PLEASE ENABLE MAC ROS FOR SPELL CHECK! '

•
PLEASE ENABLE MAC'ROS FOR SPELL CHECK!
8t. Cloud State University
Office of Sponsored Programs
NEW RESEARCHER AWARD FINAL REPORT
Prepare and submit within one month of completion of grant
Office of
DATE:
6/26/2012
TO:
Office of Sponsored Programs AS 210
FROM:
Sarah Smits-Bandstra, CSD
PROJECT:
Peer-Reviewed Dissemination of Stuttering Research
Date(s) of activity:
December, 2011 - June 26, 2012
Amount Awarded:
$4,000.00
Other funds used:
$0
JUN 2 6 2012
Spon~ored
Programs
Source:
What were the objectives of the project and to what extent have you been successful in
achieving those objectives?
The objective of this research award was to support my move toward peer-reviewed dissemination of my
research. I conducted a research project comparing adults who stutter and adults with Parkinson's disease.
I examined how each group learned new speech skills and collected their electrical brain activity while
learning. This project has important implications for speech therapy for each of these patient populations.
I have submitted two manuscripts based on this research to well-respected peer-review journals. I have
successfully published peer-reviewed articles in both of these journals before (the Journal of Fluency
Disorders, and the Journal of Clinical and Linguistic Phonetics). Both of the manuscripts I recently
submitted were well received and I was encouraged to revise and resubmit them.
During September and October, 20 II I was able to revise and resubmit my article entitled "Implicit
sequence learning and retention on nonsense syllables in persons who stutter and persons with Parkinson's
disease" to the Journal of Fluency Disorders. I received a reply in January rejecting the article (see
decision letter attached), but suggesting a very promising alternative journal to which I should send a
further resubmission. I researched several alternative journals to which I might send a resubmission, in
particular the Journal of Movement Disorders and the Journal of Clinical Neurology. Between June 13
and June 26 th 2012 I revised my article again and plan to submit it to the Journal of Movement Disorders
(see manuscript attached) before classes resume in August.
Between June 4 and June 13 th I thoroughly re-analysed my data and rewrote my manuscript entitled "Early stage chunking of finger tapping sequences by persons who stutter and fluent speakers". The completed manuscript is attached. My co-author is on holidays until August. After he reviews the manuscript I intend to resubmit the article to the Journal of Clinical Linguistics and Phonetics before classes resume in August. A second objective of this research award was to support the mentorship ofa M.Sc. thesis student in the
CSD department. This student is completing a thesis project analyzing the electrical brain activity data I
.~
PLEASE ENABLE MACROS FOR SPELL CHECK!
collected in my previous research studies. I needed support to travel with her to Montreal Quebec to the
laboratory where I originally conducted my research . At McGill University there is a laboratory dedicated
to the collection of electrical brain activity. I proposed to introduce this student (Whitney) to the process
of collecting brain activity data within this laboratory setting. She would not otherwise have access to this
advanced equipment at scsu.
In May 2012 I accompanied my thesis student to Montreal, Quebec, Canada. We flew to Montreal and
stayed at a hotel near the university (see attached Expedia receipt) and conducted our research on one
pilot subject and four experimental subjects while we were there. We also participated as experimental
subjects ourselves for two studies. My thesis student learned a great deal about recruiting subjects,
running an experiment, and collecting brain waves using advanced equipment. This trip was an
unqualified success and my student is analyzing the data we collected at this moment in order to
incorporate it into her thesis.
Discuss the outcomes and benefits of the project. Specifically comment on how the project has
moved you towards the peer-reviewed dissemination of your work or the submission of a
proposal for external funding. Include a copy of any published material, if applicable.
Primarily, this project had great significance for my own professional development. At the end of this
process, I will have published two first-author, peer-reviewed articles in respected journals. This project
will also have significance for the CSD department, in particular, and SCSU, in general, because first­
author publications will improve my chances of success when applying to external agencies for research
grants. Peer-reviewed publications will increase the visibility of SCSU within the field of communication
disorders on an international level because of the international readership base of these journals. In
addition, these publications will certainly impact student recruitment, particularly recruitment of high­
level students who wish to complete a thesis, apply for research scholarships, and perform research.
This funding also allowed me to provide my thesis student with an experience in international research
collaboration, and access to advanced technology, otherwise unavailable to her at SCSU. She plans to
publish her research and present her results at the American Speech-Language Hearing Conference in
November 2012.
Gmail- Your Submission to the Journal of Auency Disorders
hups:1 Imail.google.com/maill?ui=2&ik= 1892515e60&view=pt. ..
Sarah Smits-Bandstra <[email protected]>
Your Submission to the Journal of Fluency Di,sorders
2 messages
Journal of Fluency Disorders <[email protected]>
Wed, Jan 18, 2012 at 5:56 PM
To: [email protected]
Cc: [email protected]
Ms. Ref. No.: JFD-D-11-00059
Title: Implicit Sequence Learning and Retention of Nonsense Syllables in Persons who Stutter and Persons
with Parkinson's Disease
Journal of Fluency Disorders
Dear Sarah,
Thank you for submitting your above-referenced manuscript to JFD, which has now been reviewed by an
Associate Editor (AE) expert in the area, and two expert Editorial Consultants , who were blinded to your name
and institutional affiliation, and by myself. While the reviewers and AE believe the paper has improved, there
are still a number of problems with your paper, one being that the study does not provide a convincing
argument. Therefore, and I know this will be disappointing, I regret to inform you that we have concluded that
your manuscript is not acceptable for publication in the Journal.
The review of the Associate Editor and those of the Editorial Consultants are enclosed. I hope that a careful
reading of these enclosures will be helpful in understanding the basis for this publication decision and will
assist you in developing future research projects or clinical stUdies and that you will continue to submit your
work to the Journal of Fluency Disorders to consider for publication . Thank you for your submission to the
Journal of Fluency Disorders.
Yours sincerely,
Professor Ashley Craig , PhD Editor Journal of Fluency Disorders Reviewers' comments:
Dear authors,
Your paper has been reviewed by two experts and myself. We agree that you have done a good job in trying
to accommodate several concerns that were raised in previous reviews, but there remain some critical
problems with this paper. Although the 1st reviewer indicates that he/she is willing to accept the changes,
he/she also clearly expresses that the findings are not convincing. This sentiment is echoed by the comments
of the 2nd reviewer, who also highlights several other issues that have to do with the quality of writing, in
particular with respect to the discussion section. I agree with the assessment of the 2nd reviewer and the
overall statement that the findings as reported in this manuscript are rather ambiguous in terms of the
potential underlying sources for both the stuttering and Parkinson's individuals. This is reflected in the
discussion, which offers little more than a somewhat disjointed collection of ideas and speculations that often
have no direct
1 of 4
1/19/ 1212:47 PM
ail - Your Submi ssion
10
the Journal of AueDey Disorders
hups:l/mail.google.comimaill?u.i=2&ik=1892515e6O&view=pl. .
relationship to the original data. Moreover, you frequently refer to assumed neuropathological issues in these
populations without providing a proper evaluation of the actual processes that may have been involved In this
particular experimental paradigm. In particular, this paper falls to address one of the main issues that was
raised in previous reviews, namely how this specific type of task taps into speech motor processes in people
who stutter or people with Parkinson's disease. Mhough you have tried to avoid this question by taking out a
lot of references to speech motor control, you do highlight in the introduction and discussion problems in
speech motor processes for both populations and the way therapy may not adequately address those, thus
linking these issues to sequencing problems. However, you do not explain how the task that you used has any
direct bearing on sequencing as relevant for speech motor control and the issues you raise regarding targets
used in
therapy. For example, how leaming to produce a gentle voice onset in stuttering therapy or loudness training
for Parkinson has anything to do with identifying a pattern of nonsense syllable sequences remains totally
obscure. It is also interesting to notice that the individuals who are supposed to be mostly affected by
sequencing issues due to basal ganglia malfunction, viz. patients with Parkinson's disease , do not show the
strongest difference with control data in reaction times between pairs of nonsense syllables that form part of
the sequence vs. those that do not in the first session (and in the second session, all these trends are gone,
which makes the findings even more puzzling). In conclusion, this paper addresses an important topic but the
findings do not present a clear picture of what might be the underlying issues for either people who stutter or
those who have Parkinson's disease. Furthermore , how the specific demands related to this particular
paradigm tap
into specific speech production problems for both populations remains unclear and the suggested links to
neuropathology are highly speculative, especially for stuttering.
Specific comments:
1. Given the large inter-subject variability, figures 2, 3, and 4 would be better presented in the form of box
plots using msec as unit
2. On page 7, you refer to the study by Namasivayam & Van Ueshout (2008) but the information you provide
is msleading. First, the 2nd session was separated by a week from the 1st session not just a few days and
second, this was a study about implicit motor leaming in people who stutter (not explicit as you seem to
suggest) .
3. Page 38, line 34: "ceiling effect'; you probably mean "flooring effect" .
4. The appendices can be reduced in size by removing A-D
5. Table 6 seems to show two outliers; one in the control group (a positive difference of 330 ms for subject 7)
and one in the PWS group (a negative difference of 257 ms in subject 6). Ignoring that, overall, 2 controls, 7
PWS and 5 PPD show negative differences (i.e., faster RT for RAN vs. SEQ). Hardly strong evidence for
general problems identifying sequences in the two patient populations.
Reviewer #1 : The authors did a good job revising their manuscript according to the suggestions of myself, of
the first reviewer, and of the AE. I still find the results not very convincing, but the interpretation of the
outcome is now more adequate than before.
I still have a few mnor points:
p.3; 28 : "principles", not "principals"
of 4
lf19112 12:47 PM
Gmail • Your Submission to the Journal of Auency Disorders
https:l/mail .goog1e.com!maill?u.i=2&ik=1892515e60&view=pt ..
p. 3; 42-57 : Wikipedia should not be cited as a reference in a research article. The authors should rather refer
to some acknowledged textbook of cognitive neuropsychology!
p. 7 ; 10: remove comma
p. 8; 43-46 : kappa coefficients still quoted as percentages, - see my comments to the earlier draft
p. 11 ; 20: should probably read. (> 2SD from each group mean) .
p. 17; 28: what is meant by". aberrant white matter physiology within the motor cortex ."? Within the cortex we
only see grey matter, not white matter.? Please clarify!
p. 19; 54: "preclude". not "precluding"
Reviewer #2: This revised version is an improvement , having successfully addressed many of the complaints
of this reviewer. The conceptual background is more coherent and the experimental design has greater
clarity, helped by the schema given in Figure 1. The figures (please note subject-verb agreement error in the
fifth level down) .
As stated previously, this experiment addresses an interesting and important question about brain function in
the two clinical groups under consideration . The background and rationale are stated clearly. (However, the
use of a leng1hy paragraph from Wikipedia to define a phenomenon would be as unacceptable in a college
paper and as it is in a scholarly article.)
As an aside, acquiring a spatial shaping of an array of unrelated nonsense syllables is a long ways away from
implicit learning as it occurs naturally. Some characteristics of orderly organization and/or an unfolding with
transitional logic of SOME KIND inhere in anything we can think of as procedural memo'Y or procedural
learning. Even though this is accepted practice in psychology studies, the problem of ecological validity
remains. The authors should mention this limitation . (This limitation might contribute to the instability of the
results.)
Despite the several improvements, trouble arises again as the reader handles the overall managing of the
study's results and interpretation. Just when the exposition seems to have an orderly direction, a derailment
occurs in the form of a vagueness or contradiction . The abstrac1 states "the groups did not dilfer in retention
of implicit learning." A problem with this statement arises from results for the retention session provided on
page 12, where "PPD were significantly less accurate than PWS and near-significantly less accurate than
controls."
Even more perplexing, the authors state that the PPD group are weaker than other groups in implicit learning
tasks. How does this general statement align with the claim that there was no observed defiCit in retention of
implicit learning? The statement is consonant with results from the several studies reviewed in the
introduction showing poor implicit leaming in PPD but the resu~s of this particular study are not. Or perhaps it
has not been clearly explained for this reader.
Part of the difficu~ may arise from the many statistical comparisons that were made and the inclusion of
trends (p = .06) as meaningful to the interpretation of the study. The Inclusion of statistical trends in the
interpretation may contribute to these conundra .
While it is impressive to see the elfort that was undertaken for this study, the overall produc1 is lacking in
coherence and interpretability and therefore does not contribute significantly to our understanding of the topic.
The authors might consider limiting the report to partial components of the larger study that lend themselves
to clarity in interpretation . It is inappropriate for this elaborate and complex study to claim to refer to "prelimina'Y" results (p . 19). Editorial comments: This paper is marred by numerous grammatical errors. "Howeve(' is used in several wrong ways (e.g., page 19, bottom). Commas are missing or introduced speciously (e .g ., page 19 top) . It was demorahzlng to thiS reviewer to see that a previous waming was NOT heeded in this revision: (reviewer's previous comment) Editorial: subjec1-verb agreement , top page; on p. 7, second paragraph, run-<ln sentence (however as coordinating conjunction) . 3 of 4
111911212:47 PM
Gmail - Your Submission to the Journal of fluency Disorders
https: llrnail.google.comlrnail/?ui=2&ik=1892515e60&view=pt. ..
The authors are referred in a friendly manner to Strunk and White, Manual of Style .
For further assistance, please visit our customer support site at http://supoort.elsevier.com. Here you can
search for solutions on a range of topics, find answers to frequently asked questions and learn more about
EES via interactive tutorials. You will also find our 2417 support contact details should you need any further
assistance from one of our customer support representatives.
Sarah Smits-Bandstra <[email protected]>
Thu, Jan 19, 2012 at 9:21 AM
To: Vincent Gracco <[email protected]>
[Quoted text hidden)
Sarah Smts-Bandstra, Ph.D. CCC-SLP
Assistant Professor
Department of Communication Sciences and Disorders (CSD)
Brown Hall 103
St. Cloud State University
720 Fourth Ave S.
St. Cloud , MN 56301
Dept: 320-308-2092
Office: 320-308-2018
Fax: 320-308-6441
Website: www.stcloudstate.edu/csd
Email: [email protected]
40f4
1/19112 12:47 PM
Running Head: SEQUENCE LEARNING AND RETENTION
Verbal Implicit Sequence Learning and Retention in Stuttering and in Parkinson's Disease
2
Abstract
BACKGROUND: This study investigated the functional integrity of implicit learning systems in
persons with Parkinson's disease (PPD), persons who stutter (PWS) and control participants.
METHODS: Participants completed a verbal serial reaction time task where they were required
to say aloud four different syllables in response to four different visual stimuli. Unbeknownst to
participants, the syllables fonned a repeating 8-item sequence. General learning was assessed by
comparing groups' reaction time and accuracy over practice for Session One and after seven
days on the Retention Session. Implicit sequence learning was assessed by comparing reaction
time and accuracy for sequence syllables vs. random syllables on Session One and after seven
days on the Retention Session. Questionnaires assessed participants' explicit sequence
knowledge.
RESULTS: PWS demonstrated delays but not deficiencies in general learning (significant) and
retention (significant trend) relative to controls as measured by reaction time. PPD demonstrated
less accuracy in general skill acquisition and retention relative to controls. Control participants
demonstrated the expected pattern of implicit sequence learning in the initial session while PWS
(significant), and perhaps PPD (significant trend) differed from this pattern. The groups did not
differ in retention of implicit learning. All groups demonstrated similar limited explicit sequence
knowledge on the questionnaires.
CONCLUSIONS: PWS' differences in implicit sequence learning (involving the cortico-striato­
thalamo-corticalloop) may provide infonnation regarding possible etiological factors of
stuttering. Results of the current study are sufficiently promising to embolden further research
investigating implicit learning and retention in populations with motor speech disorders.
Educational objectives: As a result of this activity the participant will be able to: (1)
Differentially define procedural and implicit learning; (2) Summarize the reviewed literature
concerning the perfonnance ofPPD and PWS on retention tasks over practice, and (3) Explain
why a compromise in the functional integrity of the procedural learning system provides
infonnation regarding possible etiological factors of stuttering.
Keywords: Retention, Stuttering, Parkinson's disease, Nonsense Syllables, Serial Reaction Time
(SRT), Sequence, Learning, Procedural Learning, Implicit Learning, Motor Learning, Practice,
Basal Ganglia.
Introduction
Many programs designed for the treatment of various disorders of speech motor control (i.e.
apraxia, dysarthria and stuttering 1) are based on the theoretical principles of motor learning.
Duffy (2005) stated, "principles of motor learning should influence the structure of speech­
oriented treatment" (p. 449). Similarly Yorkston, Beukelman, Strand, and Hakel (2011) stated,
"the clinician must address factors that facilitate motor learning" (p. 388). In the area of
stuttering, Caruso and Max (J 997) held that a model of motor learning "may prove to be a
valuable theoretical framework to interpret changes in stuttering" (p. 213).
Most speech treatment programs are based on principles of motor learning derived from
studies of non-speech skill learning in healthy populations. For example, Yorkston et al. (2011)
reports "there is a large literature in cognitive motor learning that provides a good evidence base
for our use in planning effective therapy for people with motor speech impairment" (p. 390).
This reliance on non-speech literature is due to the unfortunate fact that the capabilities of
persons with motor speech disorders for motor learning and retention, relative to persons without
communication disorders, is relatively uninvestigated.
Many of the strategies taught in speech therapy require conscious, cognitively-based skills
(e.g., think LOUD) while other skills rely on unconscious, procedural learning skills that are
difficult to verbalize (e.g., easy onsets). That is, most clinicians teach their clients what to do, as
well as how to do it. The call of various authors to use motor learning principals intimates that, in
addition to other aspects, clinicians should make the most of the procedural learning system.
However, as mentioned above, there are few studies to date characterizing the integrity of the
procedural learning system in persons with motor speech disorders. It is essential that research
characterize the capabilities of various populations for procedural learning and retention. Only
then can we proceed to manipulate learning variables for the purpose of improving procedural
learning. The current study is a preliminary attempt to assess the integrity of procedural learning
and retention in two different popUlations and as such, does not have direct implications for
speech treatment.
Definitions ofprocedural memory, implicit sequence learning, and serial reaction time
According to Wikipedia (2011),
Procedural memory is memory for how to do things. Procedural memory guides the
processes we perform and most frequently resides below the level of conscious
awareness. When needed, procedural memories are automatically retrieved and utilized
for the execution of the integrated procedures involved in both cognitive and motor skills;
from tying shoes to flying an airplane to reading. Procedural memories are accessed and
used without the need for conscious control or attention. Procedural memory is a type of
long-term memory and, more specifically, a type of implicit memory. Procedural memory
is created through "procedural learning" or, repeating a complex activity over and over
again until we figure out how to make all of the relevant neural systems work together to
automatically produce the activity. Implicit procedural learning is essential to the
development of any motor skill or cognitive activity.
In this paper " learning" refers to consolidated learning (over four hundred trials, across
two sessions, spanning approximately seven days). This investigation focused on learning that
was procedurally based rather than declarative (recall and recognition skills), but may have had
cognitive components. Implicit sequence learning is an example of procedural learning where
participants learn to combine known movement components into a sequence but are unable to
verbally describelidentify the sequence. Speech treatments involve both explicit and implicit
learning. For example, therapists often model techniques such as easy onset, overarticulation, or
slow rate and explicitly direct the patient whenlhow to move the vocal folds or articulators. After
an initial model or explanation (explicit learning), the patient is left to 'figure it out' by trial and
error over practice (implicit learning) how to further refine and consolidate the new skill. In
contrast to explicit learning and direct procedural approaches, little is known about implicit
learning in speech pathology .
