The Use of Chaining and Reinforcement in an Obstacle Course

1
The Use of Chaining and Reinforcement in an
Obstacle Course Vivian Scott Elishia
Webster Randolph-Macon Womans College Lynchburg,
VA 24503
Results Paired sample t-tests were run in order
to examine the difference in the number of Froot
Loops needed to complete the obstacle course. The
difference from Day 4 (M 15.50, SD 1.73) to
Day 5 (M 12.25, SD 1.50) was statistically
significant, with t(3) 3.15, p lt .05. The
difference from Day 5 (M 12.25, SD 1.50) to
Day 6 (M 10.00, SD 4.32) was not
statistically significant, as t(3) 1.25, p gt
.05. The difference between Day 6 (M 10.00, SD
4.32) to Day 7 (M 3.5, SD 2.52) was
statistically significant, with t(3) 3.43, p lt
.05. Additionally, a paired sample t-test was
conducted to examine the overall difference from
Day 4 (M 15.50, SD 1.73) to Day 7 (M 3.5,
SD 2.52), the final day of the experiment. The
difference was statistically significant, as t(3)
8.15, p lt .05.
Introduction Shaping is a training procedure
through which the subject is reinforced for every
successful approximation of the desired behavior
that is produced (Chance, 2006). Every small
approximation was immediately reinforced by
providing about ¼ of a Froot Loop. Chaining is
used to teach the subject to perform each action
in a particular sequence (Chance, 2006). The
present study used forward chaining. Each link of
the chain was reinforced and the last act in the
chain (entering the exit box) produced the
delivery of a strong reinforcement (pile of Froot
Loops). We used continuous reinforcement (CRF) to
reinforce the desired behavior each time it
occurred. Once the behavior was produced at a
steady rate, the ratio of Froot Loops were
decreased by stretching the ratio. Today,
shaping and chaining is used to train animals and
even humans. Carl Cheney (1978) shaped and
chained series of behaviors including climbing
ladders, walking across a rope, crawling through
a tunnel and entering an elevator. We
hypothesized that through the use of shaping and
chaining the subject will successfully complete
the obstacle course without any reinforcements.
Our second hypothesis was that the number of
reinforcements needed to complete the obstacle
course would decrease each day.
Method Subject The subject was a male,
Long-Evans rat who was obtained from the Harlan
Breeding Company. He was on food deprivation for
the duration of the experiment in attempts to
keep him at 85 of his normal body weight. Water
was provided ad libitum. Apparatus The obstacle
course consisted of a large central box (L21 x
W17 x H12) with two smaller boxes (both
measuring L11 x W5 x H4) at opposite
corners as the starting point and ending point
for the obstacle course. Inside the large central
box there was a small tunnel (3 ½ x 3 ½), a
box (also L11 x W5 x H4) with a ramp on one
end and a tower (11 tall) consisting of five
blocks. Procedure For three days the rat was
shaped through the obstacle course through
continuous reinforcement by providing a trail of
Froot Loops cereal pieces. During Day 1 we ran 5
trials, on Day 2 we ran 7, and on Day 3 we ran 8.
A trial consisted of placing the rat in the entry
box and dropping Froot Loops along the correct
path. A trial was over when the rat correctly
finished the obstacle course by entering the
ending box or when the rat deviated from the
correct path. On Day 3 the rat correctly finished
the obstacle course. The baseline number of Froot
Loops it took to guide him through the maze was
18. The following 4 days the researchers
gradually removed Froot Loops and stretched the
ratio of reinforcement, until Day 7, when the rat
correctly finished the obstacle course with no
reinforcement at all.
Figure 1
Discussion -Hypothesis generally supported, as a
significant drop in reinforcement needed occurred
on two of the three days -Hypothesis supported,
as successful completion of the course occurred
from Day 4 to Day 7 -Limitations include small
number of subjects, environment was distracting
with noise -Possible problem may have been the
sheer amount of reinforcements needed, as the
subject began to show less interest the longer
the experiment went each day -Future research
could examine any differences between the use of
forward chaining and backwards chaining
References Chance, P. (2006). Learning and
behavior Active learning edition (5th ed.).
Belmont, CA Thomson/Wadsworth