Furthermore, a system like that could

explain the results of study II where general, cognitive control mechanisms

may constitute a general purpose resource for controlling timed behavior

and be recruited in response to the computational demand of the task, such

as increased attention.

The Striatal Beat Frequency (SBF) model described in the introduction could

be complementary to the CTBG circuit. The SBF models propose an innate

interactive nature of timing networks. Specifically, SBF model assumes a

54

dedicated timing mechanism in the basal ganglia which monitors distributed

neural activity in the cortex. The mechanisms proposed by the SBF model as

described by Matell and Meck (2004) is constructed such that it is consistent

with brain regions which are thought to be involved in timing (e.g., frontal

cortex and striatum). Furthermore, its output is consistent with physiological

recordings and behavioral results from interval-timing experiments (Matell

et al. 2003). Critically, the SBF model reproduces the scalar property,

(Gibbon et al. 1984, 1997), which occurs because of variability in the firing

patterns of striatal neurons and because cortical activity is assumed to be

oscillatory.

Implications

In general, the results of the three studies support models that propose a

core timing mechanism that has access to multi modal information and is

engaged in the synchronization and the production of timed intervals.

Furthermore, our cross-modal and transfer studies propose that there are

other behavioral parameters that can influence this core timing mechanism

in a specific way, such as higher cognitive components like attention. In

other words, in addition to bottom-up, intrinsic circuitry mechanisms can

affect timing.

Apart from the theoretical contribution of the results of the studies included

in this thesis, they do have many practical implications for both healthy

people and clinical populations. First, they provided valuable information on

how much timing training would someone need to improve in motor timing

and which way would be the most effective. That could be beneficial for

people’s everyday activities such as sports performance, speech production

and a variety of motor activities. Furthermore, the effects of near transfer

found between timing and sustained attention can be beneficial for clinical

population exhibiting attention and motor control problems, such as ADHD

and autism. Future research can test if people with severe problems in

sustained attention capacity could benefit from a motor timing intervention

program that eventually can improve cognitive performance, at least under

the experimental conditions that we employed in our studies.

Given the direct relationship found between sustained attention and

repetitive motor timing, it seems reasonable to expect an improvement in a

number of capacities, including sustained attention and motor control, at

least under the experimental conditions we employed. For example, training

in a sensorimotor synchronization task could facilitate performance on tasks

that require sustained attention and timing accuracy. These results have

55

implications not only for clinical populations but also for people exhibiting

no cognitive or motor deficits. A systematic motor timing training might

improve motor timing and sustained attention in adults and children. The

facilitation of motor and cognitive properties can lead to increased

concentration when performing several tasks that may be important for

academic achievement. In addition, it is possible that other cognitive

capacities can be improved following motor timing training. Based on the

proposed interconnectivity between motor and cognitive functions, it seems

reasonable that designing rehabilitation strategies for people with motor

problems should include a focus also on the cognitive aspects of motor

control and behaviour. Based on the results of our studies it seems crucial

that other training studies with a larger sample of participants should be

performed. Furthermore, the inclusion of various cognitive tasks seem

crucial.

To my knowledge, there are no neurological disorders that are characterized

by pure timing deficits. It is thus difficult to judge whether the observed

timing processing deficits are due to deficits in attention and working

memory since all three processes are known to engage same brain areas,

specifically, the right prefrontal cortex (PFC). Further research is warranted,

since the results of our study cannot provide direct evidence.

Given the peculiar nature of the timesense it is not surprising that there are

so many controversies and so many different models proposed.

Furthermore, neglecting various cognitive processes, which might contribute

to the perception and production of time, can cause several problems and

inconsistencies in identifying the underlying neural mechanisms of timing

(Grondin, 2010).

Limitations

The studies included in the thesis have several limitations that would

motivate further work. A methodological limitation of study I was that a

within-participant design was used. That means that the type of practice was

not varied across groups, but was mixed across the different training

sessions. Thus, although this type of design could address the effects of

training, as well as the effect of a number of different variables, such as the

feedback and the interval used, it could not provide more information about

which specific condition has been more efficient. Another methodological

consideration was the number of the training sessions included, which did

not allow us to get a clear picture of whether the training effect decreased or

flattened out, after the few hours of practice.

56

In study II, in the sensorimotor synchronization training task, participants

received both visual and auditory feedback during the training sessions.

Therefore, it was not possible to distinguish which condition (visual

auditory, or a combination) had been more beneficial for motor timing

learning. In addition, it would have been beneficial, if we had studied, not

only the immediate effects of training, but also the effects of training over

time. That would allow us to draw stronger conclusions about the

sustainability of motor timing training over time. Furthermore, we used a

passive control group, which made it possible to rule out retest effects, but a

question arising here was whether an active control group would have

improved more from pretest to posttest on the transfer task compared to a

passive control group. Having included a control group, with at least a low

level of motor training, it would have made it possible to draw firm

conclusions. Moreover, another limitation of study II, was the limited

number of cognitive tasks included. It would be more informative, if we had

incorporated test batteries that would allow for direct calculation of the g

component, as well as of other cognitive factors. Finally, more cognitive,

motor, as well as, time perception and production tasks should have been

included, in order to allow stronger inferences about the relationship found

between motor timing and cognition.

Given that several different suggestions have been proposed for the

breakpoints above and below 1 second, a limitation of study III seems to be

the limited range of intervals used. The inclusion of a wider range of

intervals would have probably provided important information for the

nature of time processing in the different timescales. Specifically, the

inclusion of intervals below 500 ms, would have made it possible to identify

several sources of motor noise, which are more likely to be reflected in the

short interval durations. It seems that an avenue for advancing our

knowledge about timing mechanisms, would be by using different

modalities, while trying to eliminate independent noise.

Another limitation concerning all the studies was the study population. A

larger number of participants would have increased the statistical power in

the studies. Moreover, the study population was highly homogeneous. This

specific selection could have led to limited generalizability of the results to

the general population, while on the other hand it might have led reduced

within groups’ variance and noise.