In the behavioral neuroscience literature, implicit sequence learning is acknowledged as a
relatively independent functional brain system associated with specific structures and
connections (particularly the basal ganglia and cortico-striato-thalamo-cortical connections) (see
Doyon & Benali, 2005). Within the motor learning literature there exists a well-established
protocol for examining implicit learning using serial reaction time tasks (Nissen & Bullemar,
1987) and verbal versions of serial reaction time tasks (Smith & McDowall, 2004; Westwater et
at. 1998). The verbal serial reaction time task of the current study was chosen as a preliminary
attempt to investigate implicit learning in patients with motor speech disorders. This task was
utilized because extensive previous research in the area of motor learning has established high
internal validity for this task. If it is determined that deficits in explicit and/or implicit motor
learning and retention exist for certain populations, future research incorporating actual speech
therapy tasks (high external validity) will be called for.
Populations o/interest
The current study examined the implicit sequence learning of persons with Parkinson's
disease (PPD) and persons who stutter (PWS). PPD were recruited for the study because of the
extensive body of research already existing regarding implicit sequence learning deficits for
speech and nonspeech tasks in this population (see the following section). Previous serial
reaction time studies with PPD were used as a guide for the methodology of the current study.
PWS were chosen as part of the study because many researchers have proposed dysfunction
within the cortico-striato-thalamo-cortical circuit as a possible etiological factor in stuttering.
This hypothesis is based on comparisons of PPD and PWS in many areas including brain activity
patterns during neuroimaging, sequencing performance, and speech symptoms (see Smits­
Bandstra & De Nil, 2007 for a review).
Hypokinetic dysarthria is the primary speech symptom associated with PD. Nevertheless ,
repetitive speech phenomena such as disfluency is commonly observed (in up to 54% of
advanced stage PD : Benke, Hohenstein, Poe we, & Butterworth, 2008), and can be a prominent
or presenting symptom of some PPD (Koller, 1983). Several studies have observed fluctuations
in stuttering dependent upon onset and progression of PD (Benke et aI., 2008), the use of
dopaminergic medications (Brady, 1998), or stereotaxic surgery of sites within the cortico­
striato-thalamo-corticalloop (Burghaus et aI., 2006). Disfluencies most often associated with PD
are stuttering, palilalia, speech freezing, and tachyphemia (see Louis, Winfield, fahn & Ford,
200 I for definitions and video-recorded examples). We also chose to examine the acquisition
and retention of implicit sequence learning by PWS to characterized implicit learning and
retention in this population. The ability of PWS to transition to automated performance (i.e. the
procedural system) has been brought into question in previous research (Smits-Bandstra, De Nil,
& Rochon, 2006; Smits-Bandstra & De Nil, 2009) and relapse is an acknowledged problem in
this population (Blomgren et aI., 2005). If, in our unprecedented direct comparison, PWS and
PPD demonstrated similar deficits on an implicit sequence-learning task known to involve the
striatal circuit, this would provide information regarding possible etiological factors of stuttering.
Moreover, consideration of the differences and similarities between stuttering and other motor
control disorders such as PO is essential to stimulate new hypothesis-driven research to
investigate the mechanisms of stuttering (Ludlow & Loucks, 2003).
The role ofthe striatal system in implicit sequence learning
In general, PPD demonstrate impaired implicit learning but relatively good explicit
learning skills. In contrast, patients with intact striatal systems (e.g., Alzheimer's disease) have
relatively good implicit learning but poor explicit learning, due to damage in the hippocampal
region (Saint-Cyr et a!. , 1988). These "double dissociation" studies emphasize the importance of
the striatal system relative to the hippocampal explicit learning system for implicit learning-type
tasks (an example of an implicit learning task is the serial reaction time task described below). It
should be emphasized, however, that these learning systems are not independent but interactive
and that, in many cases of neurological disease, aspects of both types of learning can be affected
to a certain extent. In addition, research has revealed roles for other areas of the brain (e.g., the
cerebellum), in implicit motor learning, which should be acknowledged but will not be discussed
in detail here (Sanes, Dimitrov, & Hallett, 1990).
Serial reaction time tasks, whether manual or verbal, have reliably found poor implicit
sequence learning in patients with impaired striatal systems such as Huntington ' s Disease (Saint­
Cyr et aI., 1988), and Parkinson's disease, unsurprising due to the striatal system deficits shared
by these populations (see Siegert, Taylor, Weatherall, & Abernathy, 2006, for a meta-analysis).
Impaired sequence acquisition by PPD has been found for both nonspeech and speech tasks.
Schulz et a!. (200 l) measured the accuracy of articulatory patterns of PPD relative to
control participants while learning a novel speech utterance (HThraim po /Ta mo dis" ) practiced
SO times. They found only control participants demonstrated increased accuracy with reductions
in duration and variability of productions over practice. Smith and McDowall (2004) conducted a
serial reaction time task that required verbal responses to a 12-item sequence. Twenty Parkinson
patients and 37 matched controls participated. Following a visual stimulus participants said aloud
" A, B, C, or 0 " to indicate one of four spatial locations where the next anticipated stimulus
would appear. Westwater and colleagues (1998) compared 13 non-demented PPD and II control
participants on a similar verbal serial reaction time task with a 10-item sequence. Designating
four locations on a computer screen as 1,2,3, and 4 /Tom left to right, PPD were required to say
the number aloud as quickly and accurately as possible. For both studies PPD demonstrated less
sequence specific learning relative to control participants as demonstrated by comparing the
reaction times of sequence trials to the reaction times of random trials.
Impaired acquisition of implicit sequencing has also been found in PPD both ON and OFF
dopaminergic medication (Feigin et a!., 2003; Muslimovic, Post, Speelman, & Schmand, 2007).
Muslimovic et a!. reported that the procedural learning impairment associated with PO is likely
independent of cognitive dysfunction or dopaminergic medication. It is important to note that
6
both Feigin et al. and Muslimovic et al. indicated a mild negative effect of dopaminergic
medication on declarative knowledge. Abbruzzese, Trompello and Marinelli (2009) concurred
that the mechanisms of implicit learning appeared relatively independent from dopamine­
replacement therapy . While the characteristics of implicit sequence acquisition in PPD has been
investigated by many studies to date, the retention of implicitly learned skills associated with
striatal dysfunction (e.g., Parkinson's disease) is a new area of investigation.
The role a/the striatal system in retention a/procedural skill
Retention tests measure whether performance changes due to motor learning are well
established. A retention test often consists of a sample of the practiced experimental task and
may be conducted after one or several days (Guadagnoli, Leis, Van Gemmert, & Stelmach, 2002;
Karni, 1996), or after a much longer rest period of months to years (Shadmehr & Brashers-Krog,
1997). Motor learning is not said to be well established, nor invulnerable to interference until
after consolidation (Shadmehr & Brashers-Krog, 1997). Research indicates that consolidation
occurs over a rest phase of five hours or longer (Brashers-Krog, Shadmehr, & Bizzi, 1996), after
extensive practice (Hauptmann & Kami, 2002).
A number of studies have investigated the importance of the striatal system for retention of
implicit skills. Kawai, Kawamura, and Kawacru (1999) compared a patient diagnosed with
Progressive Supranuclear Palsy (striatal involvement, poor implicit learning) and a patient
diagnosed with Alzheimer' s disease (hippocampus involvement, poor explicit memory) using a
bimanual coordinated tracing task . The patient with Progressive Supranuclear Palsy improved
with training, but this improvement disappeared within a few months. The patient with
Alzheimer's disease maintained improvements at the 18-month follow-up . This effect was
confirmed with a larger patient sample by Mochizuki-Kawai, Kawamura, Hasegawa, Mochlwki,
Oeda, Yamanaka, & Tagaya (2004).
Retention of skills has also been found to be deficient as striatal degeneration continues,
in later stages of Parkinson ' s disease (Doyon et al. 1998) and particularly for PPD with
bradyk..inesia as the primary symptom (Vak..il & Herishanu-Naaman, 1998). Eight to 16 months
after learning an implicit sequencing task, PPD who changed from Stage I to Stage 2 on the
Hoehn and Yahr scale (1967) of the disease showed less retention of sequencing skill. The
authors determined that the difference was not due merely to motor deterioration per se (Doyon
et al. 1998).
Of note, those studies which examined motor sequencing skills over one session or over a
few days found good retention by PPD (Harrington , Haaland, Yeo, & Marder, 1990), while those
whlch looked over months of practice found increasing differences in sk.ill retention between
PPD and control participants (Agostino et aI., 2004). No studies to date have examined the
retention of implicitleaming using a syllable serial reaction time task in populations with striatal
disorders.
Sequence Learning in PWS
Similar to PPD, PWS have demonstrated some evidence of sequence learning differences
relative to fluent speakers, however most experiments to date have focused on explicit sequence
learning tasks only (see Smits-Bandstra & De Nil, 2007 for a review). Webster (1986) evaluated
the ability of PWS and control participants to learn 4-element fmger tapping sequences and
7
found that PWS did not show improved accuracy after practice on sequences with no repeated
elements (2-1-3-4 as opposed to 2-1-3-1) compared to controls. In a study by Smits-Bandstra,
De Nil, and Saint-Cyr (2006) nine PWS and nine controls asked to type a IO-nwnber sequence
(i.e. 14 2 1 3 1 2 4 1 3 4f). While the finger reaction times for the first trial were almost identical
for both groups, PNS improved their reaction times more rapidly as compared to PWS.
Ludlow, Siren, and Zikira (1997) asked PNS and PWS to practice two, 4-syllable
nonsense words. Participants practiced recalling and producing the words between trials of a
listening comprehension secondary task. Ludlow et al. (1997) found that PWS were impaired in
their rate of learning as well as the overall accuracy of their nonsense word productions.
Similarly, Cooper and Allen (1977) found that PWS required significantly more repetitions to
demonstrate increased speed of repetition for read-aloud paragraphs and sentences compared to
PNS when they used between 16 and 110 trial repetitions. In a study by Smits-Bandstra and De
Nil (2009) 12 PWS and 12 control participants were presented visually with a 10-syllable
sequence (Ita ba pa ta ga pa ga ta pa baf) on a computer screen and read aloud the
sequence 30 consecutive times as quickly and accurately as possible. The results indicated that
PNS demonstrated significantly more improvement on reaction time over practice compared to
PWS.
Retention/or PWS
Smits-Bandstra, De Nil, and Saint-Cyr (2006) examined short-term retention of an
explicit sequence learning task (participants were shown the whole sequence and encouraged to
learn it). PWS demonstrated significantly slower reaction times for a ten-item nonsense syllable
sequence relative to control participants on a retention test given approximately one hour after
practice. Narnasivayam and van Lieshout (2008) measured the speed and coordination oflip
movements ofPWS and control participants using electromagnetic articulography . PWS were
asked to repeat the pronounceable nonsense words Ibapil, Ibipa/, Ibapiterl, and Ibipaterl for 12
seconds with a bite block in place. One to three days later, PWS demonstrated slower nonsense
word performance and more variable lip coordination during production as compared to control
participants.
These preliminary studies suggest some deficits in explicit sequence learning and
retention in PWS. However, the implicit sequence learning and retention capabilities of PWS
have not been investigated.
Purpose o/the current study
The primary objective of the current study was to compare implicit sequence leaming and
retention ofan implicit nonsense syllable sequence in PWS, PPD, and control participants.
Similarities or differences between PPD and PWS will provide much needed clues as to the
functional integrity of the implicit sequence leaming and retention system (Sciunidt & Wrisberg,
2004) in these populations. Functional deficits in implicit sequence learning, which involve the
cortico-striato-thalamo-corticalloop, will provide information regarding possible etiological
factors of stuttering. It has been suggested by several researchers in rehabilitative medicine that
the knowledge of all aspects of motor learning abilities in patient populations is critical in
designing more effective rehabilitative protocols (Abbruzzese et aI., 2009).
Methodology
Parlicipanls
Participant infonnation is presented in Tables I and 2. Participants included 14 English­
speaking PPD (7 females; 65.1, SD 6.9). PPD were outpatients with idiopathic Parkinson's
disease. Diagnosis by a licensed neurologist was based on the presence of a rigidity-akinesia
syndrome, and responsiveness to Levodopa, without signs of pyramidal, cerebellar, or
oculomotor deficits.
PPDs' conversational speech was rated on overall intelligibility and dysarthria
characteristics by a certified Speech-Language Pathologist (see Appendix A). The speech and
reading samples of all PPD (14) scored below (2.8, SD 2.7) the very mild range (10-17) on the
Stuttering Severity Instrument 4 (SSI-4; Riley, 2009). All PPD demonstrated less than 3% of
syllables stuttered for their speech and reading samples. Table 2 also includes a rating of speech
affectedness from the Activities of Daily Living subsection of the Unified Parkinson's Disease
Rating Scale (Fahn, Elton, & Members of the UPDRS Development Committee, 1987), and
overall stage of disease (modified Hoehn & Yahr, 1967; see Appendix B). Twelve PPD had
intelligibility between 95-100%; and two PPD had intelligibility between 85-95%. A second
rater, blind to the conditions of the study, also rated 20% ofPPDs' speech samples. Inter-rater
reliability calculated using the kappa coefficient was 100% for intelligibility ratings, 84% for
dysarthria characteristics, 94.7% for stuttering frequency of speech samples and 97.3 % for
stuttering frequency of reading samples. All PPD were tested during the ON-cycle of their
medication within 60-90 minutes of their last dose. Because dopaminergic treatment speeds up
the execution of motor sequences (Benecke, Rothwell, Dick, Day, & Marsden, 1986), to
minimize bradykinesia at baseline, we specifically studied medicated patients.
Fourteen English-speaking PWS (6 females; 65.1, SD 5.7), scored in the very mild (9), mild
(3) and moderate (2) range (10.3, SD 7.2) on the SSI-4 (Riley, 2009). All participants reported
onset of stuttering in early childhood on the screening questionnaire. It is noteworthy that the
large majority of PWS scored in the very mild to mild range on the SSI-4 . On average, PWS
demonstrated an average of 3% of syllables stuttered (ranging from 1%- 15%). A second rater,
blind to the conditions of the study, also rated 20% ofPWS' speech samples. Inter-rater
reliability using the kappa coefficient was 94.8% for stuttering frequency of speech samples and
93.7% for stuttering frequency of the reading samples. Fourteen English-speaking participants (6
females; 65.0, SD 5.8) served as matched control participants.
All PWS and PPD were screened for depression. All PPD were screened for mental state.
All participants were screened for forward and backward digit span, hearing, vision, medication
use (other than PPD medication), neurological and motor control difficulties (other than PD), and
speech and language difficulties (other than those associated with PD or stuttering). PWS and
PPD who had received speech therapy within the last 6 months were excluded as these
treatments typically teach slowed rate of speech which may have interfered with the task .
Written infonned consent was obtained and all participants were treated according to ethical
treatment of human participant guidelines established by McGill University, the University of
Toronto and the Baycrest Centre in Toronto.
9
Procedures
Familiarization
The experiment was composed of two sessions, Session One and the Retention Session.
Before Session One, and again before the Retention Session, all participants underwent 16
training trials of the experiment. Participants saw four horizontal lines on the screen (see Figure
I) and heard a beep. They were told that they should get ready and direct their attention to the
screen when they heard this beep. Participants then saw an 'X' appear over one of the four lines
and were told that each of the four 'X' locations was associated with a unique syllable.
Participants were told that if an 'X' appeared in the top left comer, they must say "PA", in the
top right comer, "PE", in the bottom left comer, "PI" and the bottom right comer, "PO".
Participants were instructed to "Say the syllable aloud as quickly and accurately as possible as
soon as the 'X' appears on the screen". In the first eight of the 16 training trials the correct
syllable was written below the X when it appeared as a learning cue. The investigator provided
spoken feedback regarding accuracy during all of the familiarization trials. The familiarization
trials had no time limit so participants could take their time and ask questions.
Verbal serial reaction time task
Stimuli were presented on a IS-inch lap top screen with a viewing distance of 18-20
inches. For each single trial a blank screen was presented for 400 ms (from 0 ms to 400 ms) (see
Figure I). Participants were presented with four horizontal lines (i.e.,::) in the middle of the
screen for 900 ms (from 400 ms to 1300 ms). This visual cue was presented simultaneously with
a 1000 Hz tone for 500 ms (from 400 ms to 900 ms). Finally an X over one of the four lines was
presented for 2400 IDS (from 1300 to 3700 ms). Subjects had to respond immediately upon
seeing the X by saying aloud the correct corresponding syllable they had leamed in training (e.g.,
"PA"). Each trial was followed by a jittered interstimulus interval of3600-3900 ms (from 3700
ms to between 7000 and 7300 ms). Jittered interstimulus intervals are variable time intervals
inserted between consecutively presented individual stimuli to prevent anticipatory responses.
Stimuli were presented using the Presentation 0.8 (Neurobehavioral Systems, 2004) software
program.
Unknown to the participants, the locations appeared in a predicable sequence of 8
locations (i.e. PO PI PO PE PI PA PE PA; see Appendix E). The participants were not made
aware of the sequence pattern but were expected to leam it implicitly (unconsciously) as
reflected in decreases in reaction time over practice. Thus it took 8 individual trials to complete
the sequence. Nine sequences were contained in one "practice block". Each practice block ran
through the eight-item sequence nine times (72 trials). There were short breaks every 2 or 3
sequences (short breaks of 5 s each). After each break the sequence trials began again, but
always at a different point in the sequence to keep the sequential nature of the stimuli less
obvious (implicit). During each break participants were given reminders of the correct syllable
that went with each 'X' location.
Session One
The overall order of Session One is presented in Appendix E. No feedback regarding
speed or accuracy was provided during Session One. The primary investigator corrected
participants only if a participant incorrectly substituted one syllable for another for a whole block
(e.g., PI for PO), indicating they had forgotten the training/instructions. After four blocks of
10
sequence practice (72 trials each), participants were presented with a fifth block. This fifth block
is hereafter referred to as 'random' for expediency but was, technically, 72 pseudo-randomly
generated locations (there were no repeated syllables).
Retention Session
The overall order of the Retention Session is presented in Appendix F. No feedback
regarding speed or accuracy was provided during the Retention Session. The primary
investigator corrected participants only if a participant incorrectly substituted one syllable for
another for a whole block (e.g., PI for PO), indicating they had forgotten the
traininglinstroctions. The Retention Session took place after approximately seven days for
controls, (M = 6.7, SD = 3.1), PWS (M = 6.1, SD = 1.2), and PPD (M = 5.9, SD = 1.8),
respectively. This length of retention period was selected because it was comparable to previous
studies looking at retention after several days (Kawai et aI., 1999; Mochizuki-Kawai et aI. ,
2004), but was still manageable for the investigator traveling considerable distances to various
participants. Tbe Retention Session was composed of two blocks of sequence practice (72 trials
each) followed by a third 'random' block of 72 pseudo-randomly generated locations.
Explicit questionnaires and generate task
Participants completed the " Day One" explicit awareness questionnaire after Session One
and the "Day Two" questionnaire after the Retention Session (see Appendices C and D). The
generate task was a computer task presented without pause following the ' random' block (block
3) in the Retention Session (see Appendix F). The generate task was identical to the experimental
task described in the following paragraphs with one exception. The exception was, after a portion
of each sequence had been completed (e.g., three trials), participants were presented visually
with a "?" on the computer screen instead of the usual 'X' on one of four horizontal lines. When
they saw the "?", participants were instrocted to "say the syllable corresponding to where they
thought they 'X' would next appear."
