ABRAPRAXIA 2020 Conference | October 3, 2020 Principles of Motor Learning in CAS

Principles of Motor Learning

and
Childhood Apraxia of Speech

Edwin Maas, PhD

emaas@temple.edu

Speech, Language, and Brain Lab

Department of Communications Sciences and Disorders
Temple University

ABRAPRAXIA Conference
October 3, 2020

Disclosures
Financial:
• I receive salary from Temple University
• I have received grant funds from Apraxia Kids and NIH

Non-financial:
• I serve as a member on the Apraxia Kids Professional
Advisory Council (PAC), and as PAC Liaison member of
the Apraxia Kids Executive Board

Edwin Maas 1
ABRAPRAXIA 2020 Conference | October 3, 2020 Principles of Motor Learning in CAS

• Preliminaries: Setting the stage

– Motor control
– Motor learning
• Principles of motor learning
– Practice conditions
– Feedback conditions

Goal: to help you better understand some of the things

you might see in your child’s treatment sessions

Premises for interest in PML

• CAS is a speech motor planning problem
• CAS requires intensive treatment (e.g., ASHA, 2007;
Campbell, 1999; Hall, 2000; Shriberg et al., 1997)

• Resources are limited (money, time, SLPs)

How can we improve speech

motor learning in CAS?

Edwin Maas 2
ABRAPRAXIA 2020 Conference | October 3, 2020 Principles of Motor Learning in CAS

Approach
• Some practice conditions reliably enhance learning of
nonspeech motor skills (“principles of motor learning”)

• Question: Do those conditions also enhance learning

of speech skills in people with motor speech disorders
– and in particular children with CAS?

(see Maas et al., 2008, 2014; Bislick et al., 2012, for reviews)

Motor Learning
“a set of processes associated with practice
or experience leading to relatively permanent
changes in the capability for movement”
(Schmidt & Lee, 2011, p. 327)
– retention vs. temporary improvements
– transfer vs. movement-specific improvements

Edwin Maas 3
ABRAPRAXIA 2020 Conference | October 3, 2020 Principles of Motor Learning in CAS

The most important message of today

Just because your child does better during

practice in a given condition does not mean
your child learns the most in that condition

Practice performance vs. retention (learning)

Scheiner, Sadagopan, & Sherwood (2014)

Edwin Maas 4
ABRAPRAXIA 2020 Conference | October 3, 2020 Principles of Motor Learning in CAS

To see whether treatment works, you

need to look at retention and transfer

more so than gains made in the session

Motor Learning
NOT simply remembering a past movement and
re-using it

Motor learning is figuring out how the motor

system works to achieve a goal in a particular
situation
• Starting position
• Intended goal
• Motor commands
• Sensory effects

Edwin Maas 5
ABRAPRAXIA 2020 Conference | October 3, 2020 Principles of Motor Learning in CAS

GOAL
Where am I?
Where is the goal?
Motor commands:
• Which leg?
• Which muscles?
• How hard?
Was the shot on goal?

eat at

Goal: [ t ]

Different
starting positions

[ it ] “eat” [ æt ] “at”

Edwin Maas 6
ABRAPRAXIA 2020 Conference | October 3, 2020 Principles of Motor Learning in CAS

Goal: [ t ]
[ æt ] “at”
ash at

What did that feel like?

Was that the right sound?

What motor command did I

use?

Where did I start?

“Principles of Motor Learning” (PML)

(see Maas et al., 2008; Schmidt & Lee, 2011; for reviews)

• Predictable learning benefits of certain conditions

compared to other conditions
e.g., more practice is better than less practice
• PML are incorporated in several current
treatment approaches for CAS, including
• DTTC and other integral stimulation-based approaches (e.g., Edeal
& Gildersleeve-Neumann, 2011; Maas et al., 2012, 2019; Strand et al., 2006)
• ReST (e.g., Ballard et al., 2010; Murray et al., 2015)
• Motor Speech Treatment Protocol (e.g., Namasivayam et al., 2015)
• Ultrasound biofeedback treatment (e.g., Preston et al., 2013, 2016, 2017)

