Behavioural Processes,
15 (1987) 13 I-142
131
Elsevier BEHAVIORAL CONTRAST IN RATS REINFORCERS ARE USED Jennifer
J. Higa
Psychology (USA)
WHEN
and Frances
QUALITATIVELY
DIFFERENT
K. McSweeney
Dept., Washington
State
Univ., Pullman,
99164-4830
WA
(Accepted 10 June 1987) ABSTRACT Higa, J. J. and McSweeney, F. K. (1987). Behavioral contrast rats when qualitatively different reinforcers are used. Behav. Process.,15:131-142.
in
Behavioral contrast was studied during multiple schedules which provided qualitatively different reinforcers in the two components. Five rats responded on a baseline schedule in which both components delivered food reinforcers (food-food), and then on a contrast schedule in which one component delivered food and the Following this, baseline was other delivered water (food-water). recovered. Five other rats responded on a baseline schedule in which both components delivered water reinforcers (water-water), then on a food-water schedule, and then on the baseline, waterwater, schedule. Contrast was not observed when relatively low rates of reinforcement were used but it was sometimes observed when The rate of responding for high rates of reinforcement were used. a constant water reinforcer decreased when food replaced water in the other component. The rate of responding for a constant food reinforcer did not change when a water reinforcer replaced food in the other component. These results are similar to those reported by Ettinger and McSweeney (1981) when pigeons served as subjects. INTRODUCTION Behavioral in response
contrast
rate during
schedule
when
Positive
contrast
during
reinforcement
during
occurs
component
improves.
1961).
Only
0376-6357/87/$03.50
the other
component
worsens.
rate during reinforcement
have examined
a few have
contrast
Sutterer,
examined
is varied.
0 1987 Elsevier
Science
Publishers
baseline,
during
the rate of Reynolds,
occurs
Ettinger,
B.V. (Biomedical
component
the other
1975;
contrast
rate
when
Negative
by changing
For instance,
is altered.
the unchanged
& Brush,
whether
of a multiple
if the response
above
(e.g. Gutman,
type Of reinforcer
1961) occurs
increases
when
or decrease
component
component
baseline,
IMost studies
component
in the other
if the response
below
as an increase
one, unchanged,
(e.g. Reynolds,
decreases
reinforcement
be defined
the reinforcement
the unchanged
contrast
may
when
the
McSweeney,
Division)
132
and Norman
(1981) observed
type of reinforcer moderately
grain
Similarly,
(oats).
grain
(split peas
or
when
presented
negative
mixed study
baseline
phase
component.
However,
water
McSweeney,s two
reasons.
Some
similarly
for all
present
using
rate
19671,
and
these
theories
produce
Two
reinforcement variable
reinforcement past
But,
can affect
was
studies
have
shown
rates
of reinforcement (1983) also
high baseline had
failed
it when
contrast Second,
conditions indicate
(Chung
of
occurs the
produce that
changes
& Herrnstein,
whether
a change
in
The present
differed
variable
used
was
interval
a lower
was
60-second
greater
when
used,
baseline
Dougan, higher
pecking
treadle-press was
of
not observed.
rate of reinforcement
McSweeney,
for pigeons
positive
only in
VI 60-s VI 60-s) schedule
11, contrast
rate of reinforcement
to find
for
1978) of the reinforcer
that baseline
contrast
were
and
the generality
1973).
what
of contrast.
found
the changed did not change
Ettinger
whether
which
(multiple
(Experiment
that
gave
for water
is important
contrast.
a multiple
used
(1986) found
Then,
now
studies.
conducted
ho-second
the magnitude
examine
delay
to a
water.
during
This
few have examined
When
rates.
baseline.
component.
studies
also produces
were
Farmer
McSweeney
19611,
add to these
interval
Past
both
exposed
to replicate
& Silberberg,
However,
experiments
the chanyed
to determine
was
the other
water
(e.g. Rachlin,
(Hamilton
will
while
questioned
1973).
the type of reinforcer experiments
grain
were
for food reinforcers
experiments
helps
were
found that responding
have
species
contrast.
water
attempt
(e.g. Reynolds, size
a less
contrast
delivered
rats as subjects.
