- Email: [email protected]
Positive contrast as a function of component duration using a within-session procedure
Behaviourul Processes, 16 (1988) 21-41 Elsevier POSITIVE
CONTRAST
WITHIN-SESSION
FRANCES
AS A FUNCTION
K. MCSWEENEY
Accepted
OF COMPONENT
DURATION
USING
A
PROCEDURE
AND
CAM L. MELVILLE
Department of Psychology, WA 99164-4830 , U.S.A.
(
21
10 November
1987
Washington
State
University,
Pullman,
)
ABSTRACT
McSweeney, F. K. and Melville, C. function of component duration using Behav. Process., 16 : 21-41
L. Positive contrast as a a within-session procedure.
Pigeons pecked keys for food reinforcers delivered by multiple schedules. Three experiments examined changes in the size of positive contrast with changes in component duration. They used a within-session procedure which measured rates of responding during the baseline and contrast phases within the same session. Contrast was largest for shorter components, a result similar to that found using traditional procedures. Therefore, the present results provide preliminary support for the use of the withinsession procedure to study contrast. They also showed that the inverse relation between key-peck contrast and component duration is a robust finding which survives major changes in procedure. Finally, they suggest that different functions relate the size of contrast to component duration for key pecking and treadle presssing. INTRODUCTION The yielded
study few
for this whether
failure
may
contrast
circumstances. failure
of multiple-schedule
generally
Such
to observe
have used
the more contrast
studies
studies
few
under
approach
to independent would
little
examine
only
particular because
to many
information. of examining
variables.
constrain
has
of the reason
conclusions
can be attributed
productive
contrast Part
frequently
to occur
yield
it provides
studies
behavioral
conclusions.
or fails
contrast
Therefore,
such parametric
be that
occurs
factors.
relating
accepted
different Few
studies
the functions
Information
theorizing
the
from
to a greater
degree.
---------Preparation of this manuscript was partially supported by NIMH Thedata werepresented at the1987 meeting of grantMH42466. the Psychonomic Society. The authors wish to thank Dr. John Hinson for his comments on an earlier version of this paper.
0376-6357/88/$03.50 0 1988Elsevier Science Publishers
B.V. (Biomedical
Division)
22
The time
required
to measure
to the failure
to conduct
is an increase
in responding
multiple
schedule
reinforcement decrease
in the other
contrast,
contrast
schedule
in responding
represented
contrast
of contrast,
and
fluctuations
study.
in response
rates
example,
subjects
may double
schedule
to its recovery
Spealman
and Gollub,
To encourage experiments problems. measures
a new
The within-session
schedules
1983;
Heyman,
equipment to
studies
measure
used
1983).
over
for a five-
time.
rates
by
For
from
one
et al., 1986;
phases
has been
similar
used
procedures
reduces
these
procedure
which
single
sessions.
within because
other
to study
(e.g., Ettinger
The replacement control
the present
which
use a within-session
of reinforcement
by computer
measurement
is confounded
of contrast,
of contrast
and contrast
approach
have successfully other
to 50
schedule.
for each
(e. g., McSweeney,
The experiments the baseline
for each
occur
First,
Thirty
1974).
parametric
test
drawbacks.
response
over
1986).
are required
which
their
that phase
in responding
the measure
a
of
the conditions
to conduct.
are required
Second,
Then
to the contrast
has two
subject
provides
to ensure
and Farmer,
450 to 750 sessions
parametric
and varies
fluctuations
sessions
90 to 150 sessions
which
components.
is recovered
Higa,
To
the conditions
the baseline
for each
is a
with
is conducted
holds
procedure
many
are conducted
Therefore,
point
requires
of a
of
contrast
component
in both
constant
than
contrast
of reinforcement.
which
Dougan,
This across-sessions
Negative
schedule
from
rather
(e.g., McSweeney,
sessions
of the conditions
the baseline
the changes
the procedure
Positive
of reinforcement
Finally,
contributed
component
conditions
in one component
in the other.
have
studies.
a constant
is conducted
may
a constant
component.
a baseline
conditions
reinforcement
during
during
in the other
the same
time
parametric
a worsening
in responding
improvements measure
with
contrast
authors
responding
and
on
Staddon,
of electromechanical
also makes
such procedures
easier
conduct. The present
experiments
the within-session whether produces component Positive
this
procedure.
procedure
the same duration contrast
offer
The experiments
produces
function
a relatively
contrast,
relating
for short
test of
determine
not only
but also whether
positive
as the traditional is largest
strong
contrast
across-sessions components
when
it
to procedure. pigeons
23
peck
keys
for
food
Kodera
and Rilling,
1971;
Spealman,
replicate
this
reinforcers
1976). finding
The present using
Three
experiments
procedures.