Future

Given that there are many time-dependent capacities, it is really surprising

that cognitive research has largely neglected the flow of time. Recent

57

reasearch has luckily been directed towards the temporal aspects of

cognition, by incorporating a range of perceptual and motor skills into time

(Battelli et al., 2008; Taatgen et al., 2007). Overall, in order to expand the

theoretical understanding of timing processes, future research should

investigate in depth the interplay between cognitive control and timing in

many timing tasks. Considering the results of our studies, as well as previous

research, it seems crucial to study how cognitive control affects accuracy in

timing performance. In addition further research is required to more closely

map the relationship between motor timing and intelligence.

Finally, future research should focus on whether time is represented in a

context-dependent manner or in a more context-independent manner in a

dedicated region (or network of regions) that acts as an internal timer. The

new models point towards a joint operation of several mechanisms, with

perhaps a switch from one to the other as the duration to be timed increases

(Lewis and Miall, 2003b; Ivry and Schlerf, 2008). In general, it is important

for future research to study how different brain areas communicate, such as

specific timing circuits, with other brain networks (Cheng, Dyke, McConnell,

& Meck, 2011; Harrington, Castillo, Fong, & Reed, 2011), as well as changes

in the brain, in cognitive and timing processes that occur throughout

childhood and into old age.

58

REFERENCES

Ackermann, H., Graber, S., Hertrich, I., &Daum, I.(1999). Cerebellar contributions

to the perception of temporal cues within the speech and nonspeech

domain. Brain and Language, 67, 228–241.

Allan, L. G., & Gibbon, J. (1991). Human bisection at the geometric mean.

Learning and Motivation, 22, 39-58.

Allan, L. G., & Kristofersson, A. B. (1974). Psychophysical theories of duration

discrimination. Perception & Psychophysics, 16, 26-34.

Allman, M. J., & Meck W. H. (2012). Pathophysiological distortions in time

perception and timed performance. Brain, 135, 656–677.

Alústiza, I., Radula, J., Albajes-Eizagirre, A., Domíngue, M., Aubá, E., & Ortuño, F.

(2016). Meta-analysis of functional neuroimaging and cognitive control

studies in schizophrenia: Preliminary elucidation of a core dysfunctional

timing network. Frontiers in Psychoogy, 7, 192.

Artieda, J., Pastor, M. A., Lacruz, F., & Obeso, J. A. (1992). Temporal

discrimination is abnormal in Parkinson’s disease. Brain, 115, 199–210.

Bangert, A. S., Reuter-Lorenz, P. A., & Seidler, R. D. (2011). Dissecting the clock:

understanding the mechanisms of timing across tasks and temporal

intervals. Acta Psychologica, 136, 20–34.

Barkley, R. A., & Biederman, J. (1997). Towards a broader definition of the age of

onset criterion for attention deficit hyperactivity disorder. Journal of the

American Academy of Child and Adolescent Psychiatry, 36, 1204-1210.

Battelli, L., Walsh, V., Pascual-Leone A., & Cavanagh, P. (2008). The when

pathway explored by lesion studies. Current Opinions in Neurobiology, 18,

120–126.

Billon, M. & Semjen, A. (1995). The timing effects of accent production in

synchronization and continuation tasks performed by musicians and

nonmusicians. Psychological Research, 58, 206-217

Billon, M., Semjen, A., Cole, J., & Gautier, G. (1996). The role of sensory

information in the production of periodic finger tapping sequences.

Experimental Brain Research, 110, 117-130.

59

Block, R. A. (2003). Psychological timing without a timer: The roles of attention

and memory. In H. Helfrich (Ed.), Time and mind II (pp. 41- 60). Göttingen:

Hogrefe & Hube.

Block, R. A., Hancock, P. A., & Zakay, D. (2010). How cognitive load affects

duration judgments: a meta-analytic review. Acta Psychologica. 134, 330–

343.

Block, R. A., & Zakay, D. (2008). Timing and remembering the past, the present,

and the future. In S. Grondin (Ed.), Psychology of time (pp. 367–394).

Bingley, England: Emerald.

Breska, A., & Ivry, R. (2016). Taxonomies of timing: where does the cerebellum fit

in? Current Opinion in Behavioral Sciences, 282-288.

Brown, S. W. (2008). Time and attention: Review of the literature. In S. Grondin

(Ed.), Psychology of time (pp. 111-138). Bingley, U.K.: Emerald Group

Brown, S. W. (2006). Timing and executive function: bidirectional interference

between concurrent temporal production and randomization tasks. Memory

& Cognition, 34(7), 1464–1471.

Brown, S. W. (1997). Attentional resources in timing: interference effects in

concurrent temporal and nontemporal working memory tasks. Perception

and Psychophysics, 59 (7), 1118–1140.

Brown, S. W. (1985). Time perception and attention: the effects of prospective

versus retrospective paradigms and task demands on perceived

duration. Perception & Psychophysics, 38, 115–124.

Brown, S.W., & Benette, E. D (2002). The role of practice and automaticity in

temporal and non temporal dual –task performance. Psychological

Research, 66, 80-89.

Brown, S. W., & Merchant, S. M. (2007). Processing resources in timing and

sequencing tasks. Perception & Psychophysics, 69, 439-449.

Bueti, D. (2011). The sensory representation of time. Frontiers in Integrative.

Neuroscience, 5:34.

Buhusi, C. V, & Meck, W. H. (2009). Relative time sharing: new findings and an

extension of the resource allocation model of temporal processing.

60

Philosophical Transactions of the Royal Society B: Biological Sciences, 364

1875–85.

Buhusi, C. V., & Meck, W. H. (2005). What makes us tick? Functional and neural

mechanisms of interval timing. Nature Neuroscience, 6, 755–765.

Buonomano, D. V. (2000). Decoding temporal information: a model based on

short-term synaptic plasticity. Journal of Neuroscience, 20: 1129–114.

Buonomano, D. V, Bramen J, & Khodadadifar, M. (2009). Influence of the

interstimulus interval on temporal processing and learning: testing the state-

dependent network model. Philosophical Transactions of the Royal Society

B: Biological Sciences, 364, 1865–1873.