(e.g. Rachlin,
in the
when
preferred
food during
group
food replaced
food during
experiments
contrast
to deliver
when
First,
contrast.
a
food, and the other
both components
experiments
study
the
grain
For one group,
delivered
in a second
which
responding
replaced
The present
from
reinforcers
to deliver
and McSweeney
decreased
behavioral
1981).
schedule
Subjects
reinforcers
when
observed a highly
different
continued
continued
Ettinger
was with
pigeons,
h McSweeney,
during
one component food.
contrast
using
of a multiple
water.
changed
when
to a less preferred
replaced
two qualitatively
one component
delivered
in pigeons
grain).
(Ettinger
components
contrast
in one component (wheat)
(oats) was
In another observed
Then,
available
preferred
preferred
positive
Higa,
keys.
contrast
but previous rate was
and
baseline
used
when
a
studies (e.g.
133
Hemmes,
1973;
McSweeney, whena
McSweeney,
and Farmer
1978;
highbaselinerate
lower
rate
baseline
was
Westbrook,
(1985) found
of reinforcement
used.
As a result
rate of reinforcement
in Experiment
2.
1973).
lever press
Contrast
observed
Dougan,
in rats
was used,butnotwhen
of these
(multiple
was
Finally,
contrast
studies
a
a higher
VI 15-s VI
15-s) was
used
in that experiment.
METHOD Subjects
rats
Ten experimentally
naive,
served
in Experiment
as subjects
male,
90 days old Sprague-Dawley
male,
Experiment feeding water
All subjects
2.
and were
weights,
for 15 minutes
used because McSweeney
120 days old Sprague-Dawley 1.
were
maintained
water
deprived
each
session.
after
they are identical
Ten experimentally
rats served
as subjects
at 90% of their
by allowing These
to those
naive,
in free-
access
conditions
to were
used by Ettinger
and
(1981).
Apparatus The rats and
cm wide,
were
4.0 cm by 3.5 cm, center
in an operant
which
an aperture
located
the right
was
1.0 cm
2.0 cm above Two
wall.
located
4.5 cm above
dipper,
respectively.
of approximately
mounted
2.0 cm above
the component chamber
Access
the floor,
associated
was enclosed
lever
with
a fan which
equipment
scheduled
inasked
floor.
Its
to a water
dipper
was
was
11.0 cm from
and 4.5 cm above operated
were the water
by a downward
each
attenuating
extraneous
the experimental
of the aperture
5.0 cm by 1.5 cm,
inside
23.5
a grid
The bottom
that lever was
in a sound
contained
was
A light
long,
a food trough,
and its center
levers,
the food trough, Each
29.5 cm
contained
in diameter.
response
0.20 N.
chamber,
The chamber
was 5.0 cm fromtheleftwall.
through was
tested
19.0 cm high.
lever
force
indicated
apparatus
when
The
in effect.
that
Electromechanical
noise.
events.
Procedure All subjects After
procedure. multiple
were
trainedtopress
this was
VI 10-s VI 10-s schedule
was graduallyincreasedto or
to
a multV1
constructed in Catania
the rats
to a 20 interval
each
lever signalled
2.
series
(1968, Appendix
placed
This
on a
schedule
60-s for Experiment
15-s VI 15-s for Experiment
according
shaping were
of reinforcement.
a multVI60-sV1
and Reynolds
the light within
a leverbya
accomplished,
2).
when
1,
Allscheduleswere using
During
the procedure the schedules,
the component
associated
134
with s.
that
lever
Sessions
session
was
were
in effect.
conducted
terminated
after
The components
daily,
six times
40 reinforcers
were
reinforcers
consisted
of one 45 mg Noyes
reinforcers
consisted
of four
During
Experiment
VI 60-s
schedule
exposed
to a baseline
food
reinforcers
contrast right
phase
phase
(food-food).
Then,
both
components
they were
exposed
the left lever
delivered
lever delivered
water
(food-water).