Different
replications
of the results
They are also between
and component that contrast
changes
responses
will
more
of Herrnstein's
for shorter
components
should
produce
observed been
more
for short
observed
Therefore, pecking,
peck
experiments
theories
might
larger
components. also
be
The
larger
m parameters
contrast
press
is not
treadles,
(McSweeney,
replicated
that
and terminal
and larger
it has
et al., 1986).
these
successfully
of
components
(1982) argued
should
pigeons
keys
theories when
for shorter
Although when
procedure.
Several
of interim
than for longer,
components
if future
different
of the relation
be largest
(1970) theory
pigeons
procedure.
using
and Staddon
contrast
to
to provide
duration. should
1982;
and Wheatley, attempt
the robustness
contrast.
when
these
are used
Ettinger
m parameter
experiments
within-session
in the competition produce
Shimp
of the within-sessions
to determine
For example,
and Staddon,
1982;
this function
procedures
predict
short.
dynamic
the
examine
contrast
contrast are
used
(Ettinger
McSweeney,
1976;
results
describe
for key
key-peck
contrast. Different procedural contrast
Positive
inversely
with
were
duration
the different
duration
contrast
duration
whether
functions
for key pecking
and negative
component
component
also used to determine
produced
to component
pressing.
with
procedures
differences
generally
for key pecking,
for treadle-pressing
relating
and treadle vary but directly
(McSweeney
et al.,
1986). The different warrant
investigation
theories
which
responses.
a simple
failure might
between
low
rates
occurs
functions
contrast
to an insensitivity
can easily
functions
1982),
for treadle explain
of component
why
pigeons
press changes
(e.g., Davison
of treadle or to poor pressing. contrast
duration
simple
For example,
in detecting
pressing
to
for all
out many
when
pressing
damaging
responses.
to difficulty
of treadle
(e.g., Staddon,
similarly
rules
between
to find positive
the components
explanations different
different
and treadle
they are particularly
contrast
be attributed
1978),
reinforcement
that
for key pecking
for the differences
treadles
in generally
because
argue
Finding
explanations
Ferguson,
functions
pressing
and to
discrimination
But, none should
of these
change
for the two
as
responses.
24
Therefore, differ This
the differences
in important
would
contrast
question
occurs
Herrnstein,
pressing
contrast several study
ways
from
and
operandum
for key
components Finally,
interval
contrast.
than
pressing,
interval that
reinforcers
for key
15
used
presented
pressing
but
timeout
but not
stimuli
The
to study
2-mext).
were
A three-second
pecking.
for treadle
schedule
in
of reinforcement
variable
for treadle
positive
to differ
2-mVI2-mandmultipleVI
operanda
treadle
in the
had
rate
extinction)
1983). and
to produce
key-peck
a higher
that
(e.g.,
Williams,
In order
15 second
the discriminative
produced
1978;
may
pressing.
for key pecking
to study
15-s
contrast
assume
responses
the experiment
used
VI
all
which
by differences
them.
of the multiple
response
for
that
and treadle
of contrast
Staddon,
used
(multipleV1
different
pecking
produced
those
suggest
functions
pressing,
multiple
The components
which
been
pressing
variable
key pecking
and
to measure
for treadle
(multiple
ways
Hinson
have
used
of treadle
second,
for key theories
the different
might
experiments
ways many
in similar
1970;
However,
in functions
on two on the same
separated
the
for key pecking.
appeared pecking
on the operandum but
not for treadle
pressing. Because
of the theoretical
is important
to determine
differences
produced
responses.
The present
contrast
Experiment
rate
of reinforcement
stimuli
by McSweeney
EXPERIMENT This McSweeney
away
Experiment
reinforcers.