Buonomano, D. V., & Karmarkar, U. R. (2002). How do we tell time?

Neuroscientist, 8, 42–51.

Buonomano, D. V., & Mauk, M. D. (1994). Temporal discrimination and the timing

of motor responses. Neural Computation, 6:38–55.

Buonomano, D. V., & Merzenich, M. M. (1995). Temporal information

transformed into a spatial code by a neural network with realistic properties.

Science 267, 1028-1030.

Burns, N. R., Nettelbeck, T., & McPherson, J. (2009). Attention and intelligence: A

factor analytic study. Journal of Individual Differences, 30(1), 44-57.

Burr, D. C., Cicchini, G M, Arrighi, R, & & Morrone, M. C. (2011). Spatiotopic

selectivity of adaptation-based compression of event duration”. Journal of

Vision, 11(2), 21, 1–9.

Burr, D. C., Tozzi, A., & Morrone, M. C. (2007). Neural mechanisms for timing are

spatially selective in real-world coordinates. Nature Neuroscience, 10, 423–

425.

Buzsáki, G., & Draguhn, A. (2004). Neuronal oscillations in cortical networks.

Science, 304, 1926-1929.

Bååth, R., Tjostheim, T. A., & Lingonblad, M. (2016). The role of ececutive control

in rythmic timing att diferent tempi. Psychomonic Bulletin & Review, 1954-

1960.

61

Casini, L. & Macar, F. (1997). Effects of attention manipulation on judgments of

duration and on intensity in the visual modality. Memory & Cognition, 25,

812- 818.

Chaston, A., & Kingstone, A. (2004). Time estimation: The effect of cortically

mediated attention. Brain and Cognition, 55, (2), 286-289.

Chen, A. H., O’ Leary, D. J., & Howell, E. R. (2000). Near visual function in young

children. Part I: near point of convergence. Part II: amplitude of

accommodation. Part III: Near heterophoria.Ophthalmic and Psysiological

Optics, 20 (3), 185-198.

Chen, Y., Repp, B. H., & Patel, H. D. (2002). Spectral decomposition of variability

in synchronization and continuation tapping: comparisons between auditory

and visual pacing and feedback conditions. Human Movement Science, 21,

515-532.

Cheng, R. K., Dyke, A. G, McConnell, M. W., & Meck, W. H. (2011). Categorical

scaling of duration as a function of temporal context in aged rats. Brain

Research, 1381, 175–86.

Coull, J. T., Vidal, F., Nazarian, B., & Macar, F. (2004). Functional anatomy of the

attentional modulation of time estimation. Science, 303, 1506–1508.

Coyle, T. R. (2003). A review of the worst performance rule: Evidence, theory and

alternative hypotheses. Intelligence, 31, 567−587.

Creelman, C. D. (1962). Human discrimination of auditory duration. Journal of

the Acoustical Society of America, 34, 582–593.

Csikszentmihalyi, M. (1990). Flow: The Psychology of Optimal Experience. New

York: Harper and Row.

Deary, I. J. (2000). Looking down on human intelligence: From psychometrics to

the brain. Oxford, UK: Oxford University Press.

Denckla, M. B., Rudel, R. G., Chapman, C., & Krieger, J. (1985). Motor proficiency

in dyslexic children with and without attentional disorders. Archives of

Neurology, 42, 228–231.

Dragoi, V., Staddon, J. E., Palmer R. G., & Buhusi, C. V. (2003). Interval timing as

an emergent learning property. Psychological Review, 110, 126–144.

62

Drewing, K., Hennings, M., & Aschersleben, G. (2002). The contribution of tactile

reaffernce to temporal regularity during bimanual finger tapping.

Psychological Research, 66, 60-70.

Droit-Volet, S., Meck, W. H., & Penney, T. B. (2007). Sensory modality and time

perception in children and adults. Behavioural Processes, 74, 244-250.

Eagleman, D. M. (2008). Human time perception and its illusions. Current

Opinion in Neurobiology, 18, 131–136,

Fitts, P. M, & Posner, M. I. Human Performance. Brooks/Cole Pub. Co; Belmont,

CA: 1967.

Flanagan, J. R., Vetter, P., Johansson, R. S., & Wolpert, D. M. (2002). Prediction

precedes motor control in motor learning. Current Biology, 13, 146-150.

Forsman, L., Madison, G., & Ullén, F. (2009). Neuroticism is correlated with drift

in serial interval production. Personality and Individual Differences, 47,

229-232.

Fortin, C., & Breton, R. (1995). Temporal interval production and processing in

working memory. Perception & Psychophysics, 57 (2), 203–215.

Fortin, C., & Rousseau, R. (1998). Interference from short-term memory

processing on encoding and reproducing brief durations. Psychological

Research, 61 (4), 269–276.

Fujioka, T., Trainor, L. J., Large, E. W., & Ross, B. (2012). Internalized timing of

isochronous sounds is represented in neuromagnetic beta oscillations. The

Journal of Neuroscience, 32 (5), 1791-1802.

Getty, D. (1975). Discrimination of short temporal intervals: A comparison of two

models. Perception & Psychophysics, 18, 1– 8.

Gibbon, J. (1977). Scalar expectancy theory and Weber’s Law in animal timing.

Psychological Review, 84, 279–325.

Gibbon, J., Church, R. M., & Meck, W. H. (1984). Scalar timing in memory. In J.

Gibbon & L. Allan (Eds.), Annals of the New York Academy of sciences (Vol.

423). Timing and time perception (pp. 52–77). New York: New York

Academy of Sciences.

63

Gibbon, J., Malapani, C., Dale, C. L., & Gallistel, C. (1997). Toward a neurobiology

of temporal cognition: advances and challenges. Current Opinion in

Neurobiology, 7, 170–184.

Gómez, J., Marín-Méndez, J., Molero, P., Atakan, Z., & Ortuño, F. (2014). Time

perception networks and cognition in schizophrenia: a review and a

proposal. Psychiatry Research, 220, 737–744.

Grondin, S. (2010). Timing and time perception: a review of recent behavioral and

neuroscience findings and theoretical directions. Attention. Perception &

Psychophysics, 72, 561–582.