Dependent Variables and Statistical Analysis
Excluded trials included yawns, sneezes, laughing, licking the lips, clearing the throat,
equipment glitches, etc., and were not included in any of the analyses. In Session One control
participants, PWS, and PPD had 7.9%, 7.4% and 7.4% excluded trials respectivel!. For the
Retention Session control participants, PWS, and PPD had 4.8%, 2.7% and 2.9% excluded trials
respectively.
The dependent variables were accuracy of production and reaction time . Inaccuracy was
defined as incorrect syllable substitutions (i.e. incorrect vowels; e.g. , !PAI for !PEl). PPD were
in early stages of the disease with well-preserved speech skills (see Table 2). Dysarthria, voicing
errors or distorted articulation of the initial /p/ consonant did not noticeably affect single syllable
productions for these patients. All productions were categorized as correct productions or
incorrect substitutions.
An entire block was removed if there were three consecutive presentations or 4/5
presentations when the same incorrect syllable (e.g., PI) was always substituted in place of the
correct syllable (e.g., PE), indicating the participant had forgotten the instructions/training. In
these cases, participants typically made the error throughout the block. Participants received
reminders of the syllable names and locations by tutorials on the computer screen during each
II
break, or by the investigator after a block of these recurring consistent errors were made. In
Session One, control participants, PWS, and PPD had 15.4%, 12.3%, and 16.9% of blocks
excluded as such, respectively. In the Retention Session control participants, PWS, and PPD bad
10.1 %, 8.9%, and 5.3% of blocks as excluded data, respectively. There were a considerable
number of trials removed due to difficulty understanding to task. This is most likely the result of
the minimal amount of training provided. In the current experiment minimal training was
provided purposefully, in order to observe the early stages of the learning process.
Disfluencies were defined as silent blocks or any repeated, prolonged, or e/Tortful sounds
or syllables observed on-line, through videotape recordings and/or spectrogram analysis.
Disfluencies were omitted from accuracy and reaction time analysis. For Session One, Control
participants, PWS, and PPD bad 70, 109, and 187 of trials excluded as disfluent, respectively. In
the Retention Session control participants, PWS and PPD bad 40, 60, and 82 of trials excluded as
disfluent, respectively.
Inordinately long reaction times were categorized as outliers (2+SD from each group
mean) and were removed from analysis. One reason for the removal of these outliers was to
control for akinesia/speech freezing in PPD. Video analysis confIrmed the presence of these
effortless/tension-free speech hesitancies (no preparatory breath, no pre-forming of lips, no
voicing or evidence of struggle) but the participant was still attending to the task. In Session One,
outliers amounted to 4.6% for controls, 5.8% for PWS, and 4.2% for PPD. In the Retention
Session outliers amounted to 3.2%, 3.0%, and 3.2% total trials for control participants, PWS, and
PPD respectively. There were no cases of improbable/anticipatory reaction time in the data. The
inclusion of a jittered interstimulus interval likely prevented such anticipatory responses.
Syllable reaction time was measured as the time (ms) from the onset of the stimulus
presentation to the voice onset of the syllable (the onset of the vowel after the initial consonant).
Participant's speech was recorded using a dual channel digital audio tape recorder (Tascam DA­
01) with 16 bit resolution and 48 kHz sampling rate. Reaction times were calculated off-line
using waveform acoustic analysis Speech Analyzer 3.0.1 (SIL,2007).
Reliability
For Session One, percentage agreement with an independent rater on the occurrence of
disfluencies vs. inaccuracies was 96.7%, and 99.9% respectively, based on 10% of the
sequences. For the Retention Session, agreement was 97.2%, and 99.4% respectively, based on
10% of the sequences.
An independent trained rater, blind to the conditions of the study, re-analyzed 10% of the
participants' syllable reading acoustic waveforms. For Session One, occurrence agreement inter­
rater reliability for cursor placement within 10 ms was 98% (r = .99, ? = .98) and for the
Retention Session, 93% (r = .99, r] = .98, a similarly high correlation). When, for example, rater
one recorded reaction time for a syllable as 500 ms, if rater two recorded reaction time for tbe
same syllables as 508 ms, they were considered in agreement within 10 ms. In both sessions
there was 99.5% agreement within 20 ms.
General learning (non-specific) vs. Implicit learning (5equence-specific)
Generalleaming can be inferred from a gradual reduction in reaction time over practice
blocks. Reaction times are a reflection of " generalleaming" because they are non-specific,
12
meaning they include some aspect of implicit sequence learning, but also more general task
adjustment and stimulus-response learning ("PA" = lower left comer). Statistical analysis of
reaction time included samples from the beginning of each block. Two means were taken from
each practice block for a total of eight reaction time means across the four blocks. This was done
to observe the large amount of change occurring during early trials as the sequence was being
learned. The mean of trials 1-8 (a total of8 trials, the first total sequence) and the mean of trials
9-16 (a total of 8 trials, the second total sequence) were taken from Block I, Block 2, Block 3
and Block 4 in Session One.
For the Retention Session, a traditional retention comparison was made between Session
One Block 4 (the mean of trials 1-8) and the Retention Session Block I (the mean of trials 1-8).
An additional, more detailed retention analysis included eight reaction time means across
Session One and the Retention Session . There were four Session One means including trials 1-8
and 9-16 of Block I, and trials 1-8 and 9-16 of Block 4. There were four Retention Session
means including trials 1-8 and 9-16 of Block I, and trials 1-8 and 9-16 of Block 2. This analysis
over eight means (each an average of 8 trials) was chosen to compare the process of "forgetting"
and "relearning" across groups in more detail. Poor retention was expected as a cubic curve
where performance improves over practice in Session One but suffers upon introduction of the
Retention Session.
Sequence Blocks vs. Random Blocks
Traditionally, implicit sequence learning is assessed by comparing reaction times of stimuli
that follow a sequence pattern to the reaction times of random stimuli (Nissen & Bullemar,
1987). This comparison is made after some amount of practice so that participants have had time
to implicitly learn the sequence. Reaction time for the random syllables should reflect several
kinds of general learning (hence non-specific learning). Reaction time for the sequence syllables
should be even shorter than reaction time for random syllables, reflecting implicit knowledge of
the sequence (hence sequence-specific). Shorter reaction time for sequence vs. random syllables
is interpreted as an indicator of implicit sequence learning. To assess implicit sequence learning
in Session One we compared the reaction time of Block 4 sequence syllables (first 16 trials) with
Block 5 random syllables (first 16 trials). To assess retention of implicit learning in the Retention
Session we compared the reaction time of Block 2 sequence syllables (first 16 trials) with Block
3 random syllables (first 16 trials).
This traditional comparison of a sequence block and a random block is problematic because
within the random block, many of the syllables happened to fall into pairs, triplets, etc. that
conformed to the practiced sequence. For example, the combination of a " PO" trial and a "PE"
trial occurred in that particular order in the sequence (see Appendix E), but this exact
combination also occurred sometimes when the syllable trials were randomly distributed.
Recently researchers have proposed the use of probability as a more sensitive indicator of
learned pair-wise associations or chunks within a sequence. Wilkinson and lahanshahi (2007)
compared reaction times of frequently paired items (the practiced sequence items) to rarely
paired items (sequence violations).
SEQ vs. RAN
Based on this more sensitive, probability-based analyses , the 'random' block (Block 5) trials
were divided into syllable pairs that conformed to the sequence (SEQ), or violated the sequence
(RAN). That is, within the random block, trial pairs that occurred in the sequence (e.g, PO PEl,
13
and were practiced in that order hundreds of times were called SEQ. Within the random block,
pairs of trials that never occurred in the sequence (e.g., PE PO) and were never encountered
before by the participant were called RAN (see Appendix E). Wilkinson and Jahanshahi (2007)
demonstrated that the reaction time of the second syllable in each SEQ pair captured the
'anticipation' associated with implicit learning. This shortened 'anticipatory' reaction time
occurred because subjects who implicitly knew the sequence could anticipate which syllable
followed another. Within Session One, the first 16 RAN trials and the first 16 SEQ trials of the
random block (Block 5) were included in analysis (see Appendix E). Within the Retention
Session, the first 16 RAN trials and the first 16 SEQ trials of the random block (Block 3) were
included in analysis (see Appendix F).
Explicit questionnaires and generate task
Some degree of explicit knowledge of a sequence/pattern of stimuli often arises during a
serial reaction time experiment and the greater the extent of explicit knowledge, the bener
sequencing performance (Mayr, 1996). Therefore, it is incumbent upon researchers to measure
explicit knowledge as a potential confound of sequence learning that is assumed to be
predominantly implicit. Traditionally, researchers have employed an explicit questionnaire
where parts of the sequence are presented for participants to recognize/identify (Eimer, Goschke,
Schlaghecken, & Sturmer, 1996; Ghilardi et aI., 2007; Russel & Rosier, 2000). We presented
two-, three-, and four-syllable portions of the sequence as suggested by Russel and Rosier (2000)
so as not to underestimate explicit knowledge of parts of the sequence, even if participants had
not realized the whole sequence. We included the explicit questionnaire score as a dependent
variable reflecting the extent to which sequencing performance can be accounted for by explicit
rather than implicit knowledge.
For the Day One questionnaire (see Appendix C), participants were awarded single points
for each portion of the sequence they correctly identified as occurring "always" or "often" in the
experiment. Additional single points were awarded for violations of the sequence they correctly
identified as "rarely" or "never" occurring in the experiment. Ten total points were possible with
two correct answers and three foils (chance performance = 40%). For the Day Two
questionnaire (see Appendix D), participants were awarded single points for each portion of the
sequence they correctly "guessed", "were somewhat sure", or "were sure" were part of the
sequence. Additional single points were awarded for violations of the sequence they correctly
"guessed", "were somewhat sure", or "were sure" were NOT part of the sequence. Ten total
points were possible with three correct answers and three foils (chance performance = 50%).
Scores greater than chance (40% for Day One and 50% for Day Two) indicated increasing
explicit knowledge of the sequence. Questionnaire scores were expected to approach chance
levels indicating predominantly implicit learning of the sequence.
Recently there has been some evidence that questionnaires may underestimate explicit
knowledge (Pe igneux et aI., 2000), because they are completed sometime after the actual
experiment and are presented in a different media than the sequence (paper and pen vs. computer
screen and keyboard). In response to these concerns we also included a computer-based
"generate" task (see Appendix F; see also "explicit questionnaires and generate task" in the
method section). Following the Retention Session participants completed the generate task. The
generate task was very similar to the experimental task. Stimuli were presented on the computer
and participants responded by saying the correct syllables aloud . Participants were awarded a
14
point for each correct guess regarding which syllable carne next in the sequence. The generate
score was out of ten possible points with one correct answer and two foils (chance performance =
33%i. Generate scores were expected to approach chance levels.
Results
Accuracy
Analysis was completed using 3866 trials (54 blocks) for controls, 4047 trials (56 blocks)
for PWS and 3815 trials (53 blocks) for PPD in Session One. For Session One a 3 (Group) x 4
(Block 1 - 4) ANOVA revealed a significant Group main effect for accuracy, F (2,37) = 4.0,p
<. 01, ,:,1 = .22. No other significant effects were found. A between-groups Least Significant
Difference (LSD) post hoc test revealed PPDs' accuracy rate of 95.6% (0.9) across blocks was
significantly lower than that of control participants, 97.5% (0.4) and PWS', 98 .0% (0.7), as
shown in Figure 2.
Analysis was completed using 2392 trials (33 blocks) for controls, 2547 trials (35 blocks)
for PWS and 2451 trials (34 blocks) for PPD in the Retention Session. For the Retention Session
a 3 (Group) x 4 (Block 1 - 4) ANOVA revealed a significant Group main effect for accuracy, F
(2,32) = 4.4,p = .02, ,:,1= .22. A between-groups LSD post hoc comparison indicated PPD were
significantly less accurate than PWS and near-significantly less accurate than controls (p = .06)
across blocks. A significant Block main effect was also found, F (2,32) = 15.9, p = .00, ,:,1= .33,
indicating better accuracy over practice during the Retention Session. No other significant effects
were found. Two 3 (Group) x 2 (Condition: Ran vs. SEQ) ANOVAs revealed no significant
differences for Session One or the Retention Session.
Reaction time for general (non-specific) skill learning
For both Session One and the Retention Session, a log transformation was made of the
reaction time data and the Greenhouse-Geisser correction was made for all comparisons.
Analysis was completed using 864 trials for controls, 868 trials for PWS and 865 trials
for PPD in Session One. For Session One a 3 (Group) x 8 (Block; 8 means across practice
blocks) ANOVA revealed a significant Block main effect for reaction time, F (7,37) = 33.2,p =
.00, ,:,1 = .47. Reaction times decreased over practice blocks (see Tables 3, 4, and 5 in Appendix
G). A significant Group x Block interaction was also found , F (14,37) = 4.8,p = .01,':' = 21
(see post hoc below). No other significant effects were found .
Three 2 (Group) x 8 (Block) post-hoc ANOVAs (contrasting the reaction time ofPPD vs.
controls, PWS vs. PPD, and PWS vs. controls over practice blocks) revealed that the reaction
times of control participants improved more quickJy after practice relative to PWS, F (I, 25) =
11.2, p =. 00, ,:,1 = .31. No other significant effects were found.
Analysis was completed using 416 trials for controls, 416 trials for PWS and 427 trials
for PPD in the Retention Session. For the Retention Session complete data were only available
for 12 PWS, 12 PPD and 13 control participants. A 3 (Group) x 2 (Session One, Block 4 vs.
Retention Session, Block 1) ANOVA found no significant differences. Figure 3 depicts a more
detailed analysis for retention described under the "general learning vs . implicit learning"
heading in the Dependent Variables section. A 3 (Group) x 8 (Block) ANOVA revealed a
significant Block main effect for reaction time, F (7,34) = 12.5, p = .00, ,:,1 = .27. Reaction
15
times decreased significantly over blocks. A near-significant Group x Practice interaction was
also found, F (14, 34) = 2,7, P = .08, ~2 = .14, indicating PWS showed slower reaction times on
the first practice block of the Retention Session relative to the other groups" No other significant
effects were found.
Reaction lime for implicit (sequence-specific) learning
Analyses were designed to assess learning of the entire sequence (Sequence Blocks vs.
Random Blocks) and learning of the pair-wise associations or chunks within the sequence (SEQ
vs. RAN; Wilkinson & Jahanshahi, 2007).
Sequence Blocks vs. Random Blocks
Analysis was completed using 431 trials for controls, 433 trials for PWS and 431 trials for
PPD in Session One. For Session One a 3 (Group) x 2 (Condition) ANOVA compared the
groups' log-transformed reaction time on Block 4 (sequence) vs. Block 5 (random). Complete
data sets were obtained from 14 control participants, 13 PWS and 14 PPD. A significant
Condition main effect was found, F(I , 38) = 23.5 ,p = .00, ~2 = .38. Block 5 reaction times were
slower than Block 4 reaction times. No other significant effects were found ,
Analysis was completed using 400 trials for controls, 413 trials for PWS and 420 trials for
PPD in the Retention Session. For the Retention Session a 3 x 2 ANOVA compared the groups'
mean reaction time for Block 2 (sequence) vs, Block 3 (random). A significant Condition main
effect was found, F (I, 38) = 4. 1, p = .05, ~2 = .10. Block 3 reaction times were slower than
Block 2 reaction times. No other significant effects were found.
SEQvs. RAN
Analysis was completed using 427 trials for controls, 421 trials for PWS and 427 trials for
PPD in Session One. For Session One a 3 (Group) x 2 (Condition: SEQ vs. RAN) ANOVA
assessed the groups' log transformed reaction time for SEQ vs. RAN pairs. A significant
Condition main effect, F (1,39) = 9.0,p = .01 , ~2 = .19, and a significant Condition x Group
interaction was found, F (2 , 39) = 3.8, p = .03, ~ 2 = .17. RAN reaction times were slower than
SEQ reaction times. Three 2 (Group) x 2 (Condition) post hoc ANOVAs confirmed a significant
Condition x Group interaction for PWS vs. control participants, F (I, 26) = 7.3, P = .0 I, ~2 = .22,
and a similar near-significant trend for PPD vs. control participants, F (1,26) = 3.2, p = .08, ~2 =
.11 (see Figure 4). Control participants had much faster SEQ vs. RAN reaction times relative to
the other groups. No other significant effects were found (see Table 6 in Appendix G).
Analysis was completed using 388 trials for controls, 411 trials for PWS and 395 trials for
PPD in the Retention Session. On tbe Retention session a Condition main effect, F (I, 36) =
25.0,p = .00, ~2 = .46, was the only significant finding (see Table 7 in Appendix G). RAN
reaction times were slower than SEQ reaction times.
Explicit questionnaires and generate task
Control participants, PWS, and PPD scored 3.1 (2.1),3.4 (\.8), and 3.2 (\.8) out of 10,
respectively on the Day One questionnaire. Participant groups' scores did not significantly differ,
and all groups performed near chance levels (see previous section). Control participants, PWS,
and PPD scored 4.3 (2.1), 5.1 (1.4 ), and 5.2 (1.3) out of 10, respectively on the Day Two
questionnaire. Participant groups' scores did not significantly differ, and all groups performed
16
near chance levels. Control participants, PWS, and PPD score 3.7 (1.5), 3.6 (1.4), and 3.4 (\.7)
out of9 respectively for the generate task. Groups' scores did not significantly differ and all
groups perfonned near chance levels.
Discussion
Accuracy
In agreement with previous research (Wilkinson & Jahanshahi, 2007), PPD demonstrated
significantly poorer accuracy in general (non-specific) learning over practice in Session One and
in the Retention Session. It is important to note, however that, while statistically significant, all
groups perfonned with high levels of accuracy. PPD's accuracy levels were only a few
percentage points less than the other groups. Based on a comparison of PPD and control
participants on a sequence learning study, Dominey et a1. (1997) concluded that PPDs' implicit
learning appeared relatively intact only when knowledge of results is present to guide learning.
In concordance, Guadagnoli, Leis, Van Gemmer!, and Stelmack (2002) reported Parkinson
patients required 100% knowledge of results (relative to 20% for control participants) in order to
demonstrate commensurate retention of a simple timing movement. In the present study, we did
not manipulate "feedback" as an independent variable, which likely explains the reduction in
accuracy.
Reaction time for general (non-specific) learning
Reaction time and accuracy differences over practice represented general (non-specific)
learning (Smith & McDowall, 2004 ; Westwater et aI., 1998). While the figures in this paper
display group means for reaction time, it is important to point out that there was individual
variability within each group and considerable overlap between groups (see Appendix G).
