Transcript Document

Using Felder’s
Index of Learning Styles
in the Classroom
Kay C Dee, Glen A. Livesay
Department of Biomedical Engineering
Tulane University, New Orleans, LA 70118 USA
Who, What, Why?
We are not here to tell you
“how you should teach.”
The Dark Side of Teaching Well
You’ll
NEVER get
tenure!!
Of course, this varies with institution and priorities.
“You prep for classes your way, Harris,
I’ll prep for classes my way.”
Overview
Broad Questions:
What are some of the different ways that
students take in and process information?
Which learning styles are favored by:
• many students?
• the teaching style of many professors?
What can we do to reach a full spectrum of
learning styles?
Outline
I
What is Felder’s Index of Learning Styles
(ILS)? Where did it come from?
II What has the ILS told us about learning
styles so far?
III Let’s be fair - are there concerns or
critiques associated with the ILS?
IV How can we use learning style information
to make informed teaching style choices?
V
Does using ILS information in the classroom
actually make a difference?
Learning Styles
There are several definitions of “learning style.”
Generally, these definitions include aspects of:
• perception, acquisition, processing, and
retention of information
• both cognitive and affective behaviors
• individuality
• maximal learning when instruction capitalizes
on an individual’s learning style preferences
- the matching hypothesis
Felder’s Index of Learning Styles
• Relatively short questionnaire
• Specifically formulated with engineering
students in mind
• Does not require professional scoring and
interpretation
• Collected data/publications available
[1]
• Dimensions well-suited for discussions of
teaching as well as learning
Index of Learning Styles: Overview
Visual
Active
Verbal
Reflective
Sensor
Sequential
Intuitor
Global
ILS Domains
Visual
Verbal
• Pictures
• Spoken words
• Diagrams
• Written words
• Flow charts
• Formulas
• Plots
“Show me the systems
you’re talking about.”
“Explain what’s going
on inside the systems.”
ILS Domains
Active
Reflective
• Tends to process
information while
doing something
active
• Tends to process
information
introspectively
• Likes group work
• May never get around
to starting tasks
• May start tasks
prematurely
“Let’s just try it out.”
• Likes independent work
“Let’s make sure we’ve
thought this through.”
ILS Domains
Sensor
• Focuses on sensory
input - what is seen,
heard, touched, etc.
Intuitor
• Focuses on ideas,
possibilities, theories
• Prefers more abstract
• Prefers concrete
information: theory and
information: facts and
models
data
“How does this class
relate to the real world?”
“All we did were plugand-chug assignments.”
ILS Domains
Sequential
• Can function with
partial understanding
Global
• Needs to see the big
picture
• Makes steady progress • May start slow and then
make conceptual leaps
• Good at detailed
analysis
“I need to focus on one
part of the project and get
it done - then I can move
onward.”
• Good at creative
synthesis
“I need to see how this
all fits together before I
can start the project.”
What We’re NOT Saying
We don’t mean to “put people in boxes.”
What We’re NOT Saying
We don’t mean to “put people in boxes.”
Everyone learns both
actively and
reflectively, both
visually and verbally,
etc.
Most people, however, have some
preferences (mild, moderate, or strong).