(e.g., multiple
Experiment
et al.
from
positive
to those
key-peck
between
1 but also
contrast
uses
a high
VI
15-s
VI
moves
the
the operanda
which
produce
replicates
(1986) to study
treadle
used
the components.
2 but also
3 essentially
it
the
study
similar
1 studies
a timeout
Experiment
3 replicates
discriminative
with
between They
do this.
results,
procedural
differences
procedures
Experiment
procedure
2 replicates
Experiment
used
functional using
of these
any of these
experiments
pressing.
a two-key
baseline
the
for key pecking
for treadle using
importance
whether
the
15-s).
procedure
pressing.
1 experiment
uses
(1982) to study
of component
duration
a procedure postive
except
that
similar
key-peck
to that
contrast
the present
used
by
as a function
procedure
uses
a
25
within-session
different
procedure, and a timeout
components,
between
operanda the
for the two
components.
Method
Subjects: their
Four
free-feeding
experience
pecking
number
model
were
3124A-300,
the
evenly
walls
and
The
hopper
hopper
was
The
key
the chamber.
centered
the
Experimental
located
All
subjects
outlined
in the order
(baseline)
key; pecking
multiple on the key
were
was
conducted a multiple
key.
The
VI l-m ext schedule. left
key,
extinction
illuminated
when
were
placed
VI were
are
houselight
noises
from by a
l-m
VI
part
The
for food
the
conditions
was
The l-m
were
were
key.
into
part
schedule.
obtained
component
was
divided
first
by pecking
(contrast)
right
keys
into
listed.
obtained
its component
key. 5-W,
programmed
session
The VI l-m on the
The
pecked
1.
component
second
center
directly
in Table they
the
successively.
for the other
the right
was
room.
the experimental
for one component
reinforcers
key
and
corner.
previously
were
in which
condition,
presented
Reinforcers
from was
with
center
A fan masked
events
had
conditions
which
key
in diameter,
right
in another
experimental
parts
The
illuminated
A 4 cm
sesion.
they
two
not
in the upper
Therefore,
In each
keys
6.5 cm
center
illuminated
light.
10.5 cm under
reinforcers.
conducted
green
model
All
used.
located The
left key was
with
4.5 by 5.5 cm.
microcomputer
were
floor.
It was
throughout
Procedure:
the
pigeon
no effect. was
located
illuminated
outside
from
keys
was
had
by a forceof
operated
side
experiment.
measured
houselight
two
them.
right
Grason-Stadler
had
in a Grason-Stadler,
chamber,
and were The
22.5 cm
on it had
A food
enclosed
85% of
All
as subjects.
three-key,
E6446C,
sound-attenuating
in this
responses
SYM
reinforcement.
between
light.
not used
was
for food
0.1 N.
spaced
white
The
keys
2.5 cm in diameter
side
at approximately
served
number
approximately
maintained
weights,
A standard,
Apparatus: station,
pigeons,
body
The
the
left
by
presented was
appropriate
available.
a
scheduled
26
TABLE
1
The conditions conducted they were conducted
in Experiment
1 in the order
in which
--______---__--___-_~~--~~~~~~__---____--__----__----___~~-_----_ Component duration Number of components per --______---______--_-~-------~_-~~~~----~__---~---~~~-~~--~~~~~~60 30 16 5 3
20 40 2 240 8
seconds seconds minutes seconds minutes
Component
duration
listed
in Table
during
which
in Table
been
some
Reinforcers mixed
responses
were
were
collected
presentation
Each
in each
becoming
extremely
of 5-s access lights
during
experimental
number
was
stable
responding
chosen
were
parts
Reinforcers
which
changed
of the
to the magazine
were
All
reinforcers in Fleshler
scheduled
held
which
and no
outlined
were
to
long.
extinguished
procedure
listed
conditions
reinforcement.
to the
a component
were
ofthetwo
from
Key
the
of components
experimental
according
of that
Sessions
the number
as
timeout,
separated
across
consisted
before
when
conditions,
a three-second
varied
grain.
(1962).
experimental
illuminated,
presented
recorded
scheduled
and Hoffman
ended
sessions
contained
across
conditions,
key was
The number
prevent
varied
In all
Sessions
lhad
session.