All groups showed significant general learning, achieving a mean reaction time of
approximately 900 ms by Block 4 in Session One and by Block 2 in the Retention Session. This
result suggested that PPD and PWS were not "slower overall". Our choice of early stage,
medicated PPD, appears to have been successful in minimizing akinesialbradykinesia effects for
motor execution of the sequence. Our results indicated poorer accuracy for PPD (see Wilkinson
& Jahanshahi, 2007) instead of slower reaction times found in other studies (Smith & McDowall,
2004; Westwater et aI., 1998). This difference may have arisen due to our diligence in
identifying and removing disfluencies and outliers (akinetic speech freezing) from the reaction
time data. PPD had more than twice as many disfluencies relative to control participants and a
great deal more than PWS (see the dependent variables section). Our finding of a high number of
disfluencies in the data agrees with previous findings of high rates of disfluency for PPD as
previously reported (Benke et aI., 2008). This difference may also have arisen due to a difference
in difficulty level. Our learning task required participants to learn an arbitrary stimulus/response
association between a spatial location and a nonsense syllable while previous studies required
participants to name presented letters or numbers appearing in different spatial locations (e.g.,
"A" or "4"). It is important to empbasize that the current study used only early stage, non­
demented PPD and this is in contrast to other studies in the literature. This may have been an
important factor explaining this discrepancy with the existing literature.
PWS' difficulties were limited to reaction time differences in early trials in Session One,
which may be attributable to difficulty with acquisition of set. While PWS had a ' slow start',
17
they were not slower overall and demonstrated good nonspecific learning (a large improvement
in reaction time) during Blocks 2, 3 and 4 so that their reaction times became equivalent to those
of controls. This "slow start" was also noted for PWS during sequence learning tasks in two
previous studies (Smits-Bandstra, De Nil, & Saint-Cyr, 2006; Smits-Bandstra & De Nil, 2009)
and on a review of PWS' performance on reaction time studies in general (Smits-Bandstra,
20 I 0). During early trials it has been hypothesized that participants use compiled sensory
information to guide recognition and facilitation of appropriate preexisting learned movement
patterns (synergies) while inhibiting irrelevant ones (Saint-Cyr, 2003). The critical role of the
cortico-striato-thalamo-cortical circuit for this process is well established within the neuroscience
literature (Grafton, Hazeltine, & Ivry, 2002; Carbon & Eidelberg, 2006), as is PPDs' difficulty
with acquisition of set (Saint-Cyr, 2003). Peigneux and colleagues (2000) suggested that the
striatum "plays a significant role in the selection of the most appropriate responses in the context
created by both the current and previous stimuli" (p. 179).
Most interesting to note was PWS' recurring 'slow starts' on the Retention Session.
4
Based on an exploratory post-hoc analysis • results indicate there are reaction time differences on
these Retention Session early trials with some PWS not responding as quickly as participants in
other groups. One speCUlative explanation for this difficulty with acquisition of set is that the
sensory context and/or movement patterns for the syllables were not easily
accessible/identifiable for selection and facilitation by the cortico-striato-thalarno-cortical circuit
due to aberrant white matter physiology within the motor cortex found in previous studies (Beal,
Gracco, Lafaille, & De Nil, 2007; Foundas, Bollich, Corey, Hurley, & Heilman, 2001; Sommer,
Koch, Paulus, Weiller, Buchel, 2002).
The results of the current study may partially explain why PWS do so well in the therapy
room but struggle when attempting to re-access learned patterns in transfer situations or in the
long term. Clearly further research is needed in this area, as an important limitation of the current
study is the lack of generalization from our visual , nonsense syllable task to conversation speech
or typical speech skills taught in speech therapy.
Reaction time for implicit (sequence-specific) learning
Sequence Blocks vs. Random Blocks
For Session One, reaction time slowing was observed for all groups upon introduction of
the random trials in Block 5 and for the Retention Session, upon introduction of the random trials
in Block 3. These findings indicated recognition of the implicit sequence in both Session One
and the Retention Session for all the groups. In contrast to previous studies PPD and control
participants did not differ on these sequence block vs. random block, sequence-specific contrasts
(Smith & McDowall, 2004; Westwater et al. 1998).
It should be noted that differences in reaction time between sequence blocks and random
blocks were typically only 40 ms smaller for PPD relative to controls in previous studies, a small
difference. In the current study existing smaU group differences may have been masked because
this paradigm had a reduced total number of practice trials relative to previous studies. In
addition, group differences may not have reached statistical significance because the current
study recruited early-stage, mildly affected PPD relative to other studies. As an alternative
interpretation it must be considered that the syllable task in the Current study may not have been
as heavily reliant on implicit memory as previous tasks. Speech, more than other movements,
likely involves a complex interaction of implicit and explicit systems. This may explain, in part,
18
why speech movements are relatively preserved in many PPD relative to other movements and
why speech problems are relatively resistant to dopaminergic treatments.
SEQvs. RAN
For both Session One and the Retention Session, SEQ reaction times were significantly
faster than RAN reaction times indicating that our experimental manipulation was successful in
eliciting implicit learning (Nissen & Bullemar, 1987). As expected, control participants
demonstrated faster reaction times for SEQ vs. RAN syllable pairs relative to PWS (significant
difference) and PPD (non-significant trend) in Session One. This finding was interpreted as more
efTective implicit sequence learning by controls relative to the other groups over practice. PPD
showed the expected pattern of slower reaction times on implicit (sequence-specific) learning
tasks relative to controls found in previous studies (Smith & McDowall, 2004; Westwater et aI.,
1998). However this difference was limited to a non-significant trend.
Surprisingly, as a group, PWS' pattern of implicit learning differed significantly from
controls (see Figure 4). One postulation is that PWS used excellent explicit and/or general
learning skills to compensate for "slow starts", and either effectively compensated for implicit
learning deficits or did not demonstrate deficits in implicit learning at all. The current study was
successful in assessing implicit learning to some degree but was limited in that there was no
opportunity to observe changes in performance as explicit knowledge developed nor was there
an opportunity to contrast implicit vs. explicit knowledge of the same task (e.g., having some
groups know about the sequence while other groups do not).
Groups did not difTer in implicit learning on the Retention Session suggesting PWS and
PPD had good retention of implicitly learned skills. This finding is in agreement with
comparable studies with PPD that examined retention over several days and found minimal or no
impairment in retention (Harrington et aI. , 1990). Alternatively, this result may have been
attributable to a ceiling efTect. Controls attained reaction times of approximately 900 ms in
Session One and no group attained reaction times faster than 900 ms subsequently for the
pseudo-random blocks.
Explicit questionnaires and generate task
It has been shown that greater explicit/declarative knowledge is associated with better
performance in healthy participants (Mayr, 1996), although not for PPD (Boyd & Winstein,
2006). No groups showed explicit knowledge for Session One or the Retention Session,
indicating better sequencing performance by any group was not likely due to a greater degree of
explicit knowledge . It has been argued that recall tests after the fact do not accurately reflect true
explicit knowledge (Russeler & Rosier, 2000) therefore we included the generate task (similar to
that used by Peigneux et aI., 2000). For both Session One and the Retention Session, generate
scores were equivalent to chance and did not differ from questionnaire scores.
Theoretical implications
Several studies have been conducted to investigate the neural correlates of intensive
training of speech " loudness" training in the LSVT program (Fox, Ramig, Ciucci, Sapir,
McFarland, & Farley, 2006; Narayana et aI., 2010). These studies seem to suggest that while
treatment is effective in changing behavior, the training required to institute such a change is
19
intensive, supplemented by many types of feedback and frequent knowledge of results, and is not
accompanied by the expected neural changes in the primary motor and sensory cortices (Karni et
aI., 1998). Instead, these studies report increased neural activity in areas associated with " top
down" (explicit) attention and monitoring, suggesting compensation and/or explicit strategy use
rather than implicit learning per se (Fox et aI., 2006; Narayana et aI., 2010). This emphasis on
monitoring may have evolved as a response to difficulties patients have transitioning to implicit
learning as well as retaining proficiency for new speech patterns. As defined in the introduction,
implicit learning, once established is relatively automatic and feed fOIWard (not reliant upon
feedback). Several researchers have presented evidence ofPPOs' (Dominey et aI., 1997;
Fattapposta et aI., 2002) and PWS' (Civier, Tasko, & Guenther, in press; Max, 2004) difficulty
with feed fOIWard motor control. Results of the current study provide preliminary support for
differences, but not necessarily deficits, in the use of feed fOIWard or implicit learning by PWS
and PPO.
11 can not be discounted that, for both PPO and PWS, the stability of the system on any
given day may detennine how welileamed sequences can be expressed as long sequences
planned in advance of the movement (Smiley-Oyen, Lowry & Kerr, 2007). Implicit learning and
feed fOIWard control may, in fact, be intact, but are only incompletely or sporadically expressed
due to instability of the motor planning/execution systems. This tentative proposal requires
further investigation.
11 must also be mentioned that sequential learning by PWS or PPO might have been
affected by differences in the focus of attention, especially in tbe beginning of the test. If
attention and cognitive processing resources required for implicit sequence learning are diverted
for speech motor execution, reduced implicit sequence learning results. Both PPO and PWS have
demonstrated difficulties in the face of dual task demands (see Smits-Bandstra & Oe Nil, 2007
for a review) indicating a need for further study of the effects of attention in this area.
Conclusion
The purpose of this study was to compare the functional integrity of procedural learning
and implicit memory systems in PPO, PWS, and control participants. In agreement with previous
studies, the current preliminary results suggest there are differences with some PPO doing not as
well as individuals in other groups on the implicit (sequence-specific) learning tasks used in the
current study (non-significant trend). Although requiring replication, the current study suggests
PPO did not demonstrate a retention deficit for implicit learning of the current serial reaction
time task (within 1-2 weeks of the Session One) and this is consistent with previous study
findings (Harrington et aI., 1990). Accuracy results also suggested that some PPO were not as
accurate as individuals in other groups when general (non-specific) learning and retention were
assessed.
Some PWS showed a pattern of implicit learning quite different from individuals in other
groups. The groups' divergent results provide preliminary evidence to negate Parkinson ' s-like
basal ganglia damage as a possible etiological factor of stuttering. However these results do not
precluding Giraud et al.'s (2008) hypothesis of cortical pathology perpetuated by cortico-striato­
thalamo-cortical connections associated with stuttering. PWS showed slow reaction times for
initial trials of Session One and the Retention Session but "caught up" after practice,
demonstrating that some PWS manifest delayed but not deficient general (non-specific) learning
relative to individuals in other groups. The difficulty some PWS showed for acquisition and "re­
20
acquisition of set" implicated cortico-striato-thalamo-cortical circuits known to be critical for
this skill (Peigneux et a!., 2000).
The findings of the current study are sufficiently promising to embolden further research
investigating implicit learning and retention in populations with motor speech disorders. As a
next step, it will be necessary to investigate how the manipulation of variables such as explicit
instruction, practice, and feedback will influence learning and retention in disordered
populations. Future research must also have greater generalizability to functional speech tasks
and speech treatment techniques than the current study.
21
Acknowledgements
We would like to acknowledge the Canadian Institute of Health Research for funding
provided to the first author and a National Science and Engineering Research Council of Canada
Grant provided to the second author. We would like to extend a special thanks to Whitney
Holman, Amy Crist, Colby Peterson and Sandra Stenerson for their detailed feedback and careful
editing.
22
References
Abbruzzese, G., Trompetto, C., Marinelli, L. (2009). The rationale for motor learning in Parkinson' s
disease. European Journal ojPhysical Rehabilitative Medicine. 45, 2, 209-214.
Agostino, R., CUITa, A., Soldati, G., Dinapoli, L., Chiacchiari, L., Modugno, N., et al. (2004). Prolonged
practice is of scarce benefit in improving motor performance in Parkinson's disease. Movement
Disorders, 19, 1285-1293.
Beal, D., Gracco, V., Lafaille, S. & De Nil, L. F. (2007). VoxeJ-based morphometry of auditory and
speech-related cortex of stutterers. NeuroReport, 18. 1257-1260.
Beck, E. , Epstein, N., Brown, G., & Steer. R. A. (\988). An inventory for measuring clinical anxiety :
Psychometric properties. Journal ojConsulting & Clinical Psychology, 56, 893-897.
Benecke, R., Rothwell, J. C., Dick, J. P. R., Day, B. L. , Marsen, C. D. (1986). Performance of
simultaneous movements in patients with Parkinson's disease, Brain, 109,739-757. Benke, Th., Hohenstein, C., Poewe, W., & Butterworth, B. (2000) . Repetitive speech phenomenon in Parkinson's disease. Journal ojNeurology, Neurosurgery, & Psychiatry, 69,319-324. Blomgren, M., Roy, N., Callister, T., & MeITil, R. M. (2005). Intensive stuttering modification therapy :
A multidimensional assessment of treatment outcomes. Journal ojSpeech, Language, and
Hearing Research, 48,509-523.
Boyd, L., & Winstein, C. (2006). Explicit information interferes with implicit motor leaming of both
continuous and discrete movement tasks after stroke. Journal ojNeurologic Physical Therapy,
30,46-57.
Brady, J. P. (1998) Drug-induced stuttering: A review of the literature. Journal ojClinical
Pharmacology, 18,50-54.
23
Brashers-Krug, T., Shadmehr, R., & Bizzi, E. (1996). Consolidation in human motor memory. Nature,
382,252-255.
Burghaus, L., Hilker, R., Thiel, A., Galldiks, N, Lehnhardt, F. G., Zaro-Weber, O. et al. (2006). Deep
brain stimulation of the subthalamic nucleus reversibly deteriorates stuttering in advanced
Parkinson's disease. Journal ofNeural Transmission, 113,625-631 .
Carbon, M., & Eidelberg, D., (2006). Functional imaging of sequence learning in Parkinson's disease.
Journal ofNeurological Science, 248,72-77.
Civier, 0., Tasko, S., Guenther, F. H. (in press). Overreliance on auditory feedback may lead to
sound/syllable repetitions: simulations of stuttering and fluency-inducing conditions with a
neural model of speech production. Journal ofFluency Disorders.
Cox, M. D. (1982). The stutterer and stuttering: Neuropsychological correlates. Journal ofFluency
Disorders, 7,129-140.
Curran, T., & Keele, S. (1993). Attentional and nonattentional forms of sequence learning. Journal of
Experimental Psychology: Learning, Memory & Cognition, 19, 189-202.
Dominey, P. F., Ventre-Dominey, J., Broussolle, E., & Jeannerod, M. (1997). Analogical transfer is
effective in a serial reaction time task in Parkinson's disease; evidence for a dissociable form of
sequence learning. Neuropsychologia, 35, 1-9.
Doyon, J. & Benali, H. (2005). Reorganization and plasticity in the adult brain during learning of motor
skills. ClITrent Opinion in Neurobiology, 15, 161-167.
Doyon, J., Laforce, R., Bouchard, G., Gaudreau, D., Roy, J., Poirier, M. et al. (1998). Role of the
striatum, cerebellum and frontal lobes in the automatization of a repeated visuomotor sequence
of movements. Neuropsychologia, 36, 625-641.
24
DuffY, J. R. (2005). Motor speech disorders: Substrates, differential diagnosis, and management (2 nd
Ed.). St. Louis, MO: Elsevier Mosby.
Eimer, M., Goschke, T., Schlaghecken, F., & Stunner, B. (1996). Explicit and implicit learning of event
sequences: Evidence from event-related brain potentials. Journal
0/ Experimental Psychology:
Learning, Memory, & Cognition, 22,970-987.
Fahn, S., Elton, R. L. & Members of the UPDRS Development Committee (1987). Unified Parkinson's
Disease Rating Scale. In S. Fahn, C. D. Marsden, D. B. Caine, & M. Goldstein (Eds.) , Recent
development in Parkinson 's disease (pp. 153- 163). Florham Park, NJ: Macmillan Healthcare
Infonnation.
Fattapposta, F., Pierelli, F., My, F., Mostarda, M., Del Monte, S., Parisi, L., et al. (2002). L-dopa effects
on preprogramming and control activity in a skilled motor act in Parkinson's disease . Clinical
Neurophysiology, 113,243-253 .
Feigin, A, Ghilardi, M.F., Carbon, M. Edwards, C., Fukuda, M, Dhawan, Y., et al. (2003). Effects of
levadopa on motor sequence learning in Parkinson's disease. Neurology, 60, 1744-1749.
Folstein, M. F., Folstein, S. E., & McHugh, P. R. (1975). "Mini-mental state". A practical method for
grading the cognitive state of patients for the clinician. Journal o/Psychiatric Research, 3,189­
98
Foundas, A. L., Bollich, A. M., Corey, D. M., Hurley, M., & Heilman, K. M. (2001) . Anomalous
anatomy of speech-language areas in adults with persistent developmental stuttering. Neurology,
5 7,207-215.
Fox, C. M., Ramig, L. 0 ., Ciucci, M. R. , Sapir, S., McFarland, D., H. & Farley, B. G. (2006). The
science and practice of LSYTILOUD: neural plasticity-principled approach to treating
25
individuals with Parkinson disease and other neurological disorders. Seminars in Speech
Language Pathology; 27, 4,283-299.
Ghilardi, M. F., Feigin, A. S., Banaglia, F., Silvestri, G., Mattis, P. , Eidelberg, D., & Di Rocco, A.
(2007). L-Dopa infusion does not improve explicit sequence learning in Parkinson's disease .
Parkinsonism and Related Disorders, 13, 146-151.
Grafton, S. T., Hazeltine, E., & Ivry, R. B., (2002). Motor sequence learning with the nondominant left
hand: A PET functional imaging study . Experimental Brain Research, 146,369-378.
Guadagnoli, M. A., Leis, B., Van Gemmert, A. W., & Stelmach, G. E. (2002). The relationship between
knowledge of results and motor learning in Parkinsonian patients. Parkinsonism and Related
Disorders, 9,89-95.
Harrington, D. L., Haaland. K. Y. , Yeo, R. A., & Marder, E. (1990). Procedural memory in Parkinson's
disease: Impaired motor but not visuo perceptual learning. Journal o/C1inical & Experimental
Neuropsychology, 12,323-339.
Hauptmann, B., & Karni, A. (2002). From primed to learn: The saturation of repetition priming and the
conduction of long-term memory . Cognitive Brain Research, /3,313-322.
Hoehn, M. & Yah!, M. (1967). Parkinsonism: onset, progression and mortality. Neurology. 17, 427-442.
Karni, A. (1996). The acquisition of perceptual and motor skills: A memory system in the adult human
cortex. Cognitive Brain Research. 5, 39-48.
Karni, A., Meyer, G., Rey-Hipolito, c., Jezzard, P., Adams, M., Turner, R. & Ungerleider, L. (1998).
The acquisition of skilled motor performance: Fast and slow experience-driven changes in
primary motor cortex. Proceedings National Academy 0/ SCience, 95, 861-868.
26
Kawai , H., Kawamura, M., & Kawachi , J. (1999). Procedural memory in a patient with progressive
supranuclear palsy and Alzheimer's disease: impairment of long-term retention following
subcortical degeneration. Japanese Journal ofNeuropsychology, 15,229-237 (in Japanese).
Kent, R. D. (2000). Research on speech motor control and its disorders: A review and prospective.
Journal ofCommunication Disorders, 33,391-428.
Koller, W. C. (1983) Dysfluency (stuttering) in extrapyramidal disease. Archives ofNeurology, 40, 175­
177.
Louis, E. D., Winfield, L., Fahn, S., & Ford, B. (2001). Speech dysfluency exacerbated by Levodopa in
Parkinson's disease. Movement Disorders, 16, 562-581.
Ludlow, C. L., & Loucks. T. (2003). Sruttering: A dynamic motor control disorder. Journal ofFluency
Disorders, 28,273-295.
Max, L. (2004). Stuttering and internal models for sensorimotor control: A theoretical perspective to
generate testable hypotheses. In B. Maassen, R. D. Kent, H. F. M. Peters, P. H. H. M. van
Lieshout, & W. Hulstijn. Speech motor control in normal and disordered speech (pp. 357-388).