Grondin, S. (2001). From physical time to the first and second moments of

psychological time. Psychological Bulletin, 127, 22–44.

Grondin, S., Bisson, N., Gagnon, C., Gamache, P. L., & Matteau, A. A. (2009).

Little to beexpected from auditory training for improving visual temporal

discrimination. NeuroQuantology, 7, 95–102.

Grondin, S., Gamache, P.L., Tobin, S., Bisson, N., & Hawke, L. (2008).

Categorization of brieftemporal intervals: An auditory processing context

impair visual performances. Acoustical Science & Technology, 29, 338–340.

Grondin, S., & Macar, F. (1992). Dividing attention between temporal and

nontemporal tasks: A performance operating characteristics – POC –

analysis. In F. Macar, V. Pouthas, & W. Friedman (Eds.), Time, action,

cognition: Towards bridging the gap (pp. 119-128). Dordrecht, The

Netherlands: Kluwer.

Grondin, S., & Rammsayer, T. H. (2003). Variable foreperiods and temporal

discrimination. The Quarterly Journal of Experimental Psychology, 56 (4),

731–765.

Grondin, S., & Rousseau, R. (1991). Judging the relative duration of multimodal

short empty time intervals. Perception Psychophysics, 49, 245–256.

Grondin, S., & Ulrich, R. (2011). Duration discrimination performance: No cross-

modal transfer from audition to vision even after massive perceptual

learning. Frontiers in Psychology, 5, 140.

64

Gualtieri, T. C., Johnson, L. G., & Benedict, K. B. (2006). Neurocognition in

depression: Patients on and off medication versus healthy comparison

subjects. Journal of Neuropsychiatry and Clinical. Neuroscience, 18, 217–

225.

Harrington, D. L., Castillo, G. N., Fong, C. H., & Reed, J. D. (2011). Neural

underpinnings of distortions in the experience of time across senses.

Frontiers in Integrative Neuroscience, 5, 32.

Harrington, D. L., Haaland, K. Y., & Hermanowicz, N. (1998). Temporal

processing in the basal ganglia. Neuropsychology, 12, 3–1210.

Harrington, D. L., Zimbelman, J. L, Hinton, S. C, & Rao, S. M. (2010). Neural

modulation of temporal encoding, maintenance, and decision processes.

Cereberal Cortex, 20, 1274–85.

Helmbold, N., & Rammsayer, T. H. (2006). Psychometric intelligence as measured

by speed and power tests. Journal of Individual Differences,, 20-37.

Helmbold, N., Troche, S., & Rammsayer, T. (2007). Processing of temporal and

nontemporal information as predictors of psychometric intelligence: A

structural-equation-modeling approach. Journal of Personality, 75 (5), 985-

1006.

Hinton, S. C, & Meck, W. H. (2004). Frontal-striatal circuitry activated by human

peak-interval timing in the supraseconds range. Cognitive. Brain Research,

21, 171–82.

Holm, L., Karampela, O. Ullén, F., & Madison, G. (2016). Executive control and

working memory are involved in sub-second repetitive motor timing.

Experimental Brain Research, 4839-4846.

Holm, L., Ullén, F., & Madison, G. (2013). Motor and executive control in

repetitive timing of brief intervals. Journal of Experimental Psychology:

Human Perception and Performance, 39, 365-380.

Holm, L., Ullén, F., & Madison, G. (2011). Intelligence and temporal accuracy of

behavior: unique and shared associations between intelligence, reaction time

and motor timing. Experimental Brain Research, 214, 175- 183.

Ivry, R. B., & Hazeltine, R. E. (1995). Perception and production of temporal

intervals across a range of duration: Evidence for a common timing

65

mechanism. Journal of Experimental Psychology: Human Perception and

Performance, 21, 3–18.

Ivry, R. B., Keele, S. W., & Diener, H. C. (1988). Dissociation of the lateral and

medial cerebellum in movement timing and movement execution.

Experimental Brain Research, 73, 167–180.

Ivry, R. B., Schlerf, J. E. (2008). Dedicated and intrinsic models of time

perception. Trends in Cognitive Science, 12 (7), 273–280.

Ivry, R. B., & Spencer, R. M. C. (2004). The neural representation of time. Current

Opinion in Neurobiology, 14, 225-232.

Ivry, R. B., Spencer, R. M. C., Zelaznik, H. N., & Diedrichsen, J. (2002). The

cerebellum and event timing. Annals of the New York Academy of Sciences,

978, 302–317.

Jahanshahi, M., Jones, C. R. G., Dirnberger, G., & Frith, C. D. (2006). The

substantia nigra pars compacta and temporal processing. Journal of

Neuroscience, 26, 12 266–12 273.

James, W. (1950). The principles of psychology. New York: Dover. (Original work

published 1890)

Jaskowski, P., Purszewicz, A., & Swidzinski, P. (1990). VEP latency and some

properties of simple motor reaction-time distribution. Psychological

Research, 52, 28–34.

Jensen, A. R. (2006). Clocking the mind: mental chronometry and individual

differences. Oxford, UK: Elsevier.

Jingu, H. (1989). Numerical processing of duration in temporal tracking behavior.

Japanese Psychological Research, 31, 169-178.

Jucaité, A., Dahlström, A., Farde, L., Forssberg, H., & Madion, G. (2008). Time

perception and production in children with ADHD correlates to the central

dopaminergic transmission (submitted for publication).

Justus, T. J. & Ivry, R., B. (2001). The cognitive neuropsychology of the

cerebellum. International Review of Psychiatry, 13, 276-282.

66

Karampela, O., Holm, L., & Madison G. (2015). Shared timing variability in eye

and finger movements increases with interval duration: Support for a

distributed timing system below and above 1 second. The Quarterly Journal

of Experimental Psychology, 68, 1965-1980.

Kaplan, B. J., Wilson, B. N., Dewey, D., & Crawford, S. G. (1998). DCD may not be

a discrete disorder. Human Movement Science, 17, 471-490.