Finally,
subjects
(Group 2) were
However,
implemented
during
reinforcers
exposed
baseline,
2 was
identical
VI 15-s VI 15-s schedule of rats were
the components baseline
2 was exposed
of reinforcement
during
during which
to baseline
both
phase
phase
five as Group
delivered
water
1, except
that a mult
was
phases
used.
Again,
two
in the type of reinforcer
baseline.
both
Group
components during
responding
stable
sessions
fell within
the same
phase.
1 was
exposed
delivered
which
both
food.
that to Group
components
1 needed
34.8 for the during
phase.
an average
23.2 during
to meet
In Experiment
24.4 sessions
criteria phase.
and
was
during
the baselines
and
emitted
five during
of sessions
29.6 for the baseline 2 took an average
37.0 sessions
Group
phase.
number
or
was
for the last rates
Group
2, responding
of 19.9 sessions
during
rates
1, the average
this
to a baseline
Responding
of the previous
contrast
baselines
the contrast
rat was exposed
stablized.
the response
In Experiment
phases
21.2 sessions
when
the range
that Group and
each
experiments,
until
consideredtobe
contrast
a second
water.
During contrast
to Experiment
used, differing
provided
phases
delivered
to a
food and the
contrast
components
first
delivered
The remaining
to the same
both
VI 60-s
(water-water).
Experiment
groups
(food-food).
dipper.
on a mult
which
was
Food
(Group 1) were
during
phase
after
responded
which
60-
and water
to the water
Five rats
during
every
and each
delivered.
pellet,
access
1, all subjects
of reinforcement.
baseline
1.
seconds
alternated
per week,
during
for Group
the baseline 2 required 25.4 during
of
the 1 stabilized
phases
an average
and of
the contrast
phase.
RESULTS Experiment Table
-1 1 contains
the mean
of each multiple
schedule,
These
calculated
means
were
response
rates during
for the subjects over
the components
in Groups
the last five sessions
1 and
2.
for each
135
Response
condition. of lever
presses
rates
made
component
was available.
presented
was
TABLE 1 The mean subjects
were
during
excluded
determined
The time from
by dividing
a componentbythe for which
the number
timethatthe the dipper
was
the calculations.
rates of responding in Experiment 1.
in responses
per minute
Group
for the
1
meet wonent
1
2
3
4
5
Baseline
food food
5.38 4.20
11.06 4.43
13.27 4.31
6.09 3.97
6.56 2.99
Contrast
food water
8.98 3.35
11.69 4.11
lY.17 3.74
6.96 7.70
11.23 2.16
Baseline
food food
9.12 5.01
13.75 5.95
11.61 4.94
8.38 5.16
5.84 4.67
9
10
Group
2
Subject Component
6
7
8
6.13 6.54
8.79 7.35
7.21 7.44
Baseline
water wa.ter
Contrast
water food
5.35 15.08
5.55 12.20
5.77 13.57
5.30 7.35
9.84 20.26
Baseline
water water
3.34 3.42
10.95 5.34
3.14 4.59
3.70 3.47
9.21 7.40
Table subjects Responding water
1 shows
in Group
that
increased
Fiqure durinq
from
water
decreased
replaced
for all subjects
the baseline
1 presents
the unchanged
the response
rates
for the
contrast
phase.
'The points
are plotted
response
rates.
Each
was
rate during
point
the unchanqed
obtained
component
for four of five
2 when
detected
to the contrast
component
8.13 6.04
food reinforcers.
in Group
subjects
Therefore,
reinforcers.
reinforcers
responding
1 when
6.41 5.21)
food replaced
the chanqe
in
phase.
for Groups
1 and
last five sessions relative
2, of the
to the baseline
by dividing
of the contrast
the response phase
by the
136
mean
response
rates were five days rates
calculated
taken
for a single
unchanged
by averaging
of the initial
were
contrast
rate of the same component
from
subjects.
because
they
component,
reinforcement
1.
