Title: Bio 345545 Week 12 day 2 Skill Acquisition in Sport: Evolving Theories
1Bio 345/545 Week 12 day 2Skill Acquisition in
Sport Evolving Theories
Masters . com
2Good News on Exam Scores
3Last Class Fundamentals of Skill Learning
- Learning is a process
- Relatively permanent change in capability
- Over-learning occurs
- Massed practice learning still occurs
- KR versus KP
- Transfer tests expose learning
- Conditions of Practice
- Variable practice beneficial
4Today How do we typically learn, or teach others
new skills?
- Practice is organized to steer tasks to match
conceptual model
- Practice schedules, organizations, constructs
debatable
Question What underlying principles are at work
in skill acquisition that might tell us something
about how practice should be structured or
organized?
5Todays Topic
- Present Two Major Approaches to conceptualizing
skill acquisition - Information Processing
- Dynamical Systems Theory (DST)
- Digress to explain emerging DST, and its
application to skill acquisition - Address ramifications to teaching or learning new
skills
6Information Processing Approach
- Has dominated theories of skill acquisition for
decades, emphasizing prescriptions for movement
The mind is conceptualized as a capacity-limited
device, into which raw sensing information is
channeled and acted upon in accordance with a
standard motor program.
Basically, the mind is a computer
7Challenges to IPT
- The mind is not a computer
- The physical framework of the brain is not
compatible with the computer metaphor
2. Great flexibility of performers in
accommodating variable task demands (even novel
tasks)
3. To adapt to variable task demands, an
incredibly fast parallel processor would be
required.
8More Problems with IPT
- The loans on rationality argument
- Imagine the computing power required if such vast
amounts of information regarding contractile
states and limb orientations are specified, a
priori and ongoingly, by an existing prescription
for movement.
Consider the factors affecting a muscles
contribution to torque about a joint
9Factors Affecting a Muscles Torque Contribution
- Muscle Force
- Activation state
- Velocity (/-)
- Length
- History
- Fatigue
- Metabolic Capability
- Electromechanical Delay
- Moment Arm
- Changing during movement
- 3. Biarticular muscle effects
10A Complex System
Over 600 Muscles Over 100 Joints Massive number
of degrees of freedom is a curse for
computational accounts of skill acquisition and
performance
Add the rapidly changing contexts of activities
11Enter Ecological Approach
3 key elements
- Actions best understood as highly specialized
relationships between the organism and the
environment
- Emphasis is on circular relationships between
perceptual systems and movement systems
- Perceptual information in the form of energy
flows can constrain the coordinated movement that
emerges during goal directed tasks
12Ecological Approach
In Essence
Perceptual information can be used to guide the
evolving dynamics of the neuromuscular system
toward the most adaptable pattern necessary for
specified task execution. Handford et al.
(1997)
DYNAMIC SYSTEMS THEORY
13A Fascinating Diversion Intro to Non-Linear
Behaviors
- Scientists, since Newton, have been preoccupied
with studying order in nature - Ordered systems are predicable, deterministic,
and conceptually reversible - But non-linear behavior is pervasive in nature,
and the study of NLD has arisen in chemistry,
biology, physics, and math
14Non-Linear Behaviors
- Self-organization (learning??) is a property of a
complex system living in far from equilibrium
conditions.
- Example 1 Bernard Instability
- 1 cm vegetable oil in pan, heat
- at low temps, molecules evenly distributed
- At critical temp, hexagonal matrix forms
- Increase temp more turbulent boiling
3 patterns of behavior are caused by changes in
an outside control parameter
15Non-Linear Behaviors
- An important feature of Bernards Instability is
steadfastness, or tenacity (resistance to change)
within the three phases
Non-linear dynamacists call this property
stability
Key Idea Your swing is grooved and consistent.
Your game is stable
16Non-Linear Behaviors
- Example 2 A flowing stream
- Low velocity laminar flow
- Moderate velocity fixed vortices
- Higher velocity separating vortices
- Highest velocity turbulent flow
17Insert Thought on Skill Acquisition
To achieve a new, qualitatively different
behavioral pattern (assumed to possess some
dynamic stability), requires a loss of stability
of a previous behavior pattern.
NEW PATTERN FORMATION
And ..Breaking of Bad Habits
18NLD and the Concept of Attractors
Think of a highly integrated, dynamical system in
nature. How could we describe its behavior at
any one time?
Break down the system into parts and plot
behavior as it changes over time yielding a
multi-dimensional state space
Within this state space, a trajectory emerges,
dictated by key environmental and internal
constraints, in the form of a flow
19Huh?
20NLD and the Concept of Attractors
Trajectories often thought of (or modeled) as
motion along the contours of a landscape
Attractors are hollows which the system could
settle into. Attractors may represent system
states in which component parts are brought into
coordinated synergy with each other.