Karmarkar, U. R., & Buonomano, D. V. (2007). Timing in the absence of clocks:

encoding time in neural network states. Neuron, 53, 427–438.

Karmarkar, U. R., & Buonomano, D. V. (2003). Temporal specificity of perceptual

learning in an auditory discrimination task. Learning & Memory, 10(2), 141-

7.

Kauranen, K., & Vanharanta, H. (1996). Influences of aging, gender, and

handedness on motor performance of upper and lower extremities.

Perceptual and Motor Skills, 82, 515-525.

Kee, D. W., Morris, K., Bathurst, K., & Hellige, J. (1986). Lateralized interference

in finger tapping: Comparisons of rate and variability measures under speed

and consistency tapping instructions. Brain and Cognition, 5, 268 –279.

Keele, S. W., & Hawking, H. L. (1982). Exploration of Individual differences

relavenat to high level skills. Journal of Motor Behavior, 3-23.

Keele, S. W., Pokorny, R. A., Corcos, D. M., & Ivry, R. B. (1985). Do perception and

motor production share common timing mechanisms: A correlational

analysis. Acta Psychologica, 60, 173-191.

Killeen, P. R., & Weiss, N. A. (1987). Optimal timing and the Weber function.

Psychological Review, 94, 455-468.

Klimesch, W., Freunberger, R, Sauseng, P, & Gruber, W. A. (2008). Short review of

slow phase synchronization and memory: evidence for control processes in

different memory systems? Brain Research, 1235, 31–44.

Koch, G., Oliveri, M., Carlesimo, G. A., & Caltagirone C. (2002). Selective deficit of

time perception in a patient with right prefrontal cortex lesion. Neurology,

59, 658–1659.

67

Kolers, P. A., & Brewster, J. M. (1985). Rhythms and responses. Journal of

Experimental Psychology. Human Perception Performance, 11,150–167.

Krampe, R. T., Kleigl, R., & Mayr, U. (2005). Timing, Sequencing, and Executive

Control in Repetitive Movement Production. Journal of Experimental

Psychology, 31, 379–397.

Kristofferson, A. B. (1980). A quantal step function in duration discrimination.

Perception & Psychophysics, 27, 300–306.

Lamotte, M., Izaute, M., & Droit-Volet, S. (2012). Awareness of time distortions

and its relation with time judgment: A metacognitive approach.

Consciousness and Cognition, 2, 835-842.

Lapid, E., Ulrich, R., & Rammsayer, T. (2009). Perceptual learning in auditory

temporal discrimination: No evidence for a cross-modal transfer to the visual

modality.Psychonomic Bulletin & Review 16, 382–389

Lejeune, H. (1998). Switching or gating? The attentional challenge in cognitive

models of psychological time. Behavioral. Processes, 44, 127–145.

Lewis, P. A., & Miall, R. C. (2006). Remembering the time: A continuous

clock. Trends in Cognitive Sciences, 10 (9), 401–406.

Lewis, P. A., & Miall, R. C. (2003a) Brain activation patterns during measurement

of sub-second and supra-second intervals. Neuropsychologia, 41(12):1583–

1592.

Lewis, P. A., & Miall, R. C. (2003b). Distinct systems for automatic and cognitively

controlled time measurement: Evidence from neuroimaging. Current

Opinion in Neurobiology, 13 (2), 250–255.

Lorås, H., Sigmundsson, H., Talcott, J. B., Öhberg, F., & Stensdotter, A. K. (2012),

Timing continuous or discontinuous movements across effectors specified by

different pacing modalities and intervals. Experimental Brain

Research, 220, 335–347.

Lustig, C., Matell, M. S., & Meck, W. H. (2005). Not “just” a coincidence: frontal-

striatal synchronization in working memory and interval

timing. Memory, 13, 441–448.

68

Macar, F., Grondin, S., & Casini, L., (1994). Controlled attention sharing

influences time estimation. Memory & Cognition, 22 (6), 673–686.

Madison, G. (2006). Duration-specificity in the long range correlation of human

serial interval production. Physica D, 216, 301−306.

Madison, G., (2001). Variability in isochronous tapping: Higher order

dependencies as a function of intertap interval. Journal of Experimental

Psychology: Human Perception and Performance, 27, 411– 421.

Madison, G., & Delignières, D. (2009a). Auditory feedback affects the long-range

correlation of isochronous serial interval production. Support for a closed-

loop or memory model of timing. Experimental Brain Research, 193, 519-

527.

Madison, G., Forsman, L., Blom, Ö., Karabanov, A., & Ullén, F. (2009).

Correlations between general intelligence and components of serial timing

variability. Intelligence, 37, 68- 75.

Madison, G., Karampela, O., Ullen, F., & Holm, L. (2013). Effects of practice

on variability in an isochronous serial interval production task:

Asymptotical levels of tapping variability training are similar to those

of musicians. Acta Psychologica, 1, 119-128.

Maes, P. J., Wanderley, M. M., & Palmer C. (2015). The role of working memory in

the temporal control of discrete and continuous movements. Experimental.

Brain Research, 233, 263–27.

Maniadakis, M., & Trahanias, P. E. (2014). Time models and cognitive processes: a

review, Frontiers in NeuroRobotics, Research Topic: Towards embodied

artificial cognition. doi: 10.3389/fnbot.2014.00007

Matell, M., & Meck, W. (2004). Cortico-striatal circuits and interval timing:

coincidence detection of oscillatory processes. Cognitive Brain Research, 21,

139-170.

Matell, M. S., Meck, W. H., & Nicolelis, M. A. L. (2003). Interval timing and the

encoding of signal duration by ensembles of cortical and striatal neurons.

Behavioral Neuroscience, 117, 760–773

Mattes, S., & Ulrich, R. (1998). Directed attention prolongs the perceived duration

of a brief stimulus. Perception & Psychophysics, 60, 1305-1317.

69

Matthews, W. J., & Grondin, S. (2012). On the replication of Kristofferson’s (1980)

quantal timing for duration discrimination: some learning but no quanta and

not much of a Weber constant. Attention, Perception and Psychophysics, 74,

1056-1072.