Points
produced
Group
Each which
represent
of the other
1
response
and recovered
Table
during
baseline.
rates
baseline
during
phases.
Baseline the last These
set of axes represents differ
changes
from
1.0 indicate
in response
by alterations
results
rates
during
in the condition
of
component.
Group 2
1 16
7
SESSIONS Fig. 1. The points represent the rates of respondinq during the constant component for the last five sessions of the contrast These points are plotted relative to mean baseline rates phase. taken from Table 1. Each set of axes presents the results for a single subject.
an
137
At first
glance,
for the constant when
water
replaced
responding contrast
phase
Responding baseline
than during Table
may
component
significance
when
water
greater
for
that
during
the
to the second
in responding
measures during
for this
over
analysis
time
food,
statistical
Therefore,
in the responding
contrast
for the constant
replaced
rather
of variance
the constant,
1 did not reach
reinforcers
occurred.
the initial
(F(2,8) = 2.52, p > 0.05).
reinforers
was
indicating
in responding
repeated
1) in Table
consistently
other
1 from
of responding
not appear
Responding
line,
contrast
the increase
a fluctuation
A one-way
(Group
whether
for subject
to the rates
component.
reinforcer
responding
systematically
baseline.
Therefore,
represent
contrast.
applied
food
that
somewhat
the horzontal
1 questions
increased phase.
to indicate
changed
food in the variable
for the constant
However,
than
1 seems
1, 3, and 5 is above
subjects
subject
Figure
food reinforcers
did food
food reinforcers
in the
component. Figure
responding
1 and Table for water
1 show
thatcontrastdid
reinforcers
changed
component.
A one-way
variance
applied
to the response
component
(Group
when
food replaced
repeated
2) in Table
measures
rates
1 was
notoccurinthe
during
not
water
analysis
in the of
the constant,
significant
water,
(F(2,8) = 0.62,
p > 0.05).
Experiment
_2 2 contains
Table subjects of each
in Groups multiple
for Table changed water
schedule.
Table
1.
component in Group
component Group
the mean
2.
response
1 and 2 of Experiment
2 shows
The means that
increased However,
did not change
rates
emitted
2 during
were
by the
both
calculated
components as they were
the rate of responding
for all subjects
when
food replaced
the rate of responding
systematically
when
in the
in the changed
water
replaced
food in
1. Figure
component are plotted calculated results represent
2 presents
for the
response
last five days
relative
behavioral
1.
Again,
subject, contrast.
during
the unchanged
of the contrast
to the baseline
as for Figure
for a single
rates
response each
and points
phase.
rates,
set of axes which
differ
The points
and were represents from
1.0
138
TABLE 2 The mean rates of responding subject in Experiment 2.
in responses
Group
per minute
for each
_1
Subject 11
Component
22
33
44
55
Baseline
food food
30.72 20.24
35.11 10.62
62.77 13.23
40.07 13.44
46.02 18.05
Contrast
food water
32.59
12.38
32.82 13.79
73.22 15.67
36.41 12.55
33.24 19.29
food food
25.37 9.67
23.43 14.24
73.65 19.10
41.37 16.43
30.04 14.35
Baseline
Group
2
Subject Component
77
66
88
99
water water
17.68
34.74 36.91
14.94 17.77
41.60 22.02
24.24 20.82
Contrast
water food
10.61 51.16
6.76 52.52
8.17 35.59
10.87 45.18
10.92 50.61
Baseline
water water
8.81 8.82
14.66 10.59
13.56 18.00
7.90 8.06
12.27 10.31
Easeline
Table responding
2 and Figure
2 show
for the constant
component analysis
16.91
changed
from
of variance
reinforcers
(Group
0.62, p > 0.05).
fooa to water.
applied
2 was
Bow~ver , Table
not
2 and Figure
for the constant
the other
component
from water
was consistently for subjects variance
(F(2,8)
lower
to the responding
(Group
= 8.35,
2) in Table
p < 0.05).
during
baseline
A one-way
repeated
that
when
2 shows phase
for one point
measures
statistically
contrast
the contrast
except
food,
reinforcers
Figure
for the constant
2 was
measures
(F(2,8) =
2 show water
to food.