1.
neither
components.
part
over
but
not
for the
next
component.
conducted condition because when
daily, was
it is more
pigeons
five
to six times
conducted
peck
for
than keys
per
week. This
40 sessions.
sufficient on multiple
to produce schedules
1982).
(e.g., McSweeney,
Results Table
2 presents
each
component
calculated component The time all
the
of each
by dividing
of responding
for which
calculations. emitted
multiple
for which
the magazine The
results
during
the
in pecks
all subjects
schedule.
the number
by the time
of the rates
rates
for the meanof
foreachsubjectand
Response
of pecks that
was
presented
presented last
emitted
component
five
per minute
responding
was was
in Table sessions
rates during
on
were a
available. excluded 2 are
from
the
for which
means each
27
schedule
TABLE
was
conducted.
2
Response rates in responses per minute during each multiple schedule conducted in Experiment
each 1
component
of
Schedule ----_--30-second components l-minute components --~~~~-~~~~-~~--_---_----_----~~~--c~~_contrast contrast baseline baseline Subject compon _-____----_----~~~--~~-~-~~~--~~---~~-------_--___~~---~~~-~~~-1
left right
48.6 39.4
65.8 14.7
50.6 57.2
88.7 23.3
11
left right
70.6 145.2
63.5 85.0
105.1 141.1
156.2 85.6
12
left right
103.4 44.0
112.5 29.8
86.7 105.4
105.3 34.5
13
left right
139.3 190.2
166.3 84.2
170.9 116.8
163.0 38.5
mean
left right
90.5 104.7
102.1 53.4
103.3 105.1
128.3 45.5
Schedule ------_16-minute components 5-second components _-_--___-__-___--___ ------------------c Subject compon baseline contrast baseline .contrast _--____________________~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1
left right
54.0 46.8
77.6 5.9
89.1 12.6
100.4 4.6
11
left right
103.6 116.1
100.6 24.2
155.0 92.3
191.8 36.9
12
left right
108.5 107.5
65.8 23.1
131.2 130.8
160.8 35.8
13
left right
135.2 91.2
113.4 23.5
165.9 47.1
181.2 23.0
mean
left right
100.3 90.4
89.4 19.2
135.3 70.7
158.6 25-l--
28 --_--____-__________~~--~~~~~-~~~~_--~~~-----------~_---~~---~~~_ TABLE 2 (contiued) Schedule ---_---3-minute component --_-----_--__-_-_Subject compon baseline contrast ___---_----~~---~~~---~~~-~~~~~-~~~ 1
left right
52.1 58.4
83.4 11.7
11
left right
106.5 157.3
122.9 57.4
12
left right
117.0 129.0
88.1 53.9
13
left right
148.2 137.8
179.7 52.8
mean
left right
106.0 120.6
118.5 44.0
.-- ~~-----___----____-___--------~~The rates baseline which
rates
differences over
Each
probably
of all
of responding ext,
reported
on the
fall
negative
induction,
defined
the
But,
left
of contrast
key
size
of responding
during
the
were
based
subject
was
on the left key
schedule,
the horizontal
which
fall
from
contrast,
the
multiple
on the data
a constant
component
a trends
the
in the
with
at 0 represent line rate
decreases
represent of in the
rate
component.
the variable
of contrast
line
below
as a decrease
in the other
1 reveals
measurement
al., 1986).
The
VI l-m VI l-m
above
Points
Figure
for an individual
rate
per
in minutes.
results
Calculations
contrast.
during
the
duration
1982).
2.
which
of reinforcement
the
two-
Davison,
in responses
of component
subjects.
positive
responding
of contrast
operandum
when
and
the
These
for one appear
(e.g., Charman
multiple
schedule.
in Table
Points
usually
of the
provided
components.
a preference
size
represents
the baseline,
VI l-m
the
components
2, or in the tables
schedules
in the two
used
by subtracting
two
the baseline
as a function
or for the mean
rate
are
the
in Table
differences
1 presents
plotted
determined
equal
represent
Such
set of axes
during
though
procedures
Figure minute
during
not
of reinforcement
the other.
operanda
were
even
follow,
same
of responding
schedules
results
which
(e.g., McSweeney, test
applied
to the
are typical
1982; points
of
McSweeney in Figure
et 1
29
revealed
a significant
contrast
and
test
matched
for
smaller
the
30-s,
relation
duration
pairs
significantly shorter,
linear
component
also
showed
for the
ones
between
the size of
(F(1,12) = 5.26, that
longest,
contrast 16-m,
ptO.05).