Mauk, M. D., & Buonomano, D. V. (2004). The neural basis of temporal

processing. Annual. Revisions in Neuroscience, 27, 307–340.

McFarland, K., & Ashton, R. (1978). The influence of concurrent task difficulty on

manual performance. Neuropsychologia, 16 (6), 735-741.

Meck, W. H. (2006a). Frontal cortex lesions eliminate the clock speed effect of

dopaminergic drugs on interval timing. Brain Research, 1108, 157–167.

Meck, W. H. (2006b). Neuroanatomical localization of an internal clock: a

functional link between mesolimbic, nigrostriatal, and mesocortical

dopaminergic systems. Brain Research, 1109, 93–107.

Meck, W. H. (1996). Neuropharmacology of timing and time perception. Cognitive

Brain Research, 3, 227–242.

Meck, W. H., & Benson, A. M. (2002). Dissecting the brains internal clock: How

frontal-striatal circuitry keeps time and shifts attention. Brain and

Cognition, 48, 195-211.

Meck, W. H., Penney, T. B, & Pouthas, V. (2008). Cortico-striatal representation of

time in animals and humans. Current Opinions in Neurobiology, 18, 145–

152.

Meegan, D. V., Aslin, R. N. & Jacobs, R. A. (2000). Motor timing learned without

motor training. Nature Neuroscience, 3, 860-862.

Merchant, H., Harrington, D. L., & Meck, W. H (2013). Neural basis of the

perception and estimation of time. Annual Revisions in Neuroscience, 36,

313–336.

Merchant, H., Zarco, W., & Prado, L. (2008). Do we have a common mechanism

for measuring time in the hundreds of milliseconds range? Evidence from

multiple interval timing tasks. Journal of Neurophysiology, 99, 939–949.

Michon, J. A. (1985). The compleat time experiencer. In: Jackson JAMJL,

editor. Time, mind, and behavior.Berlin: Springer Verlag; pp. 20–54.

70

Michon, J. A. (1966). Tapping regularity as a measure of perceptual motor load.

Ergonomics, 9, 401-412.

Miyake, Y., Onishi, Y., & Poppel, E. (2004). Two types of anticipation in

synchronization tapping. Acta Neurobiologiae Experimentalis, 64 (3), 415–

426.

Morrone, C., Ross, J., & Burr, D. (2005). Saccade cause compression of time as

well as space. Nature Neuroscience, 8, 950-954.

Nagarajan, S., Blake, D. T., Wright, B. A., Byl, N., & Merzenich, M. M. (1998).

Practice-related improvements in somatosensory interval discrimination are

temporally specific but generalize across skin location, hemisphere, and

modality. The Journal of Neuroscience, 18, 1559-1570.

Nagasaki, H. (1990). Rhythm in periodic tapping is centrally produced. Perceptual

and Motor Skills, 71, 985-986.

Nakamura, J., & CSíkszentmihályi, M. (2002). The concept of flow. Handbook of

positive Psychology, 89-105.

Nichelli, P., Always, D., & Grafman, J. (1996). Perceptual timing in cerebellar

degeneration. Neuropsychologia, 34, 863–871.

Ogden, R. S., Salominaite, E., Jones, L. A, Fisk J. E & Montgomery, C. (2011). The

role of executive functions in human prospective interval timing. Acta

Psychologica, 137 (3), 352–358.

Pariyadath, V., & Eagleman, D. (2007). The effect of predictability on subjective

duration. PLoS ONE 2(11): e1264. doi:10.1371/ journal.pone.0001264

Piek, J. P., Pitcher, T. M., & Hay, D. A. (1999). Motor coordination and

kinaesthesis in boys with attention deficit hyperactivity disorder.

Developmental Medicine and Child Neurology, 41(3), 159-165.

Pouthas, V., George, N., Poline, J. B., Pfeuty, M., VandeMoorteele, P. F.,

Hugueville, L., Ferrandez, A. M., Lehericy, S., Lebihan, D., & Renault, B.

(2005). Neural network involved in time perception: an fMRI study

comparing long and short interval estimation. Human Brain Mapping, 25,

433-441.

71

Rammsayer, T. H. (2008). Neuropharmacological approaches to human timing. In

S. Grondin (Ed.), Psychology of time (pp. 295-320). Bingley, U.K.: Emerald

Group.

Rammsayer, T. H. (2006). Effects of pharmacologically induced changes in NMDA

receptor activity on human timing and sensorimotor performance. Brain

Research, 1073, 407–416.

Rammsayer, T. H. (1999). Neuropharmacological evidence for different timing

mechanisms in humans. Quarterly Journal of experimental Psychology,

52, 273–286.

Rammsayer, T. H. (1997). Are there dissociable roles of the mesostriatal and

mesolimbocortical dopamine systems on temporal information processing in

humans? Neuropsychobiology, 35 (1):36–45.

Rammsayer, T. H (1993). On dopaminergic modulation of temporal information

processing. Biological Psychology, 36, 209–222.

Rammsayer, T. H. (1992). Effects of benzodiazepine-induced sedation on temporal

processing. Human Psychopharmacology: Clinical and Experimental

Psychology, 7 (5):311–318.

Rammsayer, T. H, & Brandler, S. (2007) Performance on temporal information

processing as an index of general intelligence. Intelligence, 35,123-139.

Rammsayer, T. H, & Brandler, S. (2002). On the relationship between general

fluid intelligence and psychophysical indicators of temporal resolution in the

brain. Journal of Research in Personality, 36,507-530.

Rammsayer, T. H, Lima, S. D. (1991). Duration discrimination of filled and empty

auditory intervals: Cognitive and perceptual factors. Perception &

Psychophysics, 50 (6), 565–574.

Rammsayer, T. H., & Troche, S. J. (2014). In search of the internal structure of the

processes underlying interval timing in the sub-second and the second

range: A confirmatory factor analysis approach. Acta Psychologica. 147, 68–

74.

Rammsayer, T. H, & Ulrich, R. (2005). No evidence for qualitative differences in

the processing of short and long temporal intervals. Acta Psychologica, 120

(2), 141–171.