reinforcers
than during
66 and 1010.
applied
reinforcers
for the water
repeated
significant
in the responding
responding
in the
the other
for the constant,
did occur
that
when
A one-way
to the points
1) in Table
changed
not occur
thatcontrastdid
food reinforcers
analysis
water significant
of
139
Group
2
66
):I;-77
k 88
;_:I-:rl99
k 1010
1.0 k
012345
1
2
3
4
5
SESSIONS
Fig. 2. The points represent the rates of responding during the constant component for the last five sessions of the contrast phase. These points are plotted relative to mean baseline rates taken from Table 2. Each set of axes presents the results for a single subject.
DISCUSSION The present conditions
which
conclusions.
further
behavioral
First,
can be produced variable
results produce
the results
by changing
component
our understanding contrast
indicate
that behavioral
the type of reinforcer
of a multiple
schedule.
of the
and support
several contrast
available
Past studies
in the
have shown
140
that
changes
in the rate
Herrnstein, reinforcer McSweeney
(e.g. Reynolds,
1967),
and size
produce
contrast.
(1981) show
(Hamilton
lY61),
delay
& Silberberg,
The present
that reinforcer
results,
ty?e
(Chung
&
1978) of the and Ettinger
should
be added
and
to this
list. Second,
the results
two different reinforcer were
species,
is manipulated.
similar
(Ettinger
to those
food replaced constant
contrast
variables
components
did not change
similar behavioral
when
the types
and McSweeney decreased
water
suggests
contrast
rats
when
Rut responding
when
conditions
using
may
for
replaced
food.
that the
be similar
of reinforcers
of
across
provided
by the
are changed. contrast
Third, reinforcement
reinforcement McSweeney,
appeared
rates were
in studies
contrast
under
for
the type
as subjects
reinforcers
component.
similarly when
study
served
the Ettinger
water
in the changed
controlling
rats and pigeons
found
for constant
water
of this
pigeons
Like
1981).
occurs
at least
The results
food reinforcers
Finding
that contrast
found when
& McSweeney,
responding
study,
show
rats and pigeons,
which
used.
Dougan,
Higa,
failed
contrast
were
which
greater
predict
(eg., Herrnstein,
contrast
1986).
responding
In those
rates
that
of
with
theories
of reinforcement
these
for different
1983;
studies,
rates
is compatible
at higher
to that
the rate of
1985; McSweeney,
low baseline
It suggests
1970).
to handle
when
This result
used.
is similar
by varying
& Farmer,
& Farmer,
to occur
reinforcement
high but not low baseline
This result
produced
(Dougan, McSweeney,
also
generalized
when
theories
types
should
be
of
reinforcers. The failure reinforcers used
to find contrast
even when
is difficult
changed
component
reinforcers
reinforcers
high baseline
for food
rates of reinforcement
to interpret.
The rate of responding
did not change
systematically
replaced
may have occurred
in the responding
either
replace
because
food,
the failure
Therefore,
food.
contrast
or because
when
does
subjects
were
during
the
water
to find contrast
not occur
when
did not detect
water this
change. Although results contrast found
the present
reported
does not occur
significant
variable
results
by Ettinger
component
when
decreases when
cannot
and McSweeney water
resolve
this
reinforcers
replace
in the rate of responding
water
replaced
issue,
(1981) suggest
food,
that
food. during
They the
but they did not find
141
contrast
in the constant
The present may
indicate
effects
results
that
on behavior.