A t-
was
component
than
for
(t(3) = 2.84, p
II
a fk 5
15 COMPONENT
DURATION
(MINUTES)
contrast as a function of component duration in Fig. 1. Positive by Experiment 1. The size of contrast has been measured subtracting the rate of responding during the left component of the multiple VI l-m VI l-m baseline from the rate of responding in the left component during the multiple VI l-m extinction The size of contrast is reported in responses per schedule. duration in minutes. minute, component
30
EXPERIMENT
2
Experiment higher
rates
study
positive
2 replicates
Experiment
of reinforcement contrast
for
used
1 except
by McSweeney
treadle
that
it uses
et al.
the
(1986) to
pressing.
Method Subiects: They
here. feeding
The
were
body
used
in Experiment
at approximately
and Procedure:
in Experiment
Experiment
subjects
1 were
used
05% of their
free-
was
one
weights.
Apparatus used
same
maintained
1.
1 except
The
that
15-sVI15-sand the
The
apparatus
procedure
was
the baseline
also
schedule
contrastschedulewas
the same
identical was
to
a multiple
a multiple
VI
VI 15-s
ext.
Results Table component meanof
3 presents of each
multiple
all subjects.
Table
2.
TABLE
3
the
rates
of responding
schedule
All data
for all
emitted subjects
during and
for
werecalculatedastheywere
Response rates in responses per minute during each multiple schedule conducted in Experiment
each 2
each the for
component
of
____c_______________~~~~~~~~--------------_--_------_----____--_-
Schedule -____--_ 30-second components l-minute components ------------------_----------------___ contrast baseline contrast baseline Subject compon _______________--_----------~---------_________--_-------_-__-___ 69.0
2.8
55.5 52.1
74.3 6.1
93.2 208.8
85.1 51.1
76.9 148.5
72.2 35.6
left right
125.3 106.9
123.7 29.1
119.4 109.4
163.7 30.0
13
left right
161.7 156.7
189.1 35.9
121.3 101.5
148.9 23.9
mean
left right
107.6 132.0
116.7 29.7
93.3 102.9
114.8 23.9
left right
50.1 55.7
11
left right
12
1
31
_____----------------------------~--------~-_~~-~--~~~~--------~ 3 (continued)
TABLE
Schedule _-------
16-minute components ---_-----_---__---_contrast Subject compon baseline __----------_-----~~~~~~~~~~~~~~~~~~~~~
5-second components ---_---_------~~~~~ baseline contrast ----_--------------~~~~
1
left right
46.7 40.7
46.8 1.8
111.6 56.3
149.2 8.6
11
left right
71.8 86.4
67.1 3.5
66.1 120.4
92.0 51.6
12
left right
70.6 115.8
70.5 0.3
142.6 134.0
177.6 27.2
13
left right
34.9 57.7
37.5 6.8
154.8 120.0
129.8 15.9
mean
left right
56.0 75.2
55.5 3.1
118.8 107.7
137.2 25.8
_____~________---_-_------------------------------~_~~~~~~~~~~~~ Schedule -------3-minute components ----------_--~~---~ Subject compon baseline contrast --------------_------~~~~--------1
left right
83.6 63.6
96.3 6.3
11
left right
59.4 133.9
80.7 30.3
12
left right
119.8 104.1
168.9 30.8
13
left right
124.4 98.2
134.2 13.2
mean
left 96.8 120.0 right 100.0 20.2 ____________________~~~~~--~~--~-_--______-__________--------~~~ Figure
2 presents
component
duration
subjects.
Points
A trends reveal
t-tests
and
test
less
size
showed
(ptO.05) the
for 3-m
as a function
for the mean
as they
relationship
duration
pairs
and
to the points
linear
(t(3) = 2.41) and
of contrast
subject
calculated
applied
component
for matched
significantly 30-s
were
a significant
contrast
the
for each
were
in Figure between
(F(1,12) = that the
the
16-m
for
1.39,
size
Figure 2 did
the
size
p>O.O5).
of contrast
components
Of
of all
than
(t(3) = 2.54) components.