72

Rao, S. M., Harrington, D. L., Haaland, K. Y., Bobholz, J. A., Cox, R. W, & Binder,

J. R. (1997). Distributed neural systems underlying the timing of

movements. Journal of Neuroscience, 17, 5528–5535.

Ren, X., Schweizer, K., & Xu, F. (2013). The sources of the relationship between

sustained attention and reasoning. Intelligence, 41, 132-128.

Repp, B. H. (2005). Sensorimotr synchronization: A review of the tapping

literature. Psychonomic Bulletin & Review, 12, 969-992.

Repp, B. H. (1998). The detectability of local deviations from a typical expressive

timing pattern. Music perception, 14, 419-444.

Schirmer, A. (2004). Timing speech: A review of lesion and neuroimaging

findings. Cognitive Brain Research, 21, 269–278.

Schmidt, R. A., & Wrisberg, C. A. (2008). Motor learning and performance (4th

ed.). Champaign, IL: Human Kinetics Publishers. (Translations into

Traditional Chinese, Polish, Greek, Japanese, Persian, Turkish, and an

electronic version).

Schweizer, K., & Moosbrugger, H. (2004). Attention and working memory as

predictors of intelligence. Intelligence, 32, 329-347

Schweizer, K., Moosbrugger, H., & Goldhammer, F. (2005). The structure of the

relationship between attention and intelligence. Intelligence, 33, 589-611.

Semjen, A., Vorberg, D., & Schulze, H. H. (1998). Getting synchronized with the

metronome: Comparisons between phase and period

correction. Psychological Research, 61, 44–55.

Serences, J. T., & Yantis, S. (2006). Selective visual attention and perceptual

coherence. Trends in Cognitve Science, 10, 38–45.

Sergent, V., Hellige, J. B. & Cherry, B. (1993). Effects of responding hand and

concurrent verbal processing on time-keeping and motor-implementation

processes. Brain and Cognition, 23(2), 243–262.

Spencer, R. M, & Ivry, R. B. (2009) Sequence learning is preserved in individuals

with cerebellar degeneration when the movements are directly cued. Journal

of Cognitive neuroscience, 21 (7), 1302–1310.

73

Spencer, R. M. C., Karmarkar, U., & Ivry, R. B. (2009). Evaluating dedicated and

intrinsic models of temporal encoding by varying context. Philosophical

Transactions of the Royal Society B: Biological Sciences, 364 (1525), 1853–

1863.

Studenka, B. E., & Zelaznik, H. N. (2011). Circle drawing does not exhibit auditory-

motor synchronization. Journal of. Motor. Behavior, 43, 185–191.

Taatgen, N., Rijnvan, H., & Anderson, J. (2007). An integrated theory of

prospective time interval estimation: the role of cognition, attention and

learning. Psychological Review, 114, 577–598.

Thomas, E. A. C., & Weaver, W. B. (1975). Cognitive processing and time

perception. Perception & Psychophysics, 17, 363-367.

Treisman, M. (1963). Temporal discrimination and the indifference interval:

Implications for a model of the “internal clock”. Psychological Monographs,

77, 576, 1-31.

Turvey, M. T. (1977). Preliminaries to a theory of action with reference to vision. In

R. E. S. J. Bransford (Ed.), Perceiving, acting and knowing (pp. 211-265).

Hillsdale, NJ: Lawrence Erlbaum.

Ullén, F., Forsman, L., Blom, Ö, Karabanov, A., & Madison, G. (2008). Intelligence

and variability in a simple timing task share neural substrates in the

prefrontal white matter. The Journal of Neuroscience, 28, 4239- 4243.

Ullén, F., & Madison, G. (2009) There is a bottom-up relation between temporal

accuracy and intelligence - further arguments from studies of correlations

between tapping variability and intelligence during high and low motivation.

In: International Society for Intelligence Research, Madrid

Ullén, F., Söderlund, T., Kääriä, L., & Madison, G. (2012). Bottom-up mechanisms

are involved in the relation between accuracy in timing tasks and intelligence

- further evidence using manipulations of state motivation. Intelligence, 40,

100-106.

Valera, E. M., Spencer, R. M. C., Zeffiro, T. A, Makris, N., Spencer, T. J, & Faraone,

S. V. (2010). Neural substrates of impaired sensorimotor timing in adult

attention-deficit/hyperactivity disorder. Biological Psychiatry, 68, 359–

367.

74

Vorberg, D., & Wing, A. (1996). Modeling variability and dependence in timing. In:

Heuer H, Keele SW, editors. Handbook of perception and action. Vol. 2.

London: Academic Press, 181–262.

Wearden, J. H., Edwards, H., Fakhri, M., & Percival, A. (1998). Why “sounds are

judged longer than lights”: Application of a model of the internal clock in

humans. Quarterly Journal of Experimental Psychology, 51B, 97–120.

Wiener, M., Turkeltaub, P., & Coslett, H. B. (2010). The image of time: a voxel-

wise meta-analysis. Neuroimage, 49(2), 1728-1740.

Wing, A. M., & Kristofferson, A. B. (1973). The timing of interresponse intervals.

Perception & Psychophysics, 13, 455– 460.

Woehrle, J. L., & Magliano, J. P. (2012). Time flies faster if a person has a high

working-memory capacity. Acta Psychologica, 139(2):314-319.

Woodrow, H. (1930). The reproduction of temporal intervals. Journal of

Experimental Psychology, 13, 473-499.

Wright, B. A., Buonomano, D. V., Mahncke, H. W., & Merzenich, M. M. (1997).

Learning and generalization of auditory temporal-interval discrimination in

humans. The Journal of Neuroscience, 17, 3956-3963.

Wulf, G. (2010). Social-comparative feedback affects motor learning. Psychology

Press, 4, 738-749.

Yamazaki, T., Tanaka, S. (2005). Neural modeling of an internal clock. Neural

Computation, 17, 1032–1058.

Zakay, D. (2000). Gating or switching? Gating is a better model of prospective

timing. Behavioural Processes, 50, 1–7.