Origins of ILS Domains
Sensor - Intuitor
• Carl Jung’s theory of psychological types:
sensing and intuition modes of perception
• Myers-Briggs Type Indicator:
sensors and intuitors as problem solvers
• Kolb’s experiential learning model:
concrete experience and abstract
conceptualization
Origins of ILS Domains
Active - Reflective
• Myers-Briggs Type Indicator:
extrovert and introvert
• Kolb’s experiential learning model:
active experimentation and reflective
observation
Kolb’s Cycle
Concrete
Experience
Feeling
Internalizing
experience
Making
something new
Reflective
Observation
Active
Experimentation
Watching
Doing
Doing it
Abstract
Conceptualization
Thinking
Developing
concepts
Kolb’s Learning Model
Concrete
Experience
Diverger
Accommodator
Reflective
Observation
Active
Experimentation
Converger
Assimilator
Abstract
Conceptualization
Kolb’s Learning Model
Concrete
Experience
Accommodator
Diverger
Active
Experimentation
Converger
Reflective
Observation
Assimilator
Abstract
Conceptualization
Kolb’s and Felder’s Models
Concrete
Experience
Accommodator
Diverger
Active
Converger
Reflective
Assimilator
Abstract
Conceptualization
Kolb’s and Felder’s Models
Sensor
Accommodator
Active
Diverger
Reflective
Converger
Assimilator
Intuitor
Index of Learning Styles: Overview
Visual
Active
Verbal
Reflective
Sensor
Sequential
Intuitor
Global
Faculty Learning Styles
100
n=568
Percent of Population
80
(national)
83
90
75
73
70
60
50
38
40
36
27
25
30
17
20
10
0
Visual
[2]
Active
Sensor
Preferred Learning Style
Global
n=12
(Tulane
BMEN)
Learning Styles - Tulane
100
Percent of Population
90
n=12
88
(BMEN
faculty)
83
75
80
70
62
n=255
(ENGR
students)
60
60
52
50
40
25
30
17
20
10
0
Visual
Active
Sensor
Preferred Learning Style
[3]
Global
Learning Styles of Other Engineers
Percent of Population
100
90
80
70
88
86
80
72
69 67
69
62
60
60 59 57
53
50
52
40
33
30
28 28
20
10
0
Visual
Active
Sensor
Global
Preferred Learning Style
Tulane, Engr (n=255) [3]
U Michigan, Chem Engr (n=143) [5]
U Western Ontario, Engr (n=858) [4]
Ryerson Univ, Elec Engr (n=87) [6]
Learning Styles and Gender
Percent of Population
100
90
80
70
Males, Engr
89
(n = 692)
69
Females, Engr
72
(n = 135)
59
60
58 61
University of
Western Ontario
50
40
35
30
25
20
10
0
Visual
Active
Sensor
Preferred Learning Style
Global
[7]
Learning Styles and Gender
Percent of Population
100
90
80
70
91
89
84
78
72
69
59
60
56
50
58
67
61
61
50
40
48
35
25
30
20
10
0
Visual
Active
Sensor
Global
Preferred Learning Style
University of Western Ontario
Males, Engr
(n = 692)
Females, Engr
(n = 135, 16.3%)
Tulane University
Males, Engr
(n = 129)
Females, Engr
(n = 63, 32.8%)
Index of Learning Styles: Critiques
Concerns which have been noted regarding the
use of the Index of Learning Styles:
• Doesn’t predict academic performance.
• The matching hypothesis - just a
hypothesis - is difficult to prove.
• Lacks statistical validation.
• Bunch’a hooey.
[11]
[8]
[9,10]
[8]
Predicting Academic Performance
We found little or no correlation between SAT score and
cumulative GPA at the end of the sophomore year.
4.0
3.5
2
GPA
3.0
R = 0.16
2.5
2.0
1.5
1.0
Tulane sophomores in Statics,
all disciplines, n=98 [3]
0.5
0.0
800
900
1000 1100 1200 1300 1400 1500 1600
SAT score
Intuitors Outperformed Sensors on SAT
Percent of Population
Sensor/Intuitor vs. SAT Score
0.60
0.50
0.40
0.30
SAT Score
0.20
Below 1100 (4)
1100-1199 (5)
1200-1299 (24)
0.10
1300-1399 (40)
0.00
Strong
1400-1499 (21)
Mod
Sensor
Mild
Mild
1500+ (4)
Mod
Intuitor
Strong
Tulane sophomores, Statics group, n=98 [3]
ILS = Academic Performance? No.
Some concerns regarding the ILS appear to
arise from a misapplication of the inventory:
• It was not developed to enable predictions of
academic performance.
• It was not developed as a selection tool to determine
‘who should be an engineer’.
Activities or tests which engage only one
learning style may not illustrate the true
potential or abilities of a group of students.
Testing the Matching Hypothesis
B = f (P, E)
Behavior-person-environment paradigm
leads to the idea of optimizing the
instructional environment for optimal learning.
Testing the matching hypothesis is difficult there are many learning style schemes to test,
not all easily comparable to each other.
Meta-analyses [9,10] have claimed that a
majority of published studies support the
matching hypothesis.
Statistical Validation
Reliability
(Precision)
a coefficient
Validity
(Accuracy)
item total
correlation (ITC)
Statistical Analysis
SPSS was used to:
• Calculate a reliability coefficients for each
learning style domain.
- a larger a value implies a more internally
consistent construct.
• Perform item and factor analyses to determine
which items were most strongly correlated
with each other and how many factors were
present within each domain.