Edwin Maas 7
ABRAPRAXIA 2020 Conference | October 3, 2020 Principles of Motor Learning in CAS

“Principles of Motor Learning” (PML)

(see Maas et al., 2008; Schmidt & Lee, 2011; for reviews)

Conditions of Practice
• Practice amount
• Practice distribution
• Practice variability
• Practice schedule
• …

Conditions of Feedback
• Feedback type
• Feedback frequency
• Feedback timing
• …

Practice Conditions

• Practice amount more > less practice

• Practice distribution distributed > massed practice
• Practice variability variable > constant practice
• Practice schedule random > blocked practice
• Attentional focus external > internal focus
• … …

Edwin Maas 8
ABRAPRAXIA 2020 Conference | October 3, 2020 Principles of Motor Learning in CAS

Amount
(‘cumulative intervention intensity’; Warren et al., 2007)

• High vs. Low number of trials, sessions

• More trials  greater retention (High > Low)

(e.g., Giuffrida et al., 2002; Keetch et al., 2005; Shea & Kohl, 1991; Shea & Park, 2005)
– More trials = more opportunities to figure out relations among
pieces of information (starting position, motor command,
movement outcome/goal, sensory effects)

Low amount
GOAL

Edwin Maas 9
ABRAPRAXIA 2020 Conference | October 3, 2020 Principles of Motor Learning in CAS

High amount

GOAL
Amount: Speech
(‘cumulative intervention intensity’; Warren et al., 2007)

3 studies in CAS (Edeal & Gildersleeve-Neumann,

2011; Maas et al., 2019; Namasivayam et al., 2015)

More > Less

Edwin Maas 10
ABRAPRAXIA 2020 Conference | October 3, 2020 Principles of Motor Learning in CAS

Amount: Example
Maas et al. (2019)
• 6 children with childhood apraxia of speech (CAS)
• Alternating treatments SCED study (data from first phase)
• ASSIST (Apraxia of Speech Systematic Integral Stimulation
Treatment) on 15 utterances, randomly assigned to
• High amount condition (5 items)
• Low amount condition (10 items)

High (5 targets)

Low (10 targets)

Amount: Example
Maas et al. (2019)
• 6 children with childhood apraxia of speech (CAS)
• Alternating treatments SCED study (data from first phase)
• ASSIST (Apraxia of Speech Systematic Integral Stimulation
Treatment) on 15 utterances, randomly assigned to
• High amount condition (5 items)
• Low amount condition (10 items)
• ~180 minutes / condition (8 sessions of ~22.5 minutes)
 ~18 minutes for each utterance in Low amount and ~36
minutes for each utterance in High amount
• Outcome measure = accuracy of whole utterance

Edwin Maas 11
ABRAPRAXIA 2020 Conference | October 3, 2020 Principles of Motor Learning in CAS

Amount: Example
d Effect Size by condition
High
4.00
Low
3.00 Control
Standardized Effect Size (d)

2.00

1.00

0.00

-1.00

-2.00

-3.00
001 002 003 004 005 008
Child ID Maas et al. (2019)

Amount:
Implementation
Ways to maximize practice amount:

• Fewer targets, more trials

• Fast reinforcers, shorter and fewer breaks
• More sessions

Edwin Maas 12
ABRAPRAXIA 2020 Conference | October 3, 2020 Principles of Motor Learning in CAS

Distribution
(‘dose frequency’; Warren et al., 2007)

• Distributed vs. Massed practice: same amount divided

over more or less time (“intensity”)
– “spacing” vs. “cramming”

• Distributed practice  greater retention (D > M)

(e.g., Baddeley & Longman, 1978; Brashers-Krug et al., 1996; Lee & Genovese, 1988,
1989; Shadmehr & Brashers-Krug, 1997; Shea et al., 2000; cf. also Bahrick & Hall, 2005)