Zakay, D. (1989). Subjective time and attentional resource allocation: an

integrated model of time estimation. In I. Levin & D. Zakay (Eds.), Time and

human cognition: A life-time perspective (pp. 365–397). Amsterdam: North-

Holland.

Zakay, R. A., & Block (2004). Prospective and retrospective duration judgments:

An executive-control perspective. Acta neurobiologiae experimentalis, 64,

319–328.

75

Zakay, D., & Block, R. A. (1995). An attentional gate moded of prospective time

estimation. In: M. Richelle (Ed.), Time and the Dynamic Control of

Behavior: IPA Symposium; Liége, November 7-8, 1994. France: Université

Liége.

Zarco, W., Merchant, H., Prado, L., & Mendez, J. C. (2009) Subsecond timing in

primates: comparison of interval production between human subjects and

rhesus monkeys. Journal of Neurophysiology, 102, 3191–3202.

Zelanti, P. & Droit-Volet, S. (2012). Auditory and visual differences in time

perception. An investigation from a developmental perspective with

neuropåsychological tests. Joural of Experimental Child Psychology, 112,

296-311.

Zelaznik, H. N., Spencer, R. M., & Doffin, J. (2000). Temporal precision in tapping

and circle drawing movements at preferred rates is not correlated: Further

evidence against timing as a general purpose ability. Journal of Motor

Behavior, 32, 193-199.

Zelaznik, H. N., Spencer, R. M., & Ivry, 2002. Dissociation of explicit and implicit

timing in repetitive tapping and drawing movements. Journal of Experiment

Psychology: Human perception & Performance, 28, 575-588.

Zelkind, I. (1973). Factors in time estimation and a case for the internal clock. The

Journal of General Psychology, 88, 295-301.

76

ACKNOWLEDGEMENTS

‘’The essence of all beautiful art, all great art is gratitude’’

Friedrich Nietzsche

After an intensive period of several months, today is finally the day! I am

really glad to write this note of thanks, the finishing touch on my thesis. This

thesis becomes a reality with the kind support of many individuals. I would

like to reflect on these people who have supported and helped me so much

throughout this period.

Firstly, I would like to express my gratitude to my principal supervisor, Guy

Madison, for giving me the chance to become a PhD student, as well as, for

all the constructive suggestions, availability and support throughout this

journey. I am also grateful to my co-supervisor, Linus Holm, for all the

guidance, support and helpful criticism through my PhD research studies. In

addition, I would like to thank my other co-supervisor, Fredrik Ullén, for the

really nice collaboration and all insightful suggestions in article writing.

Thank you all three for giving me the chance to share of your exceptional

scientific knowledge.

A big thank you goes to all the participants, for taking part in the studies.

Without you, there would be no thesis.

My gratitude goes to those who spent time reading this thesis at an earlier

stage. Anders Friberg, Anna-Maria Johansson, Erik Marsja, Johan Eriksson

and Maria Grazia Carelli, I am very grateful for the invaluable comments and

suggestions for improvement.

My appreciation also extends to all the colleagues at the Department of

Psychology at Umeå University. Particularly, I would like to thank all my

PhD colleagues for the encouragement, support, and all the fun times.

Elisabeth, once a roommate always a roommate! Your support has been

invaluable along the PhD journey. Thanks for being a good friend and a great

colleague. Thank you for helping through the hard times and then laughing

together when the hard times have passed. You are a unique person!

77

Johan, you have always trying to encourage me. I like your GO FOR IT

attitude. Thank you for listening, offering me advice, and for all the fun

times! You are a great person!

Kalyani, thank you for being part of this journey. All the help, the chats, and

the beautiful moments we shared, have been fundamental in supporting me

during these stressful moments. I really appreciate all you have done for me.

Thanks for your friendship!

Anna –Maria, I have to thank you again, as a friend this time. Your

genuine kindness is so exceptional!

Eva, our conversations on Lync should be saved for ever. Only by reading

those one could actually understand, how important social support is, during

the writing of the thesis. I do not forget our favorite word .Thank you and I

wish you the best luck!

Sincere thanks goes also to Maria Nordin, for all the support and the

valuable advice whenever needed.

I would like to send my sincere love, to the rest of my big Greek family and

friends. Thank you, one and all, for your love, support, and patience with me.

My godparents deserve a special note of thanks for providing me a blissful

and carefree childhood, as well as, for supporting me in acquiring the skills

and tools, in order to get into the higher education. I am also grateful to my

mother in law, for helping with all the housework and for taking care of

Alkinoos, while I was mentally and physically absent during the writing of

this thesis.

A very special thanks to you Giorgos, for designing the cover of the thesis.

How many talents can a person possibly have?

Aggele, thanks for all the help with the graphs and the funny moments in

your office! Please keep the Oly file.

I would also like to thank my parents for their endless support. You have

taught me about hard work and persistence. Bασίλη και Γεωργία

Καράμπελα, σας ευχαριστώ. Δεν υπάρχουν λόγια για να περιγράψω τις

θυσίες που κάνατε και τα όνειρα που αναγκάστηκατε να αφήσετε για να μου

δώσετε την ευκαιρία να πραγματοποιήσω τα δικά μου. Ελπίζω να είστε

περήφανοι για εμένα όπως είμαι εγώ για εσάς.

78

To my beloved brother, thank you for your love and support, as well as, for

the understanding you showed to me in every circumstance. You deserve the

best!

Alkinoe, my son, thank you for teaching me every day what life is all about

and for making all the troubles go away with just one smile. If you could see

yourself through my eyes, you would understand how special you are to me.

You have been a blessing from the start! Mommy is back now

Last but certainly not least, I must acknowledge with deep thanks my dearest

husband. Gianni, you have always been the driving force in our relationship.

You have given me confidence and your love has been motivated me in so

many different ways. I am sure that I could not face all the life’s challenges if

you have not been by my side. You are my personal compass, always pointing

the correct direction. Well, actually I don’t know if I can find any good words

that can express my gratitude towards you. My heart and soul want to say so

many things. Could those be possibly summed up in just two words?

Ευχαριστώ, Σ’αγαπώ.

Olympia Karampela,

Umeå, January, 2017.