Transcript Document

Chain Me, Shape Me,
Control My Stimuli!
Tristram Jones, Ph. D.
Unit VI, PS 517, Kaplan University
What is Chaining?


Chaining is the
reinforcement of
successive elements of
a behavior chain. If
you are teaching your
child to tie shoelaces,
you are attempting to
build a chain. Karate
“kata” are chains! So is
driving to work 
There are two chaining
procedures, forward
and backward
chaining.
Forward Chaining:

Forward chaining is a chaining procedure that begins with
the first element in the chain and progresses to the last
element (A to Z). In forward chaining, you start with the
first task in the chain (A). You then teach A AND B. No
concept is taught in isolation!
Backward Chaining
Backward chaining a
puzzle gives the child
the idea of what success
looks like by going
backwards from
completion. Got that? 
Chaining this way gives
a clearer idea of the next
step. Mom would be
reinforcing each step as
she teaches it, but
daughter learns last
step (first), next to last,
and so on, always
adding steps together.
Sometimes called reverse chaining…

Skinner reverse-chained a monkey to jump off a
ledge, catch a rope, swing to a platform then eat a
banana. But he did it backwards. First the monkey
was placed on the platform, where it climbed to the
banana. Then he was placed on the rope, where he
learned to swing to the platform, climb up and get
the banana. Then he was placed on the ledge, and so
on!
How to chain:
Define the
target
behavior!
To teach
someone to
perform the links
of a chain, you
need to know
what those links
are. Breaking the
chain into small
manageable
steps is called
“task
analysis.”
(Great for autistic
kids)
Reinforce!
Reinforce
successive
elements of the
chain: The
elements in the
chain must be
reinforced in
sequence.
Reinforce them as
they happen.
Monitor Results
You must keep track of the
effects of your efforts. Has
a particular element been
mastered? Should it be
taught and reinforced a few
more times? Is it time to
move on to the next
element? These judgments
must be made during the
chaining process, and can
be made accurately only by
carefully monitoring the
results.
And now for something completely
different: DIFFERENTIAL
REINFORCEMENT
INCREASE A DESIRED BEHAVIOR RELATIVE TO AN
UNDESIRED BEHAVIOR!
But does Lindsay qualify?

To differentially reinforce we must have three
things in place:

Two or more different physical behaviors
One behavior is reinforced
The other behaviors are extinguished!


So if we want Lindsay to get in less trouble we might reinforce drinking
at home instead of all over town, is that DIFFERENTIAL
REINFORCEMENT?
EXTINGUISH HER
DRINKING OUT ON
THE TOWN
PRAISE HER WHEN
SHE DRINKS AT
HOME
NOT Differential Reinforcement!


Why not? Because drinking is only one
behavior, basically! We needed TWO!
So what if we switch to reinforcing her
sobriety only? No good, that’s still
cheating. we need a second behavior!
So what if we reinforce her when she
dresses nice, and use an extinguishing
tactic on the drinking?
WOW, I’m
the next
GRACE
KELLY!!


That's THE IDEA! WE
HAVE 2 distinct behaviors!
ONE is REINFORCED, the
other is Extinguished
(sort of) and WE HAVE
DIFFERENTIAL
REINFORCEMENT!!!
One more technicality!

You cannot extinguish a behavior simply by
ignoring it—you must eliminate a previously
extant reinforcer for it!
Britney must GO!!!
So what about SHAPING?

Shaping is an operant
conditioning technique for
creating an entirely new
behavior by using rewards
to guide an organism
toward a desired behavior
by successive approximation
so that the organism is
rewarded with each small
advancement in the right
direction. Once one
appropriate behavior is
made and rewarded, the
organism is not reinforced
again until a further
advance is made, then
another and another until
the entire behavior is
performed.
Example of shaping:
1.simply turning toward the lever
will be reinforced
2.only stepping toward the lever
will be reinforced
3.only moving to within a
specified distance from the
lever will be reinforced
4.only touching the lever with any
part of the body, such as the
nose, will be reinforced
5.only touching the lever with a
specified paw will be
reinforced
6.only depressing the lever
partially with the specified
paw will be reinforced
7.only depressing the lever
completely with the specified
paw will be reinforced
Think of it this way:

SHAPING is a lot like a child’s game of
“hot and cold” in which the blindfolded
subject is brought closer and closer to a
desired outcome by successive verbal cues
that suggest approval or disapproval.
Chaining is more like assembling a
model, or flat-packed furniture!
And as for
stimulus control….
WE DON’T WANT our subjects to emit
behaviors at the wrong times—or in
places they
won’t be
reinforced!
Take the sorry
plight of THIS
out of towner!

So what is discrimination training?



It's the process of teaching subjects when to emit a
conditioned behavior.
KING KONG’s behavior made him a big hit on his island,
but it didn’t work out in nyc. Behaviors that would have
been reinforced at home got him killed in Manhattan.
So discrimination treatment teaches
subjects WHEN they should emit a
behavior—and….
Then there’s GENERALIZATION

This is when the subject’s
impression that a behavior will
garner reinforcement spreads
by generalization across a
number of similar behaviors.
We can’t teach that a behavior
will be reinforced in each
circumstance—but
generalization does the job for
us! EXCEPT when
it DOESN’T! 
Much generalization occurs as a
result of general reinforcement!

A generalized reinforcer is a conditioned
reinforcer that has obtained reinforcing
function paired with many other reinforcers
(such as money, fame, applause, etc.)
Elijah thought all animals were
dogs!
DOGGIE??
Generalization
is often
extinguished
by simple
experience
(still a lack of
reinforcement
that previously
existed!)
Fading and Programming

Make it easier to teach discrimination!
The concept of fading refers to the use of a
prompt to establish a specific
discrimination. "fading out this stimulus is
used in hopes that the desired behavior
continues to be reinforced as the
discriminative cues are gradually
withdrawn.”
Like, what if your subject won’t
respond to a stimulus, so you
CANNOT reinforce the response

Your experiment
will not work!
The dang mouse
won’t push the
button, so you
can’t feed him a
pellet! UH-OH!
But you can use FADING!

You FIRST use a prompt!
You tap the button and
eventually the mouse
pushes it! Then you can
reward him. Begin
tapping less severely,
then just point—then nod
at the button—you are
fading, get it?
What’s an SD??

Discriminative stimulus or SD: A stimulus
that controls the probability of a response.
Have a chimp press a key for food.
Then light the key green when the
best result are offered. As the chimp
gains more experience with the
contingencies, he presses less when
the green light is off and finally more
when the key is lit green. The green
illumination of the key proves
CHIMPS AREN’T COLOR BLIND! No,
seriously, it becomes an SD because
response probability (as reflected in
the rate of responding) is now much
higher in its presence than in its
absence.
Program with prompts, then…
FADE AWAY
Programming…as in learning

Two flavors: linear (Skinner’s)
And intrinsic, or branching….

Developed by Norman Crowder as electronic training
for the military. Provides a problem requiring
multiple choice. A student who responds incorrectly
may be routed through a subprogram designed to
remedy the deficiency in his original answer.
In closing, have you ever been
linearly programmed?

I just asked you if you have ever been
___________programmed.