Kagel, that
not
effect
affect
substitutable
another
water
food
changes
water.
does not change
are consistent
for one
(e.g. food
(e.g. Hursh,
theories another
propose
(e.g. two
one way. and
with
that qualitatively
differently These
responding
with
(1981) found
responding
component
and IMcSweeney responding
Other
water)
theories,
increases
varied
the
will
1980;
that
Rachlin,
reinforcers
different
affect
economic
different
reinforcers
work
component
needs
types which
of are
responding
behavioral
water
changes
on this
have
been
reinforcers
if different
conditions
deprivation
levels)
used.
were
when
found
food replaced
The present
theories, even
Ettinger
only decreases
reinforcers).
subject.
of
grainprovidedby
found
reinforcer
and
in the rate
reinforcers).
study
(non-substitutable
of the economic
might
McSweeney,
whenthetypeof
(substitutable
to be done
the details
Ettinger,
and decreases
(1981) and the present
for a constant
in the other
not test
these both
for one type of grain
the other
more
with
with
delivers
(1981)
different
way.
Consistent Norman
that
YcSweeney
Froduce
reinforcers
findings
suggest
1980).
and
reinforcers.
behavior
substitutable
will
water
a component
these
which
& Battalio,
are
grain)
for food way,
of behavior
reinforcers
during
of Ettinger of reinfcrcers
food reinforcers
rate of responding
theories
those types
Replacing
replacing
In a general
and
different
the rate of responding However,
component.
in
water
However, results
for non-substitutable
(e.g. rates
of reinforcement
REFERENCES G. S. (1968). A quantitative Catania, A. C. and Reynolds, analysis of the respondinq maintained by interval schedules of Analysis of reinforcement. Journal -Of the Experimental -Behavior, 11, 327-383. Chung, S. H. and Herrnstei -n, R. J. (1967). Choice and delay of reinforcement. Journal -of the Experimental Analysis of Behavior, lo, 67-74. Dougan, J. D., McSweeney, F. K., and Farmer, V. A. (19 85). Some warameters of behavioral contrast and allocat ion of interim behavior in rats. Journal _of the Experimental Analysis of -Behavior, 44, 325-335. Behavioral contrast Ettinger, R. H. and McSweeney, F. Ii. (1981). and responding during multiple food-food, food-water, and water-water schedules. Animal Learninq and Behavior, 2, 216^..^ Ettinger,
R. H., McSweeney,
F. K., and
Norman,
do
and symmetrical
W. D. (1981).
or
142
Contrast and undermatching as a function of reinforcer duration and quality. Journal -__ of the Experimental Analysis of Behavior, -___ 271-282. x, Gutrnan. A., Sutterer. J. R.. and Brush. F. R. (1975). Positive and negative behavioral contrast in the rat. Journal of the Experimental Analysis of Behavior, 23, 377-383. -__ iiamilton, B. E. and Silberberg, A. (1978). Contrast and autoshaping in multiple schedules varying reinforcer rate and duration. Journal ___ of the Experimental Analysis of Behavior, ~___ 30. 107-122. He mG's , N . (1973). Behavioral contrast in piyeons depends upon the operant. Journal of Comparative and Physiological Psycholoqy, ill, 171-178. Eiursh, s. 1~. (1980). ticonomic concepts for the analysis of behavior. Journal -__ of the Experimental Analysis of Behavior, -__ 34, 219-238. McSGeney, F. K. (1978). Negative behavioral contrast on multiple treadle-press schedules. Journal -~ of the Experimental __-___ Analysis of Behavior, 2, 463-473. McSweeney, F.K. (1983). Positive behavioral contrast when pigeons press treadles during multiple schedules. Journal of the Experimental Analysis of Behavior, 39, 149-156. McSweeney, F. K., Dougan, J. I)., Higa, J., and Farmer, V. A. (1986). Behavioral contrast as a function of component duration and baseline rate of reinforcement. Animal Learning __-and Behavior, 2, 173-183. 7 Rachlln, I-I.(1973). Contrast and matching. Psychological 217-234. Review, x, Racm,-tl., Kagel, J. H., and Battalio, K. C. (1980). Substitutability in time allocatjon. Psychological Review, 87, 355-3'74. Behavioral contrast. Reynolds, G. S. (lY61). Journal -__ of the Experimental Analysis of Behavior, 4, 57-71. to obtain positive contrast Westbrook, R. F. (1973). Xailure the Experimental when pigeons press a bar. Journal of -Analysis of Behavior, 3, 499-510. -