- removing poorly correlated items
increases a.
Reliability (a) of ILS Domains
0.8
0.564
0.718
0.596
0.544
achievement
Alpha
0.7
0.6
attitude[12]
0.5
0.4
0.3
0.2
0.1
0
ActiveReflective
SensorIntuitor
VisualVerbal
n=248
n=246
n=242
ILS Domain
SequentialGlobal
n=244
Reliability (a) of Core ILS Domains
0.8
0.582
0.744
0.679
0.622
achievement
Alpha
0.7
0.6
attitude[12]
0.5
0.4
0.3
0.2
0.1
0
ActiveReflective
SensorIntuitor
VisualVerbal
n=249
n=247
n=248
ILS Domain
SequentialGlobal
n=248
Measures of Reliability
a is commonly used for estimating reliability
(mean of split halves).
Challenges:
- low number of questions (11 per domain)
- mutually exclusive (dichotomous) questions
- no ‘right’ answer to questions
Test-retest reliability is what a is estimating:
to what degree will people obtain the same
ILS scores if they take the test again?
Challenges:
- requires multiple administrations
- if too long between, people may change
- if too short between, people may remember test
Test-Retests Are Correlated Over Time
Correlation Coefficient
Between Test - Retest
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0


Four
(n=24)§
Seven
(n=40)
Twelve
(n=26)§
Sixteen
(n=24)§
Months Between Test - Retest
Active-Reflective
Sensor-Intuitor
Visual-Verbal
Sequential-Global
 NOT significant (p>0.05)
§ Population includes
same students
Number of Questions*
Specific Answers Correlated Over Time
% Students Repeating Original Answers
on a Given Question in Retest
*Out of 44 questions.
Test-Retest Data
(16 month interval, n=24)
More ‘Repeatable’ Questions
Greater than 90% of students answered
test-retest identically on these questions
37) I am more likely to be considered:
a) outgoing
b) reserved
41) The idea of doing homework in groups, with one grade for
the entire group:
a) appeals to me
b) does not appeal to me
43) I tend to pictures places I have been:
a) easily and fairly
b) with difficulty and without
accurately
much detail
Test-Retest Data
(16 month interval, n=24)
Less ‘Repeatable’ Questions
50% or less of students answered
test-retest identically on these questions
16) When I’m analyzing a story or a novel:
a) I think of the incidents b) I know the themes and must
and put them together
go back to find the incidents
17) When I start a homework problem, I am more likely to:
a) start working on the b) try to fully understand the
solution immediately
problem first
36) When I am learning a new subject, I prefer to:
a) stay focused on the
b) try to make connections between
subject, learning as
that subject and related subjects
much about it as I can
44) When solving problems in a group, I would be more likely to:
a) think of steps in
b) think of possible consequences or
the solution process
applications in a range of areas
Test-Retest Data
(16 month interval, n=24)
Validation Study Summary
The ILS satisfies general guidelines for a
reliability across all domains.
• a between 0.54 and 0.72 with all questions.
 a increased in all domains with “core” questions,
especially visual/verbal, sequential/global.
Test-retest scores in all domains were
significantly correlated over various intervals.
• correlation was highest at shortest interval, and
generally reduced with longer intervals.
Recommendations
We believe Felder’s ILS to be a useful,
appropriate, statistically-acceptable tool for
characterizing learning preferences and
discussing teaching methods.
There is (as always) some room for
improvement. We encourage others to test
new questions, work on statistical validation especially when the ILS is administered to
large numbers of students at one time - and
share their findings.
Additional Comments on Validity
The nature of the ILS - to force choices for a set
of individual questions - necessarily spreads out
responses.
- Increases in variance are directly related to
lower values for a.
Guidelines for statistical validity developed for
tests of achievement (e.g., a minimum of 0.7)
should not be blindly applied to the ILS.
“An instrument is valid if it measures
what it is intended to measure”.
Dimensions of Teaching and Learning
Preferred
Learning Style
Visual
Verbal
[13]
Corresponding
Teaching Style
Input
Visual
Verbal
Presentation
Processing
Active
Passive
Student
Participation
Sensor
Intuitor
Perception
Concrete
Abstract
Content
Sequential
Global
Understanding
Sequential
Global
Perspective
Active
Reflective
The “Traditional” Lecture Format
The traditional engineering
lecture format (teaching style)
tends to be (almost exclusively):
VERBAL
PASSIVE
SEQUENTIAL
INTUITIVE
Learning Styles and “Traditional” Lectures
The traditional lecture format does match
some students’ preferred learning styles,
however, the majority of students tend to
prefer:
VISUAL,
ACTIVE, and
SENSING approaches
In fact, the teaching style utilized in the
traditional lecture does not necessarily match
the preferred learning styles of professors!