Oxford, NY: Oxford University Press.
Mayr, U. (1996). Spatial attention and implicit sequence learning: evidence for independent learning of
spatial and nonspatial sequences. Journal ofExperimental Psychology, Learning Memory &
Cognition, 22,350-364.
Mochizuki-Kawai, H., Kawamura, M., Hasegawa, Y., Mochizuki , S., Oeda, R., Yamanaka, K., Tagaya,
H. (2004). Deficits in long-term retention of learned motor skills in patients with cortical or
subcortical degeneration. Neuropsychologia, 42, 1858-1863.
Muslimovic, D., Post, B., Speelman, J. D., & Schrnand, B. (2007). Motor procedural learning in
Parkinson's disease, Brain, 130, 11,2887- 2897.
27
Nagy, H., K'eri. S., Myers, C. E., Benedek, G., Shohamy, D., & Gluck, M. (2007). Cognitive sequence
learning in Parkinson's disease and amnestic mild cognitive impairment: Dissociation between
sequential and non-sequential learning of associations. Neuropsychologia, 45, 1386-1392.
Narnasivayarn, A. K., & van Lieshout, P. (2008). Investigating speech motor practice and learning in
people who stutter. Journal ofFluency Disorders, 33, 32-51.
Narayana, S., Fox, P. T., Zhang, W., Franklin, C., Robin, D. A ., Vogel, D., & Ramig, L. O. (2010).
Neural correlates of efficacy of voice therapy in Parkinson's disease identified by performance­
correlation analysis. Human Brain Mapping, 31, 2,222-236.
Neurobehavioral Systems (2004). Presentation (Version 0.80) [Computer software]. Albany, CA:
NeuroBehavioral Systems Inc.
Nissen, M., & Bullemer, P. (1987). Attentional requirements of learning: evidence from performance
measures. Cognitive Psychology, 19, 1-32.
Oldfield, R. (1971). The assessment and analysis of handedness: The Edinburgh Inventory.
Neuropsychologia, 9,97-113.
Peigneux P, Maquet P, Meulemans T, Destrebecqz A, Laureys S, Degueld.re C, Delfiore G, Aerts J,
Luxen A, Franck G, et al. (2000). Striatum forever, despite sequence learning variability: a
random effect analysis of PET data. Human Brain Mapping, 10, 179-194.
Portney, L. G., & Watkins, M. P. (1993). Foundations ofclinical research: Applications to practice.
Stamford, CN: Appleton & Lange.
Riley, G. (2009). A stuttering severity instrument (4th Ed.). Austin, TX: Pro-Ed.
Russeler, J., & Rosier, F. (2000). Implicit and explicit learning of event sequences: evidence for distinct
coding of perceptual and motor representations. Acta Psychologica, 104,45-67.
28
Saint-Cyr, J. A. (2003). Frontal-striatal circuit functions: Context, sequence, and consequence. Journal
ofthe International Neuropsychological Society. 9, 103- 127.
Saint-Cyr, J., Taylor, A., & Lang, A. (1988). Procedural learning and neostriata I dysfunction in man.
Brain. 111,941-959.
Sanes, J. N., Dimitrov, B., Hallett, M. (1990). Motor learning in patients with cerebellar dysfunction.
Brain, 113, I, 103-120.
Schmidt, R. A., & Wrisberg, C. A. (2004). Motor learning and performance (3,d ed.). Champaign, IL:
Human Kinetics Publishers.
Schulz, G. M., Leon, S., & Sulc, S. (2001). Speech motor learning in Parkinson's disease. In B.
Maassen, W. Hulstijn, R. D. Kent, H. F. M. Peters, & P. H. H. M. van Lieshout (Eds.) Speech
motor control in normal and disordered speech: Proceedings from the fourth international
speech motor conference (pp. 332-336). Nijmegen, the Netherlands: Uitgeverij Vantilt.
Shadmehr, R., & Brashers-Krug, T. (1997). Functional stages in the formation of human long-term
motor memory . Journal ofNeuroscience, 17,409-419.
Shahed, J. & Jankovic, J. (200 I). Re-emergence of childhood stuttering in Parkinson's disease: A
hypothesis. Movement Disorders, 16, 114-118.
Siegert, R. J., Taylor, K. D., Weatherall, M., & Abernethy, D. A. (2006). Is implicit sequence learning
impaired in Parkinson's disease? A meta-analysis. Neuropsychology, 20,490-495.
SIL (2007). Speech Analyzer (Version 3.0.1.) [Computer Software). Dallas, TX; SIL International.
Smiley-Oyen, A. L., Lowry, K. A., & Kerr, J. P. (2007). Planning and control of sequential rapid aiming
in adults with Parkinson'S disease. Journal ofMotor Behavior, 39, 103-114.
29
Smith, A., Sadagopan, N., Walsh, B., Weber-Fox, C. (2010). Increasing phonological complexity
reveals heightened instability in inter-articulatory coordination in adults who stutter. Journal of
Fluency Disorders, 35, 1-18.
Smith, J. G., & McDowall, J. (2004). Impaired higher order implicit sequence learning on the verbal
version of the serial reaction time task in pat ients with Parkinson's disease. Neuropsychology, 18,
679-691.
Smits-Bandstra, S. (2010). Methodological considerations in the measurement of reaction time of
persons who stutter. Journal ofFluency Disorders, 35, 19-32.
Smits-Bandstra, S., & De Nil, 1. F. (2007). Sequence skill learning in persons who stutter: implications
for cortico-striato-thalamo-cortical dysfunction. Journal ofFluency Disorders, 32,251-278.
Smits-Bandstra, S ., & De Nil, L. F. (2009). Speech skill learning of persons who stutter and fluent
speakers under single and dual task conditions. Clinical Linguistics and Phonetics, 23,38-57.
Smits-Bandstra, S., De Nil, L. F., & Rochon, E. (2006). The transition to increased automaticity during
finger sequence learning in adult males who stutter. Journal ofFluency Disorders, 31,22-42 .
Smits-Bandstra, S., De Nil, 1. F., & Saint-Cyr, 1. A. (2006). Speech and non speech sequence skill
learning in adults who stutter. Journal ofFluency Disorders, 31,116-131.
Sommer, M. , Koch, M. A., Paulus, W. , Weiller, C., & Buchel, C. (2002). Disconnection of speech­
relevant brain areas in persistent developmental stuttering. Lancet, 360, 380--383.
Vakil, E., & Herishanu-Naaman, S. (1998). Declarative and procedural learning in Parkinson's disease
patients having tremor or bradykinesia as the predominant symptom. Cortex, 34,611-620.
Weschler, D. A. (1997). Wechsler Adult Intelligence Scale-lJI. New York: Psychological Corporation.
30
Westwater, H., McDowall, J., Siegert, R., Mossman, S., & Abernethy, D. (1998). Implicit learning in
Parkinson's disease: Evidence from a verbal version of the serial reaction time task. Journal of
Clinical & Experimental Neuropsychology, 20, 413-418.
Wikipedia. (2011) Procedural Memory. Retrieved on Oct 1,2011 from
http://en.wikipedia.orglwikilProcedural_memory.
Wilkinson, L., & Jahanshahi, M. (2007). The striatum and probabilistic implicit sequence learning.
Brain Research, 1137, 117-130.
Yairi, E., & Seery, C. H. (201 )). Stullering: Foundations and Clinical Applications. Upper Saddle
River, NJ: Pearson Education Inc.
Yorkston, K. M., Beukelman, D. R., Strand, E. A., & Hakel, M. (20) I). ManagemenJ ofMotor Speech
Disorders in Children and Adults (3rd Ed.). Austin, TX: Pro-Ed.
31
Appendix A
Dysarthria Rating Scale
_ _ _ _ _ _ _ _(rater)
_ _ _ _ _ _ _ _ _ _ (participant code)
Pitch
Low
Normal
Variation of Pitch
Normal
Lack of variation
Steadiness of Pitch
Normal
Tremor
Loudness
Soft
Normal
Loud
Rate of Speech
Slow
Normal
Fast
StresslEmphasis Pattern
Normal
Excess stress
Nasality
Hyponasal
Normal
HypernasaJ
Laryngeal
Harsh
Normal
Breathy
Articulatory Precision
Normal
Reduced clarity
High
Intelligibility
Percentage of clear words:
95-100%
85-95% 50-85%
All clear
Most clear
< 50%
More than Y:z clear
Less than Y:z clear
32
AppendixB
Unified Parkinson's Disease Rating Scale
(Fahn, Elton, & Members of the UPDRS Development Committee, 1987)
II. Activities of Daily Living
5. Speec"
o= Normal. I = Mildly affected. No difficulty being understood . 2 = Moderately affected. Sometimes asked to repeat statements. 3 = Severely affected. Frequently asked to repeat statements. 4 = Unintelligible most of the time. Modified Hoehn and Yahr Staging
(Hoehn & Yahr, 1967)
STAGE 0 = No signs of disease . STAGE I = Unilateral disease . STAGE 1.5 = Unilateral disease plus axial involvement. STAGE 2 = Bilateral disease, without impairment of balance. STAGE 3 = Mild bilateral disease, with recovery on pull test. STAGE 4 = Severe disability, still able to walk or stand unassisted. STAGE 5 = Wheelchair bound or bedridden unless aided . ])
Appendix C Day One Questionnaire 1) Did you notice anything about the syllables? YES NO. If yes, what did you notice?
2)
Circle the number that fits best with what you remember about the
syllables .
The syllables
Always
appeared
In this order
1
The syllables
Often
appeared
In this order
2
The syllables
Sometimes
appeared in
This order
3
1) PO PI 2
3
4
5
2) PE PA 2
3 4
5
3)PIPA 2
3
4
5
4) PE PO PE 2
3
4
5
5) PA PO PE 2
3
4
5
6) PI PA PE 2
3
4
5
7) PE PO PA 2
3
4
5
8) PE PA PO PE 2
3 4
5
9) PA PO PI PA 2
3
4
5
10) PI PO PE PI 2
3
4
5
The syllables Rarely appeared In this order 4
The syllables
Never
appeared In this order 5
J'
Appendix D - page one
Day Two Questionnaire
1) Did you notice any1hing about the order of the syllables? YES NO. If
yes, what did you notice?
2) What number describes your belief about the syllables during the experiment? Very sure
Ihe syllables
sometimes
appeared in
a
predictable
order
Somewhat
sure the
syllables
sometimes
appeared in
a predictab le
order 1
2
Guess the
syllables
sometimes
appeared in
a prediclable
order
Somewhat
Very sure
sure the
syllables
Ihe syllables
random
appeared in
random
random
order order
order Guess Ihe
syllables
appeared in
a
appeared in
4
3) Sometimes the syllables were part of a repeating sequence and they occurred in a pred ictable order. Put the syllables in the correct order as well as you can . (PA, PA, PE, PE, PI, PI, PO, PO)
6
35
Appendix D - page two
4) Circle the number that fits best with what you remember about the
syllables.
part of the
Guess this
is part of
the
Guess this
is not part
of the
sequence
sequence
sequence
Very sure
Somewhat
this is part
of the
sure this is
sequence
Somewhat
sure this
is not part
of the
Very sure
this is not
part of the
sequence
sequence
4
A) PO PA
2
3 4
5 6
8) PE PO
2
3 4
5 6
C)PA PE
2
3 4
5 6
D)PIPE 2
3
4
5 6
E) PO PI PO 2
3
4
5 6
F) PE PI PA 2
3
4
5
6
G)PAPE PA 2
3
4
5
6
H)PI PO PA 2
3
4
5
6
I) PO PI PO PE 2
3
4
5
6
J) PE PI PA PE 2
3
4
5 6
K) PA PE PA PO 1 2
3
4
5 6
L) PI PO PA PE 1 2
3 4 5 6
5
6
36
Appendix E
Session One
Sequence Block repeated 4 x (Blockl, Block 2, Block 3 & Block 4) = Total 288 trials
PO PI PO PE PI PA PE PA
J
1
9
10
~
II
12
PO PE PI
17
18
j
,
PO PE PI PA PE PA * - Break
PO PI
19
14
IJ
15
16
PA PE PA PO PI
20
21
22
23
24
PO PE PI
PA PE PA PO PI - Break
25
28
26
27
29
30
3I
32
PA PE P A PO PI
PO PE PI
33
38
34
35
36
37
39
40
PA PE PA PO PI
PO PE PI
41
46
42
43
44
45
47
48
PA PE PA PO PI
PO PE PI - Break
49
54
50
5I
52
53
55
56
PI PA PE PA PO PI PO PE
n
~
H
00
~
~
M
~
PI PA PE PA PO PI PO PE
6S
66
67
68
RAN I RAN 2 SEQ I SEQ2
69
70
71
72
RAN 3 SEQ 3 SEQ4
RAN 4
PI PO PE PA PO PI PA PE - Break
S
5 SEQ 6 SEQ 7 SEQ 8 SE 9 SEQ lOS
PA PO PI
PI
II SEQ 12
PI
RAN 5 SEQ 13 SEQ 14 SEQ 15 RAN 6 SEQ 16 RAN 7 SEQ 17
PO PE PI
SE
PO PE PI
PA PE - Break
18 SEQ 19 SEQ 20 SEQ 21 SEQ 22 SEQ 2J S
PO PE PA PO PE
RAN8 S
26 S
PE
27 SEQ 28 SEQ 29 SE 30 RAN 9 SEQ31
PA
S
24 SEQ 25
PA PE
PO
J2 RAN 10 RAN II SEQ J 3 SEQ 34 RAN 12 RAN IJ SEQ 35
PO
PO
RAN 14 RAN 15 SEQ 36 RAN 16 S
PI PA
PO PE - Break
J7
17 SEQ 38 S
39
PO PI
RAN 18 RAN 19 SEQ 40 SEQ 41 RAN 20 SEQ 42 SEQ 43 RAN 21
PE PI
PA PE P A PO PI
RAN 22 SEQ 44 SEQ 45 SEQ 46 SEQ 47 SEQ 48 SEQ 49 SEQ 50
Day One Questionnaire
Note. *Statistical analysis included mean reaction time for the first two sequences of each
sequence block. **
- indicates random items (RAN) which violate the sequence order
37
Appendix F
Retention Session (Session Two)
Sequence Block repeated Ix (Block 1 & Block 2)
=
Total 144 trials
PE PA PO PI PO PE PI ,PA
J
4
6
j
PEPA PO PI POPE PI PA*-Break
II
10
12
PO PE PI
PI
19
20
17
18
PI
PO PE PI
25
26
27
35
34
21
22
42
29
43
50
36
38
37
44
45
51
52
59
58
67
66
RAN I RAN2 SEQ I
PI
SE
PO
40
60
68
SEQ7
55
56
PO PE PI
62
61
63
64
PO PE PI
70
69
SEQ2 RAN3 SEQ3
SEQ8
48
PO PE PI
PO PE PA PO PI
5 SEQ6
PO - Break
47
54
53
PA PE PA PO PI
65
32
39
46
PA PE PA PO PI
57
31
30
PA PE PA PO PI
49
24
2J
PA PE PA PO - Break
PA PE PA PO PI
PE PI
41
16
15
PA PE PA PO PI
PE PI
33
28
14
13
PA PE PA PO
71
72
SEQ4
RAN4
PA PE - Break
SE 9 SEQ 10 S
PA PO PI
PI
II SEQ 12
PI
RAN 5 SEQ 13 SEQ 14 SEQ 15 RAN 6 SEQ 16 RAN 7 SEQ 17
PO PE PI
S
PO PE PI
PA PE - Break
18 SEQ 19 SEQ 20 SEQ 21 SEQ 22 SEQ 2J
PO PE PA PO PE
RAN 8 S
26 SE 27 SEQ 28 SEQ 29 S
PA
30
SEQ31
PO
PA PE
SEQ35
PE - Break
PO
RAN 14 RAN 15SEQ36 RAN 16S
PI
PA
37 RAN 17SEQ38 SE
39
PO PI
RAN 18 RAN 19SEQ40 SEQ 41 RAN 20 SEQ 42 SEQ43 RAN21
PE PI
P A PE P A PO PI
RAN 22 SEQ 44 SEQ 45 SEQ 46 SEQ 47 SEQ 48 SEQ 49 SEQ 50
Generate Task = Total 10 trials
PE PA ? PO PI ? PE PI
PA PE PA ? PA PO PI ?
PE PI PA PE ? PI PA PE
PI PO PE PI ? PI PA PO
?
PI PO PE ?
PA ?
PI ?