21Marble trajectories over surfaces
22Back to Learning Dynamics Patterning in the
Coordination of Movement
Principle of self-organization!
Kelsos Classic Bimanual Experiments (1981, 1984)
23Kelsos Study
- 2 independent variables patterns of movement
and freq. of oscillation - 2 patterns in-phase (0 deg, homologous) and
anti-phase (180 deg, alt-homologous) - stepped frequencies 1 Hz x 15, 0.25 Hz ? 3 Hz
- Calculated relative frequency phase ratios
diff in phase values
24Kelsos Results
- Mean relative phase
- Avg STD of relative phase
- Zero deg pattern stable across all freqs., STD
low - 180 deg pattern
- Achievable at low freq
- At crit. Freq, phase transition to zero deg
pattern
25Kelsos Results
ALSO
3. Before switching to in-phase pattern,
anti-phase pattern exhibited instability 4. When
frequencies are stepped down, lost anti-phase
pattern is not regained (hysterisis).
26Scaling Effect Critical
- Allows visualization changes in coordination are
affected by previous states of the system - Skilled tasks exhibit scalability
27The HKB Model
- Herman Haken, German Physicist Joined with Kelso
and Bunz to develop HKB model ( to explain
Kelsos experimental findings)
Focus on presence of only 2 stable patterns
transition from one pattern to
another presence of only 1
pattern following transition
hysterisis effect
28The HKB Model
- Modeled the attractor landscape using a
potential energy (V) function
RESULTS
29HKB Model Conclusions
When system is in an unstable pattern and control
parameter is changed, system response will be
attracted to a more stable pattern.
Note perturbation experiments other
movement patterns, non-homologous pairs
concept of attraction observed throughout
30What About Learning?
Focus coordinated patterns with different
stabilities, loss of stability due to changes in
the potential landscape induce via control
parameters, and phase transitions
Learning viewed as pattern formation new
patterns on the background of already existing
patterns.
Initial State? Learner possesses intrinsic
dynamics, based on attractor layout
31A learning variant on Kelso study
Zanone and Kelso, 1992
- Subjects learned a 90 deg relative phase pattern
by watching a pair of lights (FB). - 5 days of practice
- Cycling freq. 1.75 Hz
- Before/after each practice session, relative
frequencies scaled, no FB - Retention tests, 7 days / 2 months post, no FB
32What do you think happened?
- Did the practiced pattern become more stable?
- Did other patterns become less stable?
- Were new phase transitions introduced?
33Results of Zanone and Kelso
- Subjects learned to produce 90 deg pattern
- Mean relative frequency and ? STD
- 180 deg pattern became less stable
- Zero deg pattern stability unchanged
Some subjects learned to produce the 90 deg
pattern at the expense of the 180 deg pattern.
34Back to the Ecological Approach to Skill
Acquisition
Skill acquisition viewed (in this conceptual
arena) as a process of hollowing out a
functionally appropriate attractor (or set of
attractors) which the system can settle into
during activity
35Implications for Teaching Skills
- Variability
- Response variability often viewed as unwanted
noise in performance. - The goal of successful skill acq. is believed to
be the reduction of system noise - BUT, working conditions can never be exactly
reproduced because of underlying instabilities in
the control system. Movement patterns are always
different!
36Implications for Teaching Skills
- Variability
- Message for teachers of skills
Value Variability!
Variability enhances the search for regions of
stability or attraction
To test the robustness of possible task
solutions, performer should manipulate control
variables, disrupting the stability of existing
attractor states.
37Implications for Teaching Skills
Experience regarding what is incorrect or
unstable is as important to discovery as what is
correct or stable
38Implications for Teaching Skills
- Constraints on Skill Acquisition
- A constraint which changes scale, leading to a
qualitative change in dynamics, is termed a
control parameter - Challenge for teachers of skills Identify
control parameters acting to constrain the system
(internal and external)
Manipulation of task constraints exposes subjects
to the broader landscape, allowing more effective
probing of attractor stabilities
39Newell (1980)
- Practice is considered to be the continuous
search for solutions to a movement problem in a
perceptual-motor workspace.
Bernstein (1967)
Repetition without repetition
40Critical Role of Teacher
- Support the search process by manipulating
constraints so that exploration activity occurs
over an optimal area of the perceptual-motor
workspace.
Largely Individual Specific
41Skill Acquired
- Optimal values of control variables assigned
- Passive, inertial and mechanical properties of
system are fully exploited - Coordinated structures become extremely stable
and additional degrees of freedom are released - The system is scalable, readily adaptable to
modified environmental constraints
42Now go play on the range!