Teach to a Student’s Style, or Against?
[14]
The matching hypothesis: teaching to a
student’s learning style provides the best
opportunity for learning.
- a student functioning in their preferred modes is focused
on learning and not on overcoming a barrier.
However, should we teach to the strengths of
the student, or work to help them develop in
their areas of weakness (less preferred
modes)?
- students will need to be able to function in different
modalities at different times, e.g. both actively and
reflectively, both visually and verbally, etc.
Teach to Many Preferred Styles
[14]
With the diversity of learning styles in the
classroom, do we teach to a single, preferred
learning style? If so, which one?
- teaching to a single, preferred learning style (or using a
single style to teach) will benefit those FEW students
who prefer that chosen style.
The best solution is likely to utilize a variety
of instructional styles and modes of delivery
in a course.
- enable ALL of the students to function in their preferred
modes some of the time, and also encourage
development in less-preferred modes.
Teaching Styles - Reaching Styles
Good news:
Traditional lecturing does address
several categories of learning styles.
VERBAL, (REFLECTIVE),
SEQUENTIAL, and INTUITOR
Better news:
Engaging multiple learning styles does NOT
require complete restructuring of a course, or
eliminating traditional lectures.
Teaching Styles - Reaching Styles
Still better news:
Teaching methods that address styles shortchanged by traditional methods (e.g. VISUAL,
ACTIVE, GLOBAL, and SENSOR) often
accommodate multiple styles.
For example: [15]
• Motivate theoretical material with prior presentation
of phenomena that the theory will help explain, and
problems the theory will be used to solve (SENSOR,
GLOBAL).
• Balance concrete information (SENSOR) with
conceptual information (INTUITOR) in all courses.
Teaching Styles - Reaching Styles
• Complement oral and written explanation of concepts
(VERBAL) with extensive use of sketches, plots, etc.
and physical demonstrations where possible (VISUAL).
• Illustrate abstract concepts with at least some
numerical examples (SENSOR), in addition to
traditional algebraic examples (INTUITOR).
• Use physical analogies and demonstrations to improve
students’ grasp of magnitudes of calculated quantities
(SENSOR, GLOBAL).
• Demonstrate the logical flow of individual course
topics (SEQUENTIAL), and also highlight connections to
other material in the course and other courses, in other
disciplines, and in everyday experience (GLOBAL).
Teaching Styles - Reaching Styles
• Provide time in class for students to think about
material being presented (REFLECTIVE) and for active
participation (ACTIVE).
- pause during lecture to allow time for thinking and
formulation of questions (reflective).
- assign 1-minute papers, where students write the most
important point of the lecture and the most pressing
unanswered question (active and reflective).
- assign brief, group problem-solving exercises where
students work with neighbors (active and reflective).
• Encourage or mandate cooperation on homework, or
through team projects (ACTIVE).
Teaching Styles - Reaching Styles
Teach the Cycle!
Concrete
Experience
Reflective
Observation
Active
Experimentation
Abstract
Conceptualization
(Kolb’s Cycle, that is.)
Meta-Active Learning
What types of reasons might
professors give for not using these
ideas (for example: active learning
exercises) in their courses?
2 minutes, Go!
Fears - Active Learning
TIME-CONSUMING
LOSE CONTROL OF THE CLASS
UNPREDICTABILITY
TOO MUCH EFFORT
“How will I cover the syllabus?”
“What do I want to cover?”
“What do I want the students to be able to do?”
Active Learning: Benefits
• Students cannot be ‘passive vessels’ they must be engaged with the material
• Clearly informs instructor what students
understand and what they don’t
• Shifts focus from professor to material
(“sage on the stage” to “guide on the
side”)
• Increases and personalizes studentprofessor interactions
Active Learning: Potential Drawbacks
• Students cannot be ‘passive vessels’ they must be engaged with the material
• Clearly informs instructor what students
understand and what they don’t
• Shifts focus from professor to material
The Whole Enchilada - DOES IT WORK?