• However, evidence from the neuroplasticity literature

suggests M > D (e.g., Kleim et al., 2002; see Kleim & Jones, 2008, for review)
– Evidence mostly from animal studies
– and usually conflates amount and distribution

Distribution: Speech
(‘dose frequency’; Warren et al., 2007)

3 studies in CAS (Maas et al., 2019; Namasivayam et al.,

2015; Thomas et al., 2014)

Massed > Distributed

Edwin Maas 13
ABRAPRAXIA 2020 Conference | October 3, 2020 Principles of Motor Learning in CAS

Distribution: Example
Maas et al. (2019)
• 6 children with childhood apraxia of speech (CAS)
• Alternating treatments SCED study
• ASSIST (Apraxia of Speech Systematic Integral Stimulation
Treatment) on 20 utterances, randomly assigned to
• Distributed condition (10 items)
• Massed condition (10 items – 5 in phase 1, other 5 in phase 2)
Distributed Distributed
(10 targets) (10 targets)
Massed 1 Massed 2
(5 targets) (5 targets)

Distribution: Example
Maas et al. (2019)
• 6 children with childhood apraxia of speech (CAS)
• Alternating treatments SCED study
• ASSIST (Apraxia of Speech Systematic Integral Stimulation
Treatment) on 20 utterances, randomly assigned to
• Distributed condition (10 items)
• Massed condition (10 items – 5 in phase 1, other 5 in phase 2)
• ~360 minutes / condition (16 sessions of ~22.5 minutes)
 ~36 minutes for each utterance (achieved over 8 sessions for
Massed items, and over 16 sessions for Distributed items)
• Outcome measure = accuracy of whole utterance

Edwin Maas 14
ABRAPRAXIA 2020 Conference | October 3, 2020 Principles of Motor Learning in CAS

Distribution: Example
d Effect Size by condition
Massed
4.00
Distributed
3.50
Control
3.00
Standardized Effect Size (d)

2.50

2.00

1.50

1.00

0.50

0.00
001 002 003 004 005 008
-0.50
Child ID Maas et al. (2019)

Distribution:
Implementation
Ways to increase intensity (massed practice):

• More sessions/week over shorter period

• Smaller sets of targets
• Divide practice targets into subsets and focus
on 1 subset at a time

Edwin Maas 15
ABRAPRAXIA 2020 Conference | October 3, 2020 Principles of Motor Learning in CAS

Variability
• Constant vs. Variable practice: one vs. multiple
versions of a movement

• Constant > Variable for performance during practice,

but Variable > Constant for learning (e.g., Lai & Shea, 1998; Lai
et al., 2000; Lee et al., 1985; Shea et al., 2001; Wulf & Schmidt, 1997)
– Variability allows for greater exploration of relationships
between starting position, movement outcome/goal, motor
commands, and sensory effects

• Even better: Constant followed by Variable (Lai et al., 2000)

– First establish the basic movement pattern, then learn how
to adapt it to different situations

High variability
GOAL

Edwin Maas 16
ABRAPRAXIA 2020 Conference | October 3, 2020 Principles of Motor Learning in CAS

Low variability

GOAL
Variability: Speech
1 study in CAS (Preston et al., 2017)

Variable > Constant

Edwin Maas 17
ABRAPRAXIA 2020 Conference | October 3, 2020 Principles of Motor Learning in CAS

Variability: Example
Preston, Leece, McNamara, & Maas (2017)
• 6 children with CAS
• SCED alternating treatments design
• Ultrasound treatment 2x / week for 7 weeks, with one
target sound randomly assigned to each condition:
– Variable practice: Practice utterance with different prosodic
goals (neutral, question, command, slow, fast, and loud)
– Constant practice: Practice utterance with single prosodic
goal (neutral)
• Outcome measure = accuracy of target sounds in words
on generalization probes

Variability: Example
d Effect Size by condition
Variable
47 Constant
Standardized Effect Size (d)