Day Two Questionnaire
Note. *Statistical analysis included mean reaction time for the first two sequences of each
sequence block. **
- indicates random items (RAN) which violate the sequence order
Sequence Learning and Retention I
Appendix G: Table 3 Mean (SD) Reaction Time in milliseconds for Control Participants (C) Group
Session One Blocks 1 & 4
Differences b
Retention Session Blocks 1 & 2
l.l1:ID '
.1l'Z.:1ll
4 (l-8)
C1
1087 (325)
899 (169)
C2
934 (060)
957 (090)
C3
1272 (121)
1092 (242)
C4
1033 (282)
913 (134)
4 (9-16)
l.l1:ID
1 (9-16)
2 (1-8)
2 (9-16)
Diff 1
Diff2
882 (090)
785 (164)
751 (099)
803 (137)
919 (124)
831 (134)
302
-34
-80
820 (080)
808 (087)
725 (123)
700 (106)
583 (091)
637 (224)
126
-83
88
1239 (143)
1193 (167)
859 (107)
652 (145)
740 (202)
680 (160)
80
-334
179
943 (241)
983 (098)
875 (103)
869 (070)
792 (084)
766 (058)
50
-108
109
Diff3
C5
1132 (201)
913 (179)
920 (170)
895 (151)
1015 (214)
1003 (226)
1025 (076)
965 (165)
237
120
50
C6
859 (248)
1176 (296)
1050 (341)
1158 (185)
858 (156)
853(210)
728 (115)
794 (120)
-299
-301
64
C7
896 (232)
759 (091)
790 (156)
939 (064)
827 (116)
598 (278)
782 (147)
723 (120)
-43
-112
104
C8
1223 (183)
1246 (150)
1073 (101)
1047 (175)
981 (364)
1062 (217)
176
-66
981
-68
C9
940 (225)
852 (078)
953 (187)
836 (154)
878 (334)
941 (180)
892 (073)
946 (164)
104
43
C10
1034 (256)
1000 (144)
791 (274)
804 (226)
797 (305)
742 (096)
779 (098)
822 (067)
230
-7
-26
C11
850 (084)
791 (127)
760 (153)
830 (126)
998 (391)
1055 (462)
1286 (231)
1240 (166)
21
168
-242
C12
930 (187)
797 (180)
1119 (215)
1004 (273)
874 (339)
882 (150)
906 (159)
867 (145)
-74
-130
7
C13
1111 (271)
1048 (138)
815 (089)
783 (140)
894 (361)
1025 (215)
848 (161)
931 (160)
328
111
-37
C14
888 (202)
558 (290)
656 (140)
649 (138)
870 (331)
1015 (074)
971 (216)
1017 (221)
240
221
-147
MEAN
1014 (206)
929 (165)
915 (170)
908 (153)
871 (239)
871 (183)
866 (137)
863 (146)
106
-37
70
'(1-8) indicates Mean (SD) of trials 1 though 8, (9-16) indicates Mean (SD) of trials 9 through 16. b Diff 1 = Block 1 (1-8) - Block 4 (9-16), Diff 2 = Ret Block 1 (1-8)- Block 4 (9-16), Diff 3 = Block 1 (1-8) - Ret Block 2 (9-16) Sequence Learning and Retention· J
Appendix G: Table 4 Mean (SD) Reaction Time (milliseconds) [or Persons who Stutter (PWS) Group
Session One BlockS 1 & 4
1 (1-8] '
1 (9-16J
Differences b
Retention Session Blocks 1 & 2
4 [1-8J
4 [9-16]
1 (1-8)
1 [9-16)
2 (1-8J
2(9-16)
Oiff 1
OifCZ
Diff3
PWS1
1353 (247)
1373 (284)
1049 (111)
1031 (082)
1191 (165)
989 (196)
1098 (160)
1011 (131)
321
160
180
PWS2
834 (150)
853 (134)
873 (132)
800 (079)
902 (056)
744 (345)
882 (119)
870 (071)
34
102
32
PWS3
1109 (251)
900 (111)
971 (208)
903 (185)
925 (109)
959 (144)
966 (125)
859 (106)
207
23
66
PWS4
992 (136)
982 (145)
831(061)
793 (036)
1272 (355)
1135 (352)
847 (181)
774 (153)
199
480
498
PWS5
1030 (131)
1061 (137)
781 (172)
729 (116)
845 (139)
820 (093)
787 (068)
733 (083)
301
116
112
PWS6
1034 (193)
888 (185)
608 (130)
628 (170)
740 (141)
829 (191)
836 (122)
836 (049)
406
112
7
PWS7
1253 (240)
1106 (136)
910 (075)
957 (156)
920 (120)
872 (173)
1258 (347)
1072 (210)
296
-37
84
PWS8
1061 (166)
1215(177)
809 (127)
862 (169)
984 (194)
1210 (163)
1027 (200)
951 (164)
200
123
-88
PWS9
996 (222)
69
899 (261)
1017 (163)
898 (132)
1020 (125)
899 (084)
791 (284)
733 (173)
98
122
PWS10 1243 (216)
1017 (232)
997 (231)
973 (185)
831 (268)
1059 (189)
827 (070)
768 (061)
270
-142
99
PWS11
902 (326)
689 (078)
680 (085)
727 (154)
837 (127)
936 (104)
858 (156)
417
47
-41
1097 (275)
PWS12 1218 (349)
913 (419)
1054 (199)
951 (607)
827 (109)
894 (055)
938 (128)
846 (274)
268
-124
-31
PWS13 989 (137)
940 (141)
880 (103)
841 (058)
1029 (116)
1007 (155)
935 (093)
867 (089)
148
187
183
PWS14 1012 (237)
1028 (236)
902 (396)
873 (242)
895 (093)
905 (057)
703 (074)
771 (107)
139
22
28
MEAN
1006 (210)
884(156)
851 (164)
936 (153)
940(153)
916 (148)
854 (130)
236
085
086
1087 (211)
• (1-8) indicates Mean (SO) of trials 1 though 8, (9-16) indicates Mean (SO) of trials 9 through 16. b Oiff 1 =Block 1 (1-8) - Block 4 (9-16), Oiff 2 = Ret Block 1 (1-8)- Block 4 (9-16), Oiff 3 = Block 1 (1-8) - Ret Block 2 (9-16) Sequence Learning and Retention'2
Appendix G: Table 5 Mean (SD) Reaction Time (milliseconds) for Persons with Parkinson's Disease (PPD) Group
Session One Blocks 1 & 4
Diff~rence~ b
Retention Session Blocks 1 & 2
1 (1-8)'
1 [9-16]
4 [1-8)
PPDl
1387 (408)
1029 (271)
1137 (177)
4[9-16)
1 [1-8]
1 (9-16)
2 [1-8)
2 (9-16)
Diff 1
Diff2
Diff3
1236 (173)
1315 (307)
1289 (264)
1052 (133)
1053 (107)
152
79
262
34
PPD2
990 (191)
937 (214)
610 (132)
629 (172)
915 (113)
857 (081)
793 (090)
881 (094)
360
286
PPD3
1137 (148)
1088 (111)
865 (082)
850 (142)
1052 (207)
932 (266)
955 (123)
889 (198)
287
202
163
PPD4
1301 (493)
997 (250)
1111 (157)
1399 (287)
978 (238)
972 (161)
1066 (118)
1083 (074)
-98
-421
-106
PPD6
1014 (213)
941 (240)
907 (244)
873 (229)
729 (056)
715 (071)
633 (086)
705 (058)
141
-144
24
PPD7
1230 (296)
971 (252)
989 (172)
892 (143)
698 (240)
743 (128)
609 (216)
673 (224)
338
-194
26
PPD8
870 (085)
807(130)
748 (072)
765 (111)
754 (135)
842 (209)
954 (165)
1077 (235)
105
-11
-322
PPD9
1133 (300)
887 (244)
731 (164)
741 (171)
783 (246)
782 (244)
699 (086)
694 (106)
392
41
89
60
PPD10
902 (104)
932 (076)
1032 (170)
1031 (120)
927 (152)
887 (087)
863 (085)
867 (170)
-129
-104
PPDll
1063 (289)
967 (189)
904 (304)
780 (125)
693 (194)
592 (150)
607 (164)
834 (103)
282
-88
-141
PPD12
1026 (165)
939 (250)
839 (098)
759 (079)
912 (100)
1053 (140)
1259 (184)
1056 (219)
267
152
-144
PPD13
999 (291)
870 (244)
933 (231)
991 (179)
857 (238)
1032 (365)
781 (156)
703 (127)
8
-134
154
PPD14
646 (277)
719 (388)
566 (102)
592 (147)
MEAN
1053 (251)
929 (220)
866 (162)
888 (160)
884 (180)
866 (183)
856 (134)
876 (143)
-028
008
054
166
'(1-8) indicates Mean (SD) of trials 1 though 8, (9-16) indicates Mean (SD) of trials 9 through 16. Diff 1 = Block 1 (1-8) - Block 4 (9-16), Diff 2 = Ret Block 1 (1-8)- Block 4 (9-16), Diff 3 = Block 1 (1-8) - Ret 810ck 2 (9-16) b
"
Sequence Learning and Retention 3
Appendix G: Table 6
Mean (SO) Reaction Time (milliseconds) for Control Participants (C), Persons who Stutter (PWS) and Persons with Parkinson's Disease (PPO) for
Sequence-Specific Learning in Session One
Partici~ant
Controls
PWS
PPD
Random'
Sequence b
Diff
Random
Sequence
Diff
Random
Sequence
Diff
946 (226)
848 (081)
098
1198 (200)
1109(317)
090
1393 (262)
1348 (241)
046
2
840 (097)
790 (089)
051
871 (117)
827 (110)
044
724 (156)
720 (164)
004
3
1220 (252)
1074 (134)
146
838 (083)
897 (074)
-058
899 (258)
912 (108)
-013
4
1052 (200)
939 (170)
113
939 (142)
931 (079)
008
1133 (130)
1166 (236)
-034
5
1005 (168)
1018 (226)
-013
893 (094)
946(147)
-053
1321 (216)
1340 (196)
-019
6
1176 (246)
1067 (095)
109
731 (165)
988 (199)
-257
1100 (222)
1004 (117)
097
7
1276 (232)
946 (239)
330
941 (166)
891 (281)
049
1113 (256)
1140(218)
-027
8
1038 (218)
912 (091)
126
779 (218)
835 (218)
-056
764(140)
750 (103)
014
9
891 (200)
794(116)
098
984 (156)
1078 (199)
-094
967 (124)
962 (241)
005
10
817 (223)
876 (252)
-058
938 (140)
888 (133)
050
1157 (154)
1059 (112)
099
11
873 (107)
814 (101)
059
796 (114)
741 (091)
055
967 (146)
930 (247)
036
12
1125 (164)
1027 (179)
098
1219 (161)
1225 (186)
-005
792 (125)
827 (202)
-035
13
879 (098)
867 (102)
011
897 (227)
819 (121)
077
1161 (317)
1120 (249)
040
14
789 (175)
729 (058)
060
946 (299)
949 (225)
-002
818 (273)
748 (227)
070
MEAN
995 (186)
907 (138)
087
926 (163)
937 (170)
-011
1022 (198)
1002 (190)
020
• Mean (SD) of the first 16 random trial pairs from Block 5 (random block) of session one (values are rounded)
b Mean (SD) of the first 16 sequence trial pairs from Block 5 (random block) of session one (values are rounded)
c Difference =(random mean reaction time) - (sequence mean reaction time)
Sequence Learning Wld Retention '4
Appendix G: Table 7
Mean (SO) Reaction Time (milliseconds) Jar Control Participants (C), Persons who Stutter (PWS) and Persons with Parkinson's
Disease (PPD) Jar Sequence-Specific Learning on the Retention Test
Particigant
PWS
Controls
2
PPD
Random'
Sequence"
Diff
Random
Sequence
Diff
Random
Sequence
Diff
782 (135)
775 (193)
007
1111 (233)
1161 (211)
-050
988 (126)
1048 (160)
-050
807 (256)
690 (104)
118
966 (081)
909 (154)
056
850 (157)
870 (154)
-056
3
719 (114)
N/A
915 (122)
943 (149)
-028
988 (232)
1028(216)
-028
4
877 (147)
870 (105)
007
1002 (198)
914 (250)
088
1153 (157)
945(131)
088
5
1060 (132)
1044 (127)
016
898 (115)
748 (101)
150
1377 (221)
1201 (181)
150
6
796 (089)
812 (135)
-016
875 (068)
866 (055)
008
703 (133)
623 (097)
008
7
921 (175)
782 (124)
139
1227 (538)
1099 (278)
129
884 (317)
622 (147)
129
1194 (197)
N/A
984 (136)
1026 (176)
-042
1174 (178)
1073 (304)
-042
065
8
9
992 (206)
928 (206)
064
1015 (162)
950 (142)
065
982 (251)
848 (181)
10
832 (073)
874 (083)
-042
810 (100)
788 (150)
022
960 (196)
834 (183)
022
930 (121)
724 (139)
611 (175)
037
11
991(118)
1143 (225)
-152
893 (140)
037
12
1027 (212)
786 (156)
241
995 (306)
880 (202)
115
1161 (167)
1082 (151)
115
13
1021 (161)
991 (256)
030
945 (140)
860 (067)
085
834 (217)
803 (166)
085
14
1066 (185)
993 (162)
073
810 (131)
765 (124)
045
MEAN
915 (154)
914 (159)
040
963 (175)
914 (157)
048
983 (192)
891 (173)
040
'Mean (SD) of the first 16 random trial pairs from Block 5 (random block) of session one (values are rounded)
b Mean (SD) of the first 16 sequence trial pairs from Block 5 (random block) of session one (values are rounded)
'Difference = (random mean reaction time) - (sequence mean reaction time)
<,
Sequence Learning and Retention'5
Footnotes
1 Stuttering is not traditionally categorized as a "motor speech disorder", though it has been so categorized by some researchers in the past
(Kent, 2000). We acknowledge that there is continued controversy over the importance of many other variables besides motor control that
may influence stuttering onset, persistence and manifestations. However, because of the undeniable difficulty in controlling motoric aspects
of speech inherent to stuttering we felt justified in our classification for the purposes of this paper.
2 Despite investigators best efforts, one control participant dozed several times during the experiment and these approximately 70
(excluded) trials were not included in the above sum. As this participant was still one of the fastest control participants, the remaining trials of this drowsy participant, which were included in the analysis, had minimal, if any, negative effect on the group score. ) If the generate task stimuli were fPOI, fPI!, /?I, the correct answer would be PO, the foils would be PE and PA. PI was not a foil because PI never followed PI, not even in the pseudorandom condition and the participants understood this. 4 Exploratory post hoc analyses were performed. Three 2 (group) x 8 (Block) ANOVA (p = .05) were conducted to assess reaction times across Session One and the Retention Session for PWS vs. control participants, PPD vs. control participants and PWS vs. PPD. PWS' reaction times showed significant cubic polynomial contrasts relative to both control participants, F (1,23) = 4.5, P = .04, ~2= .16 and PPD, F (1,23) = 4.5, P = .05, ~2= .17. Although uncommon, we felt it was justified to conduct post hoc tests based on a near-significant comparison (p = .08), in the current study because this is an unprecedented explorative study with a relatively small sample size. Researchers have reasoned that a type 2 error is more dangerous than a type 1 error in this earliest stage of research (Portney & Watkins, 1993). Nevertheless the reader should be cautioned to the increased risk of type 1 error for these post hoc comparisons when interpreting these results. Sequence Learning and Retention (;
Documentation of Author Roles:
Documentation of Author Roles:
Sarah Smits-Bandstra
I. Research project: A. Conception, B. Organization, C. Execution
2. Statistical Analysis: A. Design, B. Execution, C. Review and Critique
3. Manuscript Preparation: A. Writing of the first draft, B. Review and Critique
Vincent Gracco
I. Research project: A. Organization, Execution
2. Statistical Analysis: A. Review and Critique
3. Manuscript Preparation: A. Review and Critique
Runrung Head: CHUNKING AND STUTTERING
Early stage chunking of finger tapping sequences
by persons who stutter and fluent speakers
Keywords: Chunking, Sequence, Practise, Stuttering, Finger Tapping
Chunking and Stuttering
2
Abstract
We explored the hypothesis that chunking differences underlie the slow finger-tap
sequencing perfonnance reported in the literature for persons who stutter (PWS) relative
to fluent speakers (PNS). Early stage chunking was defined as an immediate and
spontaneous tendency to organise a long sequence into pauses, for motor planning, and
chunks of fluent motor perfonnance. A previously published study in which 12 PWS and
12 matched PNS practised a 10-item finger tapping sequence 30 times was examined.
Both groups significantly decreased the duration of between-chunk intervals (BCls) and
within-chunk intervals (WCls) over practise. PNS had significantly shorter within-chunk
intervals relative to PWS, but minimal differences between groups were found for the
number of, or duration of, between-chunk intervals. Results imply that sequencing
differences found between PNS and PWS may be due to differences in automatizing
movements within chunks or retrieving chunks from memory rather than chunking per se.
Chunking and Stuttering
3
Introduction
Wingate (1969) stated the difficulty persons who stutter (PWS) experienced while
speaking was not related to producing sound units for speech; rather the difficulty was in
moving on from one element to the next. More recently Venkatagiri (2005) and Huinck,
van Lieshoul, Peters and Hulstijn (2004) have suggested that PWS were slower and less
accurate than fluent speakers (PNS) in their ability to pull from memory and concatenate
(link) syllable sound units into a fluent sequence. In particular, Huinck et al. found that
PWS differed from PNS in the fluent transition across syllable boundaries. The purpose
of this research note was to investigate whether stuttering was associated with differences
in sequence organization or 'chunking' of sequence elements with a focus on nonspeech
movement.
Neuroscience research has reliably shown that certain aspects of sequence
organization are effector non-specific (Cohen, Ivry, & Keele, 1990; Hikosaka et aI.,
1995; Hazeltine, Teague & Ivry, 2002). Basic motor programmes or synaptic tendencies
are thought to be primarily housed in the roughly approximate 'homonculous' in either
the articulator area (for speech movements) or the hand and finger area (for manual
movements). However, it is well-established that the organization of these
programmes/synaptic biases are supervised/developed according to input from
subcortical structures such as the basal ganglia and cerebellum. These structures send
timing/organization signals that are critical to most sequential movements regardless of
whether they are eventually routed to hand or mouth execution areas of the motor cortex
(Graybiel, 1998).
Chunking and Stuttering
4
Nonspeech tasks were examined in the current research note because tbey have
the additional advantage of avoiding confounds associated with top-down organization of
syllables into meaningful words or syntactic units.
Cbunking theory
After practise, the performance ofa sequence of movements becomes
increasingly integrated, with less hesitancies, false starts and errors (Lashley, 1951; Pew,
1966). The resultant well-practised movement is quick and fluent. It has been postulated
that one way sequence production becomes increasingly fluent is through the process of
'chunking'. For the purposes of this research note chunking was defined as a continuing
process where the composite elemental units of movement within a sequence become
temporally linked into stable chunks of more than one unit of movement (Sternberg et aI.,
1990; Klapp, 1995; Verwey & Dronkert, 1996; Immink & Wright, 2001 ; Klapp, 2003).
Early stage chunking was defined more specifically as the immediate self-generated
tendency to organise or 'chunk' a long sequence into shorter components.
Unchunked , a sequence remains broken down into small independent units, is
executed more slowly, and is more susceptible to errors as each transition is processed
(Fisk & Schneider, 1984. Chunking is thought to be responsible for increased efficiency
of processing, and increased fluency of sequencing performance over pratise (Newell &
Rosenbloom, 1981; Fisk & Schneider, 1984). Chunking is also associated with enhanced
memory retrieval for individual sequential movements so that performance is relatively
impervious to interference from ongoing cognitive distractors (Schneider, Dumais, &
Shiffrin, 1984; Logan, 1988).
Klapp (2003) compared motor chunks to short-term memory chunks. As in
memory, the motor chunk can be processed as a unit and fluent performance span is
Chunking and Stuttering
5
based on the number of chunks regardless of the internal complexity or informational
content of each chunk. Sakai, Kitaguchi, and Hakosaka (2003) stated that a chunk
representation can overcome the limitation of working memory capacity by forming a
hierarchical structure for many memory items. Enhanced memory retrieval reduces
competing demands, alleviating processing bottlenecks and structural interference.
Chunking processes
INT
Based on a series of robust reaction time studies, Klapp (1995) postulated the
existence of SEQ and INT, two independent chunking processes. 'INT' generates a
perceptual motor representations of single-chunk response elements and is thought to be
dictated by underlying structural changes (including axon growth, synapse formation,
etc,; Dariam-Smith & Gilber 1995; Karni, Meyer, Rey-Hipolito, Jezzard, Adams &
Turner, 1998) that take place over several days/sessions of practise in both humans
(Klapp, 1995; Verwey & Dronkert, 1996; Immink & Wright, 2001) and primates
(Trembley et aI., 2009). INT is also responsible for retrieving old chunks from memory.
INT processes are observable in simple reaction time studies with a single
movement/chunk and are thought to be able to prepare single chunks in advance of a 'go'
signal if the response is known. INT is hypothesized to be effector non-specific,
unconscious, and responsible for automatization of chunks, or retrieving automatized
chunks from memory. INT is thought to be observable in reduced pauses within chunks
of movement (WCI) and within-chunk movement durations in longer sequences (Klapp,
2003).
Chunking and Stuttering
6
SEQ
Klapp (1995) further identified across-chunk motor programming as an effector­
independent, attention-demanding process entitled 'SEQ'. SEQ is an immediately
apparent, hierarchical organization of multiple appropriate responses to a particular
stimulus (Jimenez, 2008). This hierarchical plan is based on previous experience
mapping stimuli to the individual components identifiable within a novel sequence
(Darian-Smith & Gilbert, 1995; Karni et aI., 1998). SEQ is apparent in both explicit
(conscious/declarative learning; Kock, Philip & Gade, 2006) and implicit sequence
learning (unconscious/procedurallearning; Kock & Hoffman, 2000) and is highly
consistent within an individual (Kermerly, Sakai & Rushworth, 2004).
Several studies have readily observed practise effects in 'SEQ' processing over a
single session as multiple movement elements/gestures were temporarily organised into
chunks (Klapp, 1995; Immink & Wright, 2001; Klapp, 2003). SEQ processing is
attention demanding, reflects plarming across chunks and can be inferred from pauses
between movement chunks (BCI).