A longitudinal study of chemical engineering
students has shown that courses designed to
accommodate a spectrum of learning styles:
• increased students’ confidence in their
academic preparation [17]
• raised overall academic performance [16]
(even in subsequent courses taught
“traditionally” by other instructors [17])
• increased student retention [16]
• increased graduation rate
[17]
Data from Tulane BMEN
We have modified several junior-level courses
to address the sensing and active learning
preferences of our students.
New lab components were made possible
through a National Science Foundation “Course,
Curriculum and Laboratory Improvement”
grant.
A team of biomedical and psychology faculty
designed an assessment questionnaire to be
used as part of the evaluation plan.
TUBA Model
Tulane University Biomedical Assessment
(TUBA) model
Five constructs:
1. My perception of what happened in the course
2. Laboratory or “laboratory-like” experiences
3. How my skills and abilities were enhanced in
the course
4. My assessment of the course
5. The instructor
Administered to 134 students in Spring 2001,
113 students in Fall 2001, and 77 students in
Spring 2002.
TUBA Model
53 “fill-in-the-blank” questions using the scale
1. strongly disagree
2. disagree
3. neutral or undecided
4. agree
5. strongly agree
Example:
1. This course included a number of “handson” projects or exercises. _____
Statistical validation for TUBA model was conducted using
data from the three administrations. [18]
Assessing the Impact
Three courses were assessed in Spring 2001
and Spring 2002.
Longitudinal results were assessed by
performing independent t-tests on each item.
The number of items which demonstrated
significant (p < 0.05) improvement were
summed and reported.
No items showed ‘negative improvement’ from
Spring 2001 to Spring 2002.
Results
BMEN 340 - Spring 2001 and Spring 2002
Items
Improved
Total
Items
Perception of course
5
8
Laboratory-like
experiences
5
6
Skill and ability
enhancement
5
16
Assessment of course
3
11
Instructor
5
12
Construct
What Happened?
BMEN 340 incorporated no hands-on activities
in 2001 but added three laboratories in 2002,
teaching students a new skill set (cell culture
experiments).
Students in 2002 expressed higher ratings of
their:
• teamwork skills
• interest in conducting research or working in
the area
• confidence in their abilities
• instructor’s knowledge of the material
QUIZ
I
What is Felder’s Index of Learning Styles
(ILS)? Where did it come from?
II What has the ILS told us about learning
styles so far?
III Let’s be fair - are there concerns or
critiques associated with the ILS?
IV How can we use learning style information
to make informed teaching style choices?
V
Does using ILS information in the classroom
actually make a difference?
Wrap-up and Summary
I
What is Felder’s Index of Learning Styles
(ILS)? Where did it come from?
*
Visual
Verbal
Sensor
Intuitor
*
Active
Reflective
Sequential
* Similar to stages of Kolb’s cycle.
Global
Wrap-up and Summary
II What has the ILS told us about learning
styles so far?
Students tend, in general to prefer visual,
active, and sensing learning styles.
Not all populations of students (or faculty)
are the same.
Wrap-up and Summary
III Let’s be fair - are there concerns or
critiques associated with the ILS?
Yes.
BUT: We believe Felder’s ILS to be a useful,
appropriate, statistically-acceptable tool for
characterizing learning preferences and
discussing teaching methods. Nothing more,
nothing less.
Wrap-up and Summary
IV How can we use learning style information
to make informed teaching style choices?
There are many ways.
Start small. Try an approach more than once
before giving up. Tell students what you are
doing and why.
If you try only two things:
• show pictures or models (visual)
• provide short times to think and short times
to interact (reflective / active)
Wrap-up and Summary
V
Does using ILS information in the classroom
actually make a difference?
Yes.
Thank you.
Acknowledgements
We thank:
• The Rose-Hulman Institute of Technology, for
the opportunity to present this material.
• The students who participated in our studies,
for their time and good will.
• Rich Felder, for mentoring and inspiration.
• The National Science Foundation for support
provided by awards DUE-0088333,
BES-9983931, and BES-0093969.
Percent Information Retained
Active Learning and Info Retention
100
70
Lecture with
active breaks
20
Lecture
0
10
50
Time Into Lecture When Information Was Presented
(minutes)
R. Brent, R. Felder, J. Stice, National Effective Teaching Institute, 1998.