Slide 1 - Christine M. Pense by yaofenji

VIEWS: 4 PAGES: 96

									On Models
and Modeling
    You may recall this scene
from Monty Python’s Holy Grail…
“Camelot!”
“Camelot!”
“It’s only a model”
Why concern ourselves
    with models?
     Models are
pervasive & powerful.
Models guide emergency plans
Models guide political policy
 Models
determine
  teams
   tasks
& rewards
  in the
workplace
Models train pilots

Basic
passenger
Safety.
Mastery of
unusual
situations.
New
equipment.
Models shape learning.




  Everyone visualizes this key
  relationship as a model.
Models interpret cultural behavior


“the pursuit of
reputation… is
the overriding
preoccupation
of human life.”


                        Social Psychology
Models interpret cultural behavior

“Consumption is
always social;
partitioning of
goods … reflects
an underlying
partitioning of
society.”
                        Social Psychology
Models interpret cultural behavior
THEORY OF EMULATION
the upper class is happy all
the time;




                               Social Psychology
Models interpret cultural behavior
THEORY OF EMULATION
the upper class is happy all
the time;
the upper middle class is
happy 50% of the time;




                               Social Psychology
Models interpret cultural behavior
THEORY OF EMULATION
the upper class is happy all
the time;
the upper middle class is
happy 50% of the time;
the lower middle class is
happy 25% of the time;



                               Social Psychology
Models interpret cultural behavior
THEORY OF EMULATION
the upper class is happy all
the time;
the upper middle class is
happy 50% of the time;
the lower middle class is
happy 25% of the time;
the lowest class is
(relatively) unhappy all of
the time.                      Social Psychology
Models
guide your
financial
choices:
What will I
need to be
comfortable
20 years
from now?
Models
determine
how your
personal
choices
are
tracked
and met.
Models are an
   important
 intersection
    between
  technology
  and human
     values.
Models simplify complexity
When scientific knowledge
is too complicated for us
to fully understand.
Models make justifications
for past and future action.
Human values are
expressed as variables in
the model.
Models make justifications
for past and future action.
 “What if we do this?
 What if that happens?
 What does the model
 show?”
If models
simplify
complex information
and offer
justifications for action…
…then what questions
should you ask of a
model?
Our goal:
To develop questions
that can be applied to
any model from any
discipline.
I. Types of Models
II.Definitions of Models
III.Questions
IV.When to Question
 I. What are the
different types of
     models?
(we encounter all of these
      in childhood)
A model is a rough representation of reality.


                            The purpose
                            is just play,
                            fun, taking on
                            a challenge.
 A model is an exact representation of reality.




The goal is
exactitude. Details &
scale are important.
A model can become
a big project.
A model is a convenient representation of reality.




                     Not to scale.
                     Conceptually
                     easier to grasp
                     because not to
                     scale.
A model is a
normative
representation
of reality.


Variant forms and
details are omitted.
A model is a focused representation of reality.
 A model is a focused representation of reality.




Among all possible relationships, one is featured.
         A model can be a meta-metaphorical
                     representation of reality.

“The Circulatory
System is the
main
transportation
and cooling
system for the
body.”
         A model can be a meta-metaphorical
                     representation of reality.
“The Red Blood
Cells act like
billions of little
UPS trucks
carrying all sorts
of packages that
are needed by all
the cells in the
body….”
         A model can be a meta-metaphorical
                     representation of reality.

“White Blood
Cells are the
paramedics,
police and
street cleaners
of the circulatory
system.”
                         Analogous relationships
                         are featured.
By compiling our experiences, we see
6 aspects of models:
Rough – artistic, play, fun
Exact – artistic, but also predictive use
Normative – heuristic, applicable to all cases
Convenient – conceptually easier to grasp
Focused – conceptually easier to grasp
Meta-Metaphorical – conceptually easier to grasp
We can focus on 4 of these aspects:
Rough – artistic and fun
Exact – artistic and predictive
Normative – applicable to all cases
Convenient, Focused,
& Meta-metaphorical –
                easier to grasp
The Sony
Aibo: fun,
trainable,
the everydog
…& easier to
manage, too!
I. Types of Models
II.Definitions of Models
III.Questions
IV.When to Question
 II. Definitions of
      models
(What do professionals say
     about models?)
The give the elements
    of models, the
rationale for models,
   and the uses of
       models.
Elements of models
In mathematics, model theory is
the study of the representation of
mathematical concepts in terms of
set theory...
Wikipedia
 Elements of models
…It assumes there are pre-existing
mathematical objects, and asks
what can be proven given the
objects, the relationships amongst
the objects, and a set of axioms.
Wikipedia
Elements of models