Chunking in the literature
Chunking theory has been consistently supported by many past studies which
confirmed between chunk intervals (BCI) and significantly smaller within-chunk
intervals (WCI) of maze tracing movements; Van Meir & Hulstien, 1993), writing and
drawing movements (Hulstijn & Van Galen, 1983), finger tapping sequences (Taylor,
1943; Sternberg, Monsel, Knoll & Wright, 1978; lanacinski et a1., 1980; Garcia-Colera &
Semjen, 1987; Stadler, 1993) and speech (Klapp, 2003).
Based on results of the above studies researchers suggested that between-chunk
intervals (BCI; related to SEQ processing) shortened reaction time by reducing the
Chunking and Stuttering
7
demand for motor programming during the reaction time interval (Garcia-Colera &
Semjen, 1987; Klapp, 1995; Immink & Wright, 2001; Klapp, 2003). Researchers further
hypothesized sequences were executed more quickly when SEQ processing relegated
motor programming to BCI rather than co-occurring/interfering with execution of
ongoing movements (concurrent interference; Cohen et aI., 1990: Koch & Hoffman,
2000).
In contrast, within-chunk intervals (WCI; related to TNT processing) decreased
over successive practise trials as a complex response was organised into fewer but larger
chunks over practise. Klapp (2003) hypothesized that processing based on SEQ is
graduaUy replaced by INT processing to reduce reliance on attentive processing.
Several studies have investigated SEQ and TNT processing by manipulating the
manner in which participants subdivide sequences into chunks over practise. This was
done several ways including predictably arranging repeated items within the sequence
(Cohen, Ivry & Keele, 1990; Curran & Keele, 1993; Stadler, 1993), or arranging the
timing of sequence item presentation into groups (Verwey & Dronkert, 1996). Although
the existence of effectiveness of both TNT and SEQ chunking has been noted extensively
in the literature, the self-generated process of chunking over practise for individuals
(Verwey & Dronkert, 1996; Sakai et a., 2003) and patient populations (Doyon, 2008) has
only recently come under investigation.
Chunking and sequence learning in persons who stutter (PWS)
A review of reaction times studies reveals many instances where PWS show slow
simple reaction time for single movements/chunks relative to PNS, providing indirect
evidence implicating inefficient INT processing (Smits-Bandstra, 2010). PNS' reaction
Chunking and Stunering
8
times are most different from those of PWS when there is no/limited preparatory period,
further implicating inefficient TNT processing (Klapp, 2003).
However, PWS also show slow reaction times and durations for sequences of
movements, implicating inefficient SEQ processing. For example, many studies have
found poor overall speech (Watson & Alfonso, 1982; Peters, Hulstijn & Starkweather,
1989; Bishop, Williams, & Cooper, 1991 ; KJeinow & Smith, 2000; Smith & KJienow,
2000; Huink et aI., 2004) and nonspeech (Webster, 1986; Weinstein, Caruso, Severing
&VerHoeve, 1989; Williams & Bishop, 1992) sequencing performance and, more
specifically, poor sequence learning (Smits-Bandstra, De Nil & Rochon, 2006; Smits­
Bandstra, De Nil & Saint Cyr, 2006, Smits-Bandstra & De Nil, 2009) by PWS relative to
PNS.
Smits-Bandstra, De Nil and Saint Cyr (2006) compared the performance of PWS
and PNS while practising a 10-item finger tapping sequence and a 10-item nonsense­
syllable sequence over practise. PWS demonstrated significantly slower reaction times
for finger tapping and a trend for slower reaction times for syllable reading relative to
PNS. PWS also demonstrated a trend for slower syllable sequence durations. In a dual
task extension of this initial study, Smits-Bandstra, De Nil and Rochon (2006) reported
PWS had significantly slower sequence durations for a 10-item finger tapping sequence
in the single task condition relative to PNS. In another dual task extension of the initial
study, Smits-Bandstra and De Nil (2009) also reported that the syllable sequence reaction
times of PWS were significantly slower than those of PNS.
Based on these results researchers have proposed that PWS may differ from PNS
in the ability to organise speech and nonspeech movements into a fluent sequence
(Wingate, 1969; Huink, van Lieshout, Peters & Hulstijn, 2004; Venkatagiri, 2005; Smits­
Chunking and Stuttering
9
Bandstra & De Nil, 2007). It remains to be determined whether differences in SEQ
and/or !NT underlie the sequencing difficulties associated with stuttering. This
information is of no little consequence as insight into specific deficient processes may
point to neurological correlates of stuttering and suggest targeting of implicit vs. explicit
training in treatment.
Primary objective
Based on review of the literature, we assumed that PWS showed slower reaction
times and sequence durations relative to PNS due to difficulty with I) the timing or
organization of chunks (related to SEQ processing), or 2) automatization of, and/or
retrieval of chunks from memory (related to INT processing). We postulated that, if
PWS were less efficient at timing or organization of chunks, BCI would be less frequent
and oflonger duration in PWS relative to PNS (as is associated with ineffective SEQ
processing). We further postulated that, if PWS were less efficient at INT processing,
WCI would be oflonger duration in PWS relative to PNS (as is associated with
ineffective !NT processing).
It must be acknowledged that data described herein are derived from a previously
published study not originally set up to manipulate chunking behaviors. Neither was the
original study designed to identify SEQ vs. INT processing deficits. The main purpose of
this research note was limited to observation of pOtential trends in specific early stage
chunking processes by PWS and PNS during sequencing. This note served the purpose of
alerting researchers to these trends and encouraging further empirical study in this area.
Method
Details of participant recruitment, ethics approval, screening and methodology
can be found in a previously published study (Smits-Bandstra, De Nil & Rochon, 2006).
Chunking and Stuttering
10
This study reported significantly slower reaction time and total sequence duration over
practise for PWS relative to PNS. However, the timing and organization of individual
sequential elements and inter-element durations over practise revealed in the current
investigation have never been reported.
Participants
Twelve English-speaking males who stutter were matched for age and handedness
and compared with 12 English-speaking male PNS. PWS scored in the very mild to
moderate range (16.7 SO 6.8) on the Stuttering Severity Instrument-Ill (Riley, 1994).
Participants reported hours per week spent typing as 3.8 SD 1.8, and 3.9 SD 1.8 for PWS
and PNS respectively. Self-ratings of typing speed (2 slow, 6 average, 3 fast vs. 3 slow, 4
average and 4 fast) were reported for PWS and PNS respectively.
Procedure
Finger lapping sequence lask
Participants were seated comfortably in front of a computer monitor. Participants
were presented with an auditory 'ready' tone (a 1000 Hz tone for 500 ms) followed by a
fixed interstimulus interval of 1500 ms and catch trials to prevent anticipatory reaction
times. Participants were then presented visually with a I O-number sequence (3 1 3 2 4 I
lil.D printed in black.
Ten items were considered sufficient based on studies by Sakai
et al. (2003). They were instructed to use the four fingers of their right (dominant) hand
to press the four buttons arranged horizontally on a response box (Cedrus 610, Johnson,
1996) corresponding to the sequence of numbers on the screen. Participants were
instructed to look at the monitor and not the response box in order to prevent visual
feedback about their finger movements. Participants placed the index finger on the left
most button (button I), the middle finger on button 2, etc. They were instructed to 'type
Chunking and Stuttering
II
as fast you can without making mistakes' and to ' begin as soon as the sequence appears
on the screen ' . The finger sequences disappeared as soon as the participants completed
the sequence.
The number sequence was designed so that no number triplet was used more than
once, no number pair was used consecutively (e.g.l.l.lJ..), and no naturally occurring
triplets were used (e.g. I 2 3 or 4 3 2). Thirty repetitions of the trial were considered
adequate as the present study was limited to observing SEQ-type chunking in the earliest
stages (within one session). Several previous studies have observed significant practise
effects over 30 trials (Smits-Bandstra, De Nil & Rochon, 2006; Smits-Bandstra, De Nil &
Saint Cyr, 2006; Smits-Bandstra & De Nil, 2009).
Dependent variables and statistical analysis
Chunking of a novel, arbitrary sequence of known constituent movements (button
presses) was assessed over a single session of practise in the experiment examined.
Dependent variables included reaction time, between-chunk intervals (BCI), within­
chunk intervals (WCI), as well as a number of BCI (number of chunks) and consistency
of chunking patterns over practise. These dependent variables are well-established in the
chunking literature (Klapp, 1995; Immink & Wright, 2001; Sakai et aI., 2003).
Data regarding outliers and excluded error trials (minimal) are available in a
previously published study. Participant groups did not differ in outliers, total accuracy,
or accuracy patterns over practise (Smits-Bandstra, De Nil & Rochon, 2006).
Sequences were divided into taps separated by within-chunk intervals (WCIs).
WCI were measured as the time (in ms) from the fist button press to the second button
press, or the second button press to the third button press, and so on to the tenth button
Chunking and Stuttering
12
press. WCls were recorded online using the Superlab Pro 1.04 (Abboud & Sugar, 1997)
software programme.
Similar to previous studies (Sternberg et a!., 1978; Verwey & Oronkert, 1996)
unusually large WCls were identified as BCI. Past studies employing short sequences
(four items; Garcia-Colera & Semjen, 1987; Klapp, 1995; Immink & Wright, 2001;
Klapp, 2003) defined BCls as those WCls that were more than one or two SOs above the
mean WCI duration. However, SO was not an adequate threshold to identify BCI in the
current study because our sequence was long and often contained more than one BCI. If
there were two or more BCI the SO became very large and the BCI were not identified.
Similarly, if there were no BCI and very little variability around the mean, WCI that were
slightly larger than other WCI were mislabeled as BCI.
Similar to Sakai et a!. (2003), we defined WCI and BCI in distributional, rather
than absolute tenns so that the number and duration of scores did not mirror the group
differences in duration already published. Z-tests were conducted to test the probability
that a single WCI had a larger value than the rest of the WCI distribution for each
participant, for each I O-item sequence. Only those WCI identified by z-tests as having a
probability of95% or more of being larger than the rest of the WCI distribution were
categorized as BCI. This high z-test probability criterion argued against the assumption
that WCl and BCI were randomly varying values belonging to the same distribution of
scores. Because the number of BCI and duration of BCI and WCI fell on a relatively
nonnal distribution, it was deemed appropriate to define BCI based on z-tests of the raw
scores.
Chunking and Stuttering
13
Results
One subject from each group (one PWS and one PNS) performed extremely
poorly on the task and had reaction time, BCI and WCI values more than 2 SO from their
respective group means. The data from these subjects were excluded from analyses. One
PWS missed five sequences in a row (due to errors) and this resulted in an incomplete
data set for this PWS, therefore most comparisons involved complete data from II PNS
and 10 PWS.
Number of BCI
A 6 x 2 mixed ANOVA was used to assess the effect of the within-subjects factor
of Practise (trial 1-5, 6-10, etc.) and the between subjects factor Group (PWS vs. PNS) on
mean number of BCI per sequence. The 30 practise trials were subdivided into six equal
groups of five (trial 1-5, 6-10, 11-15, 16-20,21-25, 26-30) to minimize transient
fluctuations in performance. Practise effects were tested with polynomial contrasts to
find the most reasonable description of continuous data. Such contrasts allow comparison
of complex performance patterns across groups using linear, quadratic, cubic, quartic and
other higher order contrasts (Portney & Watkins, 1993).
Visual examination of a histogram suggested a good fit of the data to the normal
curve. Assumptions of covariance, sphericity and equality of error variance were met. A
significant quadratic Practise by Group interaction effect was found, F(5, 19) = 5.29, p =
1
.03, partial 1/ = .22. A trend for a quadratic Practise main effect was also found, F(5, 19)
1
= 4.15, P = .06, partial 1/ = .18. No Group main effect was found. Figure 1 displays the
unique pattern PNS and PWS show for the number of BCI/chunks per sequence over
practise.
<insert figure I about here>
Chunking and Stuttering
14
Reaction Time
A 6 x 2 mixed ANOVA assessed the effect of the within-subjects factor of
Practise (trial 1-5,6-10, etc.) and the between subjects factor Group (PWS vs. PNS) on
reaction time. Reaction time was considered a special kind of BCI and was analysed
separately from BCI duration. Visual examination of a histogram suggested a good fit of
the data to the normal curve. Assumptions of covariance, sphericity and equality of error
variance were mel. A significant quadratic Practise by Group interaction suggested that
the reaction times of PNS improved significantly more quickly over practise relative to
PWS, F(5 , \9) = 6.29, p = .02, partial 1/ 2 = 0.25 (see figure 2). A significant quadratic
Practise effect, and a trend for a Group main effect was also found, F(5, 19) = 9.08, p =
2
2
.0 \, partial 1/ = 0.32 and F( I, 19) = 3.49, p = .08 , partial 1/ = .16, respectively . Both
groups improved reaction times over practise, and PNS tended, on average to have
shorter reaction times than PWS.
<insert figure 2 about here>
Duration of BCl
A 6 x 2 mixed ANOVA was used to assess the effect of the within-subjects factor
of Practise (trial \-5,6-10, etc.) and the between subjects factor Group (PWS vs. PNS) on
BCI duration. Visual examination of a histogram suggested a good fit of the data to the
normal curve. Assumptions of covariance, sphericity and equality were mel. Levene' s
test indicated that in five of the six comparisons, the equality of error variance
assumption was mel. A significant linear Practise by Group interaction, F(5, 19) = 5.42,
2
P = .03, partial 1/ = .22, indicated that PNS demonstrated shorter BCI, on average,
relative to PNS for early practise trials only (see figure 3). A significant linear Practise
Chunking and Stuttering
IS
main effect indicate{j the BCI durations of both groups became shorter over practise, F(S ,
19) = 42 .91,p = .00, partial
r/ = .69.
No Group main effect was found .
<insert figure 3 about here>
Duration of WCI
A mixed 2 x 6 x 8 ANOVA assesse{j the effect of Group (PNS vs. PWS), Practise
(triall-S, 6-10, etc.) and WCI Position on sequence duration (WCIl = pause between I"
and 2 nd button press, WCI2 = pause between 2nd and 3'd button press, etc.). Eight WCI
positions were analyse{j including WCI2, WCB, WCI4 , WCIS, WCI6, WCl7, WCI9 and
WCI I O. WCI I was analyse{j separately as reaction time and WCI8 did not have
sufficient data to be analyse{j as it was use{j by the majority of participants as a BCI (only
two subjects did not use WCI8 as a BCI) to analyse this position. Levene's test indicate{j
unequal variances between the groups. A logarithmic transformation was used to
successfully equalize group variance. Mauchley's test of Sphericity was significant
therefore the Greenhouse Geisser Correction was applied to statistical results .
Significant main effects of Practise (linear), F(S, 19) = 160.78,p = .00, partial
r/
= .S4 and Group, F(I , 19) = 2.S,p = .01 , partial r/ = . \3, suggeste{j groups decreased
WCI durations over practise and PNS' WCls were shorter than PWS' WCls. As shown
in figure 4, a significant cubic Group x Practise interaction suggested that group
differences were largest early in practise with PWS 'catching up' in later practise trials,
F(S , 19) = 3.59, p = .02, partial
,.,2= . OS).
<insert figure 4 about here>
A significant Position main effect suggeste{j that WCI durations were dependent
to some degree on where they occurred in the sequence, F(8, 138) = 7.30 , p = .01, partial
,.,2 =
.OS. Post hoc analyses using Tukey's LSD indicated that WCI position 10 was
Chunking and Stuttering
16
significantly shorter than all other positions while WCI position 9 was significantly
shorter than WCI position 5 (see figure 5). No other contrasts were significant.
<insert figure 5 about here>
Discussion
In previous studies PWS have shown slower reaction times and sequence
durations relative to PNS (Smits-Bandstra, De Nil & Rochon, 2006; Smits-Bandstra, De
Nil & Saint-Cyr, 2006). We postulated these results were due to difficulty with temporal
organization of chunks (related to SEQ processing), or automatization of, and/or retrieval
of chunks from memory (related to INT processing). Based on previous research we
postulated that inefficient SEQ processing could be inferred from longer and less frequent
BCI. We further postulated that inefficient INT processing could be inferred from longer
WCI.
Empirical testing of INT and SEQ processing was beyond the scope of this
research note and any inferences made about SEQ and INT were based on comparison of
results found in previous chunking studies. Nevertheless potential trends were
observable in the post hoc analyses we conducted. Specifically, PWS demonstrated
significantly longer WCI but not BCI relative to PNS. As stated in the introduction, BCI
and WCI are not direct or sole indicators of SEQ or INT, respectively. With this caveat
in mind, results of the current study can be interpreted to suggest that the INT processing
of PWS appeared to be a promising avenue of future research.
BCI
One might expect fewer BCl (larger chunks) to evolve over pratise as this result is
reported in the classic literature (e.g. Taylor, 1943). However previous studies examined
chunking after extensive pratise while the current study reported immediate, early-stage
Chunking and Stut1ering
17
changes in a single session (Jimenez, 2008). As expected due to limited practise, we
found no significant changes in the number of BCI over pratise by either group.
As described in the introduction, behavioral studies have confirmed that when
BCI were incorporated into sequence training, sequence reaction time and overall
sequence durations were shorter compared to sequences with no trained structure (Cohen,
et aI., 1990; Koch & Hoffinan, 2000; Stadler, 1993; Verwey & Dronkert, 1996). Among
other things, SEQ processing is thought to be associated with initiating BCI into a
sequence, thereby decreasing reaction time and sequence duration.
As reported previously by Smits-Bandstra, De Nil, and Rochon (2006), PNS
demonstrated significantly faster reaction times and sequence durations (631.1 SD 200.5
and 3253.8 SD 946.9, respectively) relative to PWS (855.2 SD 298.8 and 3935.8 SD
1662.1, respectively). Results of the current analysis suggest that slow reaction times and
sequence durations were not due to lack of BCI or slower BCI durations. From this result
we can infer that this particular aspect of SEQ processing appeared relatively unaffected.
WCI
PWS had significantly longer WCI relative to PNS throughout the practise trials.
Previous studies suggest that, among other things, WCI reflect aspects of INT processing,
therefore we can infer from the current results that some aspect of PWS' INT processing
may be less efficient than is typical. Research has not conclusively shown that BCI are
solely programmed by SEQ, or WCI by INT. Therefore, at this time, we cannot entirely
rule out potential inefficiencies in SEQ when differences in WCI are found. In addition,
INT processes are thOUght to occur gradually, over the long term. The current
experiment was only one session, providing a limited sample of potential changes ofINT
with practice. The extent of practice should be considered in future chunking research.
Chunking and Stuttering
18
Longer WCI durations may indicate that PWS take longer/more pratise to
'automatize' or bind together sequence elements, potentially having to initiate each
segment individually. INT processing differences may also be due to inefficient retrieval
of chunks from memory. Neuroimaging studies of PWS suggest atypical white matter
underlying the primary motor area (REF), thought to be associated with movement chunk
memories/programmes (REF). Neuroimaging studies of PWS have also implicated the
cortico-striato-thalamo-cortical circuit (REF), though to be associated with
automatization of movement (REF). !NT processing inefficiency associated with
stuttering may be partially explained by these neurophysiological phenomena.