37

27

17

-3
Danica Ethan Finn Greg Hannah Isaac
Child Preston et al. (2017)

Edwin Maas 18
ABRAPRAXIA 2020 Conference | October 3, 2020 Principles of Motor Learning in CAS

Variability: Example
d Effect Size by condition Variable
5 Constant

4
Standardized Effect Size (d)

-1
Finn Greg Hannah Isaac
Child Preston et al. (2017)

Variability:
Implementation
Ways to increase variability

• Vary phonetic context (embed target in different sound

or phrase environments)
• Vary prosody (e.g., question/statement intonation)
• Vary other suprasegmentals (e.g., loud/soft, fast/slow)
• Vary elicitation tasks (e.g., repetition, picture naming)

Edwin Maas 19
ABRAPRAXIA 2020 Conference | October 3, 2020 Principles of Motor Learning in CAS

Feedback Conditions

• Feedback type KR > KP

• Feedback frequency less > more feedback
• Feedback timing delayed > immediate feedback
• Feedback control self-controlled > clinician-controlled
• … …

Feedback Frequency
• Feedback: How often? (Low vs. High Frequency)
• Low (e.g., 60%) vs. high FB frequency (e.g., 100%)
• High > Low for performance during practice, but
Low > High for learning (e.g., Bruechert et al., 2003; Lai & Shea,
1998; Swinnen et al., 1990; Winstein & Schmidt, 1990; Winstein et al., 1994; but see
Wulf et al., 1994; Wulf & Schmidt, 1989)
• Low forces greater reliance on internal response monitoring
mechanisms (e.g., Bruechert et al., 2003; Sullivan et al., 2008)
• But children may actually need High, perhaps due to
limited cognitive resources (Sullivan et al., 2008)

Edwin Maas 20
ABRAPRAXIA 2020 Conference | October 3, 2020 Principles of Motor Learning in CAS

Feedback Frequency: Speech

1 study in CAS (Maas et al., 2012)

Mostly Low > High

Feedback Frequency: Example

Maas, Butalla, & Farinella (2012)
• 4 children with CAS

• Alternating treatments SCED with multiple baselines

• Integral stimulation Tx, 3-5 words/phrases per condition

• HiFF: feedback on every trial (100%)
• LoFF: feedback on 60% of trials (randomly distributed across
session)

• Outcome measure = accuracy of whole utterance

Edwin Maas 21
ABRAPRAXIA 2020 Conference | October 3, 2020 Principles of Motor Learning in CAS

Feedback Frequency: Example

5.00
HiFF
4.50
LoFF
Combined effect size d (both phases)

4.00 control

3.50

3.00
LoFF LoFF
2.50

2.00

1.50
HiFF? neither
1.00

0.50

0.00
001 002 005 012

Feedback Frequency:
Implementation
Ways to implement Low FB frequency:

• For random (non-tapered) Low FB frequency: Create

10 flashcards per target, mark 6 for FB. Shuffle cards.

• For fading (tapered) Low FB frequency: Create trial

tracking sheet with some slots marked for FB

Edwin Maas 22
ABRAPRAXIA 2020 Conference | October 3, 2020 Principles of Motor Learning in CAS

PML in CAS: Summary

• PML do appear to be relevant for CAS
• relatively consistent effects on learning
• but so far, few PML have been studied in CAS
• even fewer replications of effects
• Also:
• almost all studies involve small sample sizes
• Effects may interact with other conditions, child
factors (age, severity), …
So be skeptical of firm recommendations!

PML in CAS: Summary

• TENTATIVE recommendations (thus far):
‒ Amount: More > Less
‒ Distribution: Massed > Distributed**
‒ Variability: Variable > Constant
‒ Feedback Frequency: Low > High Frequency?

• Definite recommendation:
• test for retention and transfer
to measure learning!

Edwin Maas 23
ABRAPRAXIA 2020 Conference | October 3, 2020 Principles of Motor Learning in CAS

Questions?
Comments?

emaas@temple.edu

Edwin Maas 24