•pre-existing objects
•relationship
amongst the objects
•a set of axioms
 Rationale for models
A method of expressing
relationships when measuring the
real world is impractical.
mc2consulting Company
 Rationale for models
Models help us to visualize the
problem, to break it down into
discrete, manageable units…Like
any other instrument, a model
assumes a specific intention of its
user.
Educational Design
Rationale for models

•Measuring the
real world is
impractical
•Visualizing the
problem, making it
manageable
 Uses for models
A prototype or surrogate of a
complex situation. It can be a
physical model, such as an
architectural model of urban
design, or a mathematical model of
interactions of many variables...
Course on Future Studies, University of
Arizona
 Uses for models
It can be a stand-alone tool to
evaluate different approaches
using different assumptions.
Course on Future Studies, University of
Arizona
Uses for models
A representation of the
components of a process, system,
or subject area, developed for
understanding, analysis,
improvement, and/or replacement
of the process.
Interoperability Clearinghouse
 Uses for models

•to evaluate different
approaches using
different assumptions.
•for understanding,
analysis, improvement,
and/or replacement of
the process.
I. Types of Models
II.Definitions of Models
III.Questions
IV.When to Question
How do we question models?
TYPES:
Rough – fun
Exact – predictive
Normative – applicable
Convenient, Focused, &
Meta-Metaphorical –
easier to grasp
How do we question models?
TYPES:                   ELEMENTS
                         •pre-existing objects
Rough – fun
                         •operations amongst the objects
Exact – predictive       •a set of axioms
Normative – applicable
Convenient, Focused, &
Meta-Metaphorical –
easier to grasp
How do we question models?
TYPES:                   ELEMENTS
                         •pre-existing objects
Rough – fun
                         •operations amongst the objects
Exact – predictive       •a set of axioms
Normative – applicable   RATIONALE
Convenient, Focused, &   •real measuring is impractical
Meta-Metaphorical –      •to visualize & manage the
easier to grasp          problem
How do we question models?
TYPES:                   ELEMENTS
                         •pre-existing objects
Rough – fun
                         •operations amongst the objects
Exact – predictive       •a set of axioms
Normative – applicable   RATIONALE
Convenient, Focused, &   •real measuring is impractical
Meta-Metaphorical –      •to visualize & manage the
easier to grasp          problem
                         USES
                         •evaluate approaches using
                         different assumptions.
                         •for understanding, analysis,
                         improvement, and/or
                         replacement of a process.
Questions: 5 are about information
1)What kind of data is being
 represented?
2)Of this set, is all data being shown,
 or is only some data being shown?
 Is the data exact or normative?
Questions: 5 are about information
3) Which data are being tracked and
 measured in real time, and which
 are generated by the model’s
 relationships and axioms?
Questions: 5 are about information
4) Which relationship is being
 highlighted? Is this relationship
 linked analogously to another
 model?
5) If the ultimate purpose of the
 model is predictive, is it possible to
 prove and/or falsify the data?
Questions: 3 are about interest
6) Who commissioned the model?
7) What was the purpose of the
 model? Was understanding,
 analysis, improvement or
 replacement of the current system
 explicitly mentioned?
8) Is the new model replacing an
 existing model?
Now we’ve
developed
some
questions.
How do we
apply them?
Information:
What kind of
data is being
represented?
Information:
Is the data
exact or
normative?
Information: Are the data are being tracked and
measured in real time? Are the data generated
by a model or by instruments?
Information:
Is it possible
to prove
or falsify the
predictions?
Information:
Can these
relationships
be linked by
analogy to
another
model?
Interest: Who commissioned the model?
Interest:
What was the
purpose of the
model?
Interest:
Is the new
model
replacing
an existing
model?
Do all the questions
apply to all models?
YES.
But if you can only ask
2 questions….
Information
Is there any data in the model which
is not falsifiable?
Interest
Is replacement one of the stated
purposes of the model?
I. Types of Models
II.Definitions of Models
III.Questions
IV.When to Question
When
Models
Go Bad…
5 Reasons
to question the model
1) When something just
looks wrong.
Godzilla in the bathtub
5 Reasons
to question the model
2) When there is only one
possible choice to be
made.
5 Reasons
to question the model
3) When data
consistently cannot be
explained.
5 Reasons
to question the model
4) When the model is not
elegantly economical.
The Ptolemaic Universe
The Ptolemaic Universe
5 Reasons
to question the model
5) When the category
“human” is in question.
5 Reasons
to question the model
When a group of “former”
humans are re-classified
as non-human.
(losing rights)
Cesare Lombroso’s criminal types
5 Reasons
to question the model
When a group of
non-humans are
re-classified as human.
(obtaining rights)
So long,
and
thanks
for all
the fish.
Great Model Shifts
 Happen When…
1) When something just looks wrong.
2) When there is only one possible choice.
3) When data consistently cannot be explained.
4) When the model is not elegantly economical.
5) When the category “human” is in question.
IMAGE CREDITS