Both SEQ and !NT processes are active during reaction time for sequential
responses (Klapp, 2003). One explanation for the reaction time findings is that PWS'
!NT process had difficulty retrieving chunks from memory. An alternative explanation is
that PWS' SEQ process was able to instigate BCI appropriately , but only after a long
reaction time delay . Potential Iy there were inefficiencies associated with both !NT and
SEQ processes. It could also be argued that reaction time delays demonstrated by PWS
(in this study and in previous studies; BIoodstein, 1995) are carried over to delays in the
subsequent movement and/or movement transition after reaction time or a BCI. Because
these results are tentative and based on exploratory post hoc analyses further research in
this area is needed.
Chunking Strategies
One of the many possible explanations ofPWS' pattern of results is that PWS
tended to rely on the more consciously controlled SEQ processing strategy during
sequencing, rather than relegating chunks to more automatized !NT processing. There is
cODsiderable support in the literature for the proposal that PWS tend to rely on attention­
Chunking and Stuttering
19
demanding, consciously controlled movement strategies (Bosshardt, 1999; Bosshardt,
2002; van Lieshout et aI., 1996,2004).
The postulation that PWS continue to use executive motor control strategies
serves as a viable explanation for the Group by Practice interaction for reaction time
found in the current study. PNS were apparently able to quickly delegate motor control
to automatized self-evolving INT processes, resulting in drastic reaction time reductions
over the first few pratise trials.
One possible interpretation of the current results is that PWS' chunking is
primarily supervised by executive planning circuits with limited delegation to more
automatic subcortical circuits (Graybiel, 1998; Trembley et aI., 2009). As a result, SEQ
processing is relatively typical but INT processing is less automated (requiring increased
planning and monitoring of executive planning centres), and susceptible to interference
and competition for resources.
Stutterers often report difficulty initiating or moving past the first element of a
speech sequence (blocks, repetitions, prolongations). The speech of PWS is also slower
(Max & Yudman, 2003), increasingly dis fluent as length and complexity of the sequence
increase (Smith & Kleinow, 2000), and susceptible to concurrent tasks (Bosshardt, 2002).
This makes sense if PWS have to initiate each element individually and/or plan the entire
sequence at the beginning of the utterance. One could hypothesize that speech would be
more susceptible to interfering tasks if it were consciously controlled rather than
relegated to more automatized circuits.
Conclusion and Future Directions
A post hoc comparison of the sequencing performance of PWS and PNS revealed
a potential trend indicating INT processing (automatizing and/or retrieving automatized
Chunking and Stuttering
20
movement chunks from memory) rather than SEQ processing may be a promising avenue
of future research. It is recommended that future investigations of INT processing
examine simple reaction time and WCI of short sequences under conditions of extensive
practise over several sessions to allow for automatization of chllilks.
Differences in SEQ processing cannot be ruled out however. Further questions
arise as to whether SEQ processing might be affected for non-repetitive stimuli. Figure 3
displays a significant Practise by Group interaction revealing that BCI durations were
only different between PWS and PNS during the first few trials of the sequence. It is
well known that stuttering frequency decreases for repeated stimuli (adaptation effect). A
study designed to assess PWS' SEQ processing of novel sequences is merited,
particularly as most speech is composed of novel sequences of known words and
syllables.
We analysed the consistency of BCI number but did not assess consistency of
BCI position. This variable would be more appropriate to assess in short sequences with
typically one BCI. Given the current findings, variability (of both number and placement
ofBCl) in sequence organization is an interesting avenue for further investigation.
Although the current data are analysed within a group model it is important to
note that we found individual differences in the process of chunking over pratise. Our
results indicated that, similar to previous studies (Sakai et aI., 2003; Verwey & Dronkert,
1996), the number of BCl differed across individuals and individuals demonstrated
evidence of hypothesis testing using different BCl patterns across pratise trials.
Chunking and Stut1ering
Acknowledgments
We would like to acknowledge editorial assistance from Sophie Lafaille.
21
Chunking and Stuttering
22
Declaration of interest
We would like to acknowledge a grant from the Toronto Rehabilitation
Institute to the first author and a grant from the Natural Science and Engineering
Research Council of Canada to the second author. The post hoc analyses on the
study described herein was funded by a Canadian Institute of Health Research
grant to the first author. The authors report no conflicts of interest.
Chunking and Stuttering
23
References
References:
Abboud, H., & Sugar, D. (1997) . SuperLab Experimental Laboratory Software (Version 1.04)
(Computer software]. Phoenix, AZ: Cedrus Corporation.
Anderson, J. R. (1982). Acquisition of cognitive skill. Psychological Review, 89,369-406.
Bernstein, N. (1967) . The co-ordination and regulation ofmovements. Oxford: Pergamon Press.
Bishop, H. J., Williams, H. G., & Cooper, W. A. (1991). Age and task complexity variables in
motor perfonnance of stuttering and nonstuttering children. Journal ofFluency
Disorders, 16,207-217.
Bloodstein, O. (1995). A Handbook ofStuttering. (4 'h ed.; p. 190). Chicago, IL: Easter Seals
Society.
Bosshardt, H. G. (1999). Effects of concurrent mental calculation on stuttering, inhalation, and
speech timing. Journal ofFluency Disorders, 24,43-72.
Bosshardt, H. G. (2002). Effects of concurrent cognitive processing on the fluency of word
repetition: Comparison between persons who do and do not stutter. Journal ofFluency
Disorders, 27,93-114.
Clark, H., Robin, D., McCullagh, G., & Schmidt, R. (2001). Motor control in children and adults
during a non-speech oral task. Journal ofSpeech, Language and Hearing Research, 44,
1015-1025.
Cohen, N., Ivry, R. I., & Keele, S. (1990). Attention and structure in sequence learning. Journal
ofExperimental Psychology: Learning, Memory, and Cognition, 16, 17-30.
Curran, T., & Keele, S. W. (1993). Attentional and nonattentional fonns of sequence learning.
Journal ofExperimental Psychology: Learning. Memory, and Cognition, 19,189-202.
Darian-Smith, C., & Gilbert, C. D. (1995) . Topographic reorganization in the striate cortex of the
adult cat and monkey is cortically mediated. Journal ofNeuroscience, 15, 1631-1647.
Doyon, J. (2008). Motor sequence learning and movement disorders. Current Opinions in
Neurology, 21,478-483 .
Fisk, A. D., & Schneider, W. (1984) . Memory as a function of attention, level of processing, and
automatization. Journal ofExperimental Psychology: Learning, Memory, and Cognition,
10,181-197.
Fitts, P. M. (1964). Perceptual-motor skilileaming. In A. Melton (Ed .), Categories ofhuman
learning: Symposium on the psychology ofhuman learning (pp. 243-285). New York,
NY: Academic Press.
Garcia-Colera, A. , & Semjen, A. (1987). The organization of rapid movement sequences as a
function of sequence length. Acta Psychologica, 66,237-250.
Graybiel, A. M. (1998). The basal ganglia and chunking of action repertoires. Neurobiology of
Learning and Memory, 70, 119-136.
Hansen, S., Tremblay, L., & Elliot, D. (2005). Part and whole practise: Chunking and online
PNS in the acquisition of a serial motor task. Research Quarterly for Exercise and Sport,
76,60-66.
Hazeltine, E., Grafton, S. T, & Ivry, R. B. (1997). Attention and stimulus characteristics
detennine the locus of motor-sequence coding. Brain, 120, 123-140.
Hidding Bishop, J. , Williams, H., & Cooper, W. (1991). Age and task complexity variables in
motor perfonnance of stuttering and nonstuttering children. Journal ofFluency
Disorders, 16,207-217.
Chunking and Stuttering
24
Hikosaka, 0., Rand, M. K., Miyachi, S., & Miyashita, K. (1995). Learning of sequential
movements in the monkey: Processes of learning and retention of memory. Journal 0/
Neurophysiology, 74,1652-1661.
Huinck, W. J., van Lieshout, P. H. H. M., Peters, H. F. M., & Hulstijn W. (2004). Gestural
overlap in consonant clusters: Effects on the fluent speech of stuttering and non-stuttering
subjects. Journal 0/ Fluency Disorders, 29,3-25.
Hulstijn, W., & Van Galen, G. (1983). Programming in handwriting: Reaction time and
movement time as a function of sequence length, Acta Psychologica, 54,23-49.
Immink, M. A., & Wright, D. L. (2001). Motor programming during practise conditions high and
low in contextual interference. Journal 0/ Experimental Psychology: Human Perception
and Performance, 27,423-437.
Jagacinski, R. J., Shulman, H.G. , & Burke, M. W. (1980). Motor programming and alerting.
Journal 0/ Human Movem enJ Studies, 6, 151-164.
Jimenez, L. (2008) Taking patterns for chunks: is there any evidence of chunk learning in
continuous serial reaction-time tasks? Psychological Research, 72, 387-396).
Johnston, D. (1996). Cool Edit 96 (1996 version) [Computer Software]. Phoenix, AZ:
Syntrillium Software Corporation.
Karni, A., Meyer, G., Rey-Hipolito, C., Jezzard, P., Adams, M. M., & Turner, R. (1998). The
acquisition of skilled motor performance: Fast and slow experience-driven changes in
primary motor cortex. Proceedings o/the National Academy o/Science, USA, 95,861­
868.
Klapp, S. T. (1995). Motor response programming during simple and choice reaction time: The
role of practise. Journal 0/ Experimental Psychology: Human Perception and
Performance, 21, 1015-1027.
Klapp, S. T. (2003). Reaction time analysis of two types of motor preparation for speech
articulation: Action as a sequence of chunks. Journal 0/ Molor Behavior, 35, 135-150.
Kelso, S. (1995). Dynamic palterns: The self-organization 0/ brain and behavior. Boston, MA;
MIT Press.
Kennerley, S. W., Sakai, K. & Rushworth, M.F.S. (2004). Organization of action sequences and
the role of the pre-SMA. Journal of Neurophysiology, 91, 978-993.
Kleinow, J., & Smith, A. (2000). Influences of length and syntactic complexity on the speech
motor stability of the fluent speech of adults who stutter. Journal 0/ Speech, Language,
and Hearing Research, 43, 548-559.
Koch, I. & Hoffmann, J. (2000) Patterns, chunks, and hierarchies in serial reaction-time tasks.
Psychological Research, 63,22-35
Koch, I., Philipp, A. M., & Gade, M. (2006). Chunking in task sequences modulates task
inhibition. Psychological Science, 17,346-350.
Lashley, K. S. (1951). The problem of serial order in behavior. In L. A. Jeffress (Ed.), Cerebral
mechanisms in behavior, (pp. 112-136). New York: Wiley.
Logan, G. D. (1988). Toward an instance theory of automatization. Psychological Review, 95,
492-527.
MacKay, D. (1982). The problems of flexibility, fluency and speed-accuracy trade-off in skilled
behavior. Psychological Review, 89,483-506.
Max, L., & Yudman, E. M. (2003). Accuracy and variability of isochronous rhythmic timing
across motor systems in stuttering versus nonstuttering individuals. Journal o/Speech,
Language, and Hearing Research, 46, 146-163.
Chunking and Stuttering
25
Newell, A., & Rosenbloom, P. (1981). Mechanisms of skill acquisition and the law of practise.
In J. Anden;on (Ed.), Cognitive skills and their acquisition (pp. I-56), Hillsdale, N.J.:
Erlbaum.
Peters, H. F. M. , Hulstijn, W., & Starkweather, C. W. (1989). Acoustic and physiological
reaction times of stutterers and nonstutterers . Journal oJSpeech and Hearing Research.
32, 668-680.
Pew, R. W. (1966). Acquisition of hierarchical PNS over the temporal organization of a skill.
Journal ojExperimental Psychology, 71, 764-771.
Portney, L. G., & Watkins, M. P. (1993). Foundations ojclinical research: Applications to
practise. Stamford, CN: Appleton & Lange.
Posner, M. I., & Keele, S. W. (1968). On the genesis of abstract ideas. Journal ojExperimental
Psychology, 77, 353-363.
Riley, G. (1994). A stuttering severity instrument Jor children and adults (3rd ed .). Austin, TX:
Pro-Ed.
Sakai, K., Kitaguchi, K., & Hikosaka, O. (2003). Chunking during buman vi suomotor sequence
learning. Experimental Brain Research, 152,229-242
Sclunidt, R. A., & Wrisberg, C. A. (2004). Motor learning and perJormance (3,d ed.).
Champaign, IL: Human Kinetics Publishers.
Schneider, W., Dumais, S. T., & Shiffrin, R. M. (1984) . Automatic and PNS processing and
attention. In R. Parasuraman, & D. Davies (Eds .), Varieties ojattention (pp. 1-27).
Orlando: FL: Academic Press Inc.
Smith, A., & Kleinow, J. (2000). Influences of length and syntactic complexity on the speech
motor stability of the fluent speech of adults who stutter. Journal oJSpeech, Language,
and Hearing Research. 4,548-559.
Smits-Bandstra S., & De Nil, L. F. (2009). Speech skill learning of persons who stutter and
fluent speakers under single and dual task conditions. Clinical Linguistics & Phonetics.
23, 38-57
Smits-Bandstra, S., & De Nil, L. F. (2007). Sequence skill learning in persons who stutter:
implications for cortico-striato-thalamo-cortical dysfunction. Journal ojFluency
Disorders, 32, 251-278.
Smits-Bandstra, S., De Nil, L. F., & Rochon, E. (2006). The transition to increased automaticity
during finger sequence learning in adult males who stutter. Journal ojFluency Disorders,
31,22-42.
Smits-Bandstra, S., De Nil, L. F., & Saint-Cyr, J. A. (2006). Speech and nonspeech sequence
skill learning in adults who stutter. Journal ojFluency Disorders, 31, 116-131.
Stadler, M. (I993). Implicit serial learning: Questions inspired by Hebb (1961). Memory &
Cognition, 21, 819-827.
Sternberg, S. , Knoll, R.L. , & Turock, D.L. (1990) Hierarchical control in the execution of action
sequences: Tests of two invariance principles. In M. Jeannerod (Ed.) Attention and
PerJormance Xlll Hillsdale, N.J.: Erlbaum, Pp. 3-55
Sternberg, S., Monsel, S., Knoll, R. R., & Wright, C. E. (1978). The latency and duration of
rapid movement sequences: Comparisons of speech and typewriting. In G.E. Stelmach
(Ed.), InJormation Processing in motor PNS and learning (pp. 117-152). New York:
Academic Press.
Taylor, D.W. (1943). Learning telegraphic code. Psychological Bulletin, 40,461-487.
Chtmking and Stuttering
26
Trembley, P-L. Bedard, M-A ., Levesque, M., Chebli, M ., Parent, M., Courtemanche, R. , &
Blanchet, P-J. (2009). Motor sequence learning in primate: Role of the D2 receptor in
movement chtmking during consolidation. Behavioral Brain Research, 198,231-239.
van Lieshout, P. H. H. M. , Hulstijn, W., & Peters, H. F. M. (1996) . From planning to articulation
in speech production: What differentiates a person who stutters from a person who does
not stutter? Journal oJSpeech and Hearing Research, 39,546-564.
van Lieshout, P. H. H. M., Hulstijn, W., & Peters, H. F. M. (2004). Searching for the weak link
in the speech production chain of people who stutter: A motor skill approach . In B.
Maassen, R. D. Kent, H. F. M . Peters, P. H. H . M. van Lieshout, & W. Hulstijn (Eds.).
Speech motor control in normal and disordered speech (pp. 313-356). Oxford, UK:
Oxford University Press.
Van Mier, H., & Hulstijn, W. (1993). The effects of motor complexity and practise on initiation
time in writing and drawing. Acta Psychologica, 84, 231-251.
Venkatagiri, H. S. (2004). Slower and incomplete retrieval of speech motor plans is the proximal
source of stuttering: Stutters occur when syllable motor plans stored in memory are
concatenated to produce the utterance motor plan. Medical Hypotheses, 62,401-405.
Verwey, W. B. (1994). Evidence for the development of concurrent processing in a sequential
key pressing task. Acta Psychologica, 85, 245-262.
Verwey, W., & Dronkert, Y. (1996). Practising a structured continuous key-pressing task : Motor
chunking or rhythm consolidation? Journal ojMotor Behavior, 28,71-79.
Watson, B. C., & Alfonso, P. J. (1982). A comparison ofLRT and VOT values between
stutterers and nonstutterers. Journal ojFluency Disorders, 7, 219-241.
Webster, W. G. (1986). Neuropsychological models of stuttering-II: Interhemispheric
interference. Neuropsychologia, 24,737-741.
Weinstein, J., Caruso, A. J., Severing, K., & VerHoeve, 1. (1989). Abnonnalities of oculomotor
PNS in stutterers [Abstract]. Investigative Ophthalmology and Visual Science, 30,78.
Williams, H. G., & Bishop, J. H. (1992). Speed and consistency of manual movements of
stutterers, articulation-disordered children, and children with nonnal speech. Journal oj
Fluency Disorders, 17, 191-203.
Wingate, M. E. (1967). Stuttering and word length. Journal oJSpeech and Hearing Research,
10,146-152.
Chunking and Stuttering
27
Figure Caption
figure 1. Number of between-chunk intervals (BCI) per sequence averaged across practice trials for participant groups (error bars represent I SO of intersubject variability). figure 2. Average finger sequence reaction time for participant groups over practice (error
bars represent 1 SO of intersubject variability).
figure 3. Average between-chunk interval duration (BCI) for participant groups over
practice (error bars represent 1 SO of intersubject variability).
figure 4. Average within-chunk intervals (WCI) for participant groups over practice
(error bars represent I SO ofintersubject variability) .
figure 5. Average duration of within-chunk intervals in all positions across the IO-item
sequence (error bars represent 1 SO of intersubject variability).
Chunking and Stuttering
3 ~--------------------------------------------------------
"5
j
§
z1.5 ~----------------------------------------------------------
1 ~----------------------------------------------------------Practfce TrIals Praetice Trials Practlce Trials PractJc:e TrIals Practice Trials Practlce TrIals
1·5
6-10
11·15.
16-20
21·25
26-·30
figure 1.
28
Chunking and Stuttering
1300 1200 1100 i•
1000 ,~
:!
:=
E 900
.E
~
1=
aDo
iI:
700 c
600 "­
~
~ I­
~"-
' ~ .....
-.~
....
­
500 400 Practice Trials Practice Trials Practice Trials Practice Trials PracUce Trials Practice Trials 1-5
6-10
11·15
16-20
21·8
26-30 figure 2.
29
Chunking and Stuttering
1100
-A 1100
I
~1000
..1
e
.,
100
!
1­
~
"
i~
~ Io.... ..........
'&100
j
'"
aoo
-. ....... ~
""
....
~
PrKtIIf TrIaIa PI'Ktl<lI Tna PrKUaI T.... Pra*l Trial. Pfldlal
1-5
figure 3 . &-10
1145
18-2Q
Tr1."
21-25
PI"Ktlal Trt4afs
21-30 30
Chunking and Stuttering
soo
T
!
.11.
'"
'"
.~
~~
~
-.......,
L.
~ .......
I
200 PractIce TrAI5 PractJot TrIa.b Practk:e Trial. practfQ Trtlll Practice Trials
1-6
6-10
11·15
15-20
J1"25
figure 4.
p~
'frqls 26030 31 Chunking and Stuttering
~o
r------------------------------------------------------­
4~ +----------------------------------------r~--------------
50
o
WCl2
figure 5.
Wet5
WCI6
WCI7
WC\4
wCla
WCI9
Position of wlthln-ehunk Intervals within the 10-item sequence WCl3
WCl10 32