Cindy Crawford
http://cutechoice.com/cgi-bin/show.pl?celebname=Cindy_Crawford&pno=8


Limits to Growth
http://www.healthtreasures.com/limits-to-growth-preface.html


Monty Python’s Holy Grail
http://www.intriguing.com/mp/holygrail.asp


Katrina Track University of Wisconsin-Madison
Cooperative Institute for Meteorological Satellite Studies
http://cimss.ssec.wisc.edu/tropic/archive/montage/atlantic/2005/KATRINA-track.gif


World Gas Reserves, Society of Petroleum Engineers
http://www.spe.org/specma/binary/images/2919957WORLDGASRESERVES.gif
http://www.globalclassroom.org/hemo.html


Myers-Briggs Personality Profiles
http://www.uwsp.edu/education/lwilson/learning/graphics/kirbys1.gif
IMAGE CREDITS
Pilot Training
http://www.link.com/gallery/f22_fmt_high_res.html


Supply and Demand
http://www.sci.wsu.edu/math/Lessons/SupplyAndDemand/theory.html


Notes on Miller
Personal notes from ES 10

Insurance Life Expectancy
http://web2.minnesotamutual.com/IMAGES/EDUCATE/LIFE_EXPECTANCY.GIF


My Simon
http://www.mjweber.com/Confessionsmm/bots/bots.htm


Popsicle Projects
http://pacifi.ca/models.html
http://www.tinypineapple.com/archives/2003/10/disneyland_day_five_a_theme_park_with_a_
theme.html

Car Models
www.oakridgehobbies.com
IMAGE CREDITS

Solar System Models
http://www.brownmunoz.com/Universe/Universe_intro.html


Visible Woman
http://www.plumcreekmarketing.com/07science/01images/fullsize/160321.jpg



USGS Water Cycle
http://ga.water.usgs.gov/edu/watercycle.html


Circulatory System
http://www.globalclassroom.org/hemo.html

Aibo and Aibone
http://gamma.cs.unc.edu/COMP290-58/aibo.jpg
http://www.mobilityparty.com/mobility_2003/press/

Axiom
http://www.jtauber.com/blog/2005
IMAGE CREDITS

Ocean
http://www.spartanburg2.k12.sc.us/OES/Ocean%20waves%20clipart.jpg
http://hyperphysics.phy-astr.gsu.edu/hbase/waves/watwav2.html

Replacement Model
http://web.pdx.edu/~stipakb/download/PA557/ReplacementModelExample.gif


Mae West
http://photos1.blogger.com/img/96/3696/320/Mae%20West.jpg


Godzilla
http://www.freerepublic.com/focus/chat/1160331/posts


Lombroso Criminal Types
http://ledroitcriminel.free.fr/utilitaires/iconographie/iconographie_1/lombroso_1_jpg_view.htm


Dolphins – So Long, and Thanks for All the Fish
http://www.spawar.navy.mil/sandiego/technology/mammals/animals.html

Ptolemaic Universe
http://www.luminarium.org/encyclopedia/medievalcosmology.htm
http://abyss.uoregon.edu/~js/ast121/lectures/lec02.html

								
To top