1. not of But, was for the
32
13
I
8 w
N
5 2o
5
0
Is
15
MEAN
15
5
COMPONENT
DURATION
(MINUTES) Fig. 2. The size of positive contrast as a function of component duration in Experiment 2. The figure has been plotted as in Figure 1. EXPERIMENT
3
Experiment discriminative
3 was
identical
stimuli
multiple
schedule
operanda
which
which
appeared
produced
to Experiment
signalled
2 except
the components
on the center
key
instead
that
the
of the of on the
reinforcers.
Method Subjects:
Four
pigeons
maintained
at approximately
85% of
33
their had
free-feeding
pecked
placed
keys
directly
into
Apparatus one
used
and
to that
discriminative and
right
key
It was
produced
white
The sixteen-minute
relatively
small.
reinforcers
multiple
reinforcers
alleviated
Pecks
center left
red light left and
during
also
when
right
both was
also
of them
failed
delivered
during
long
requires
the
that
problem
the
from
longest
Failure
all
on this very
to collect
all
The
of reinforcement
the baseline
components
to some
to collect
impossible.
rate
by a
were
components
of contrast
on the
were
replaced
some
component
pecks
components.
Therefore,
schedule.
was
produced
keys
experiment
15-s
on the
key key
in this
in one
this
The
to the
the
key.
on the
duration
the measurement
Shortening
phase.
light
component
of contrast
constant
with The
was
The subjects
15-s VI
made
definition held
VI
subjects were
identical
that
center
pecks
illuminated
reinforcers.
duration.
procedure
on the
when
was
2 except
reinforcers.
light
with
and therefore
apparatus The
2.
appeared
produced
ten-minute
rich
The
in Experiment
green
All
as subjects.
experiment.
1 and
used
illuminated
of the
served
experiment,
Procedure:
keys
with
reinforcers.
both
the
stimuli
illuminated
right
weights
in Experiments
identical
left
body
in a previous
be
to the contrast
to ten minutes
extent.
Results Table component mean Table
4 presents of each
of all
subjects.
2 with
subjects
did
the
following always
this
rich
in the data 4 are not When
were
analysis. always
the
mean
sessions,
five
sessions
then
the reported
in which
these
As discussed,
the
this
reinforcers happened
the data
of the
from
the
reinforcement. conditions Subject
83005
last
data
all reinforcers
suchsessionsdidnotoccur,
full
and were
all of their
in which
each
as they
calculated
Therefore,
during subject
last
for the for
from
could
presented
five
not
be
in
sessions.
a subjectdidnotobtainallreinforcersinthelast
five
sessions,
for each
exceptions.
collect
Sessions
schedule.
of responding
schedule
Rates
not
used Table
the rates
multiple
five,
Footnotes
are the mean were
the data
are the mean
in which
the subject
in the
table
of the
obtained.
indicate
last
When
five
of two did
obtain
whenever
occurred. also
failed
to obtain
all
reinforcers
in any
34
session indicate all at
for two when
subjects all
TABLE
component
that are
durations.
occurred.
the
mean
The
Blanks
data
for the three
in the table
reported
for the mean
subjects
which
durations.
4
Response rates in responses per minute during each multiple schedule conducted in Experiment
each 3
component
~~~~~~~----------~-~~~~~-----_--~~~~~~~~--_----~__-----_~------Schedule ____-___ l-minute components 30-second components ------_-_--_-_----_----_--------~----baseline contrast contrast Subject compon baseline ~~~~~~~---_-----_-------_______---------___________------_____ left right
48.8 60.3
77.0 6.9
51.9 77.6
77.9
83004
left right
29.0 21.2
37.4 6.6
43.11 34.1
48.5l 9.1
83005
left right
19.0 27.3
39.3 4.6
31.31 49.8
49.11 4.5
2457
left right
51.2 60.8
60.0 10.2
34.5 24.7
58.3 23.8
mean of 3
left right
43.0 47.4
58.4 7.9
43.2 45.5
61.6 17.2
83002
of
responded
18.8
____---_____-__-_------_______----------___------~~~~~~~~-~~~~~~Schedule ------__ lo-minute components 5-second components ------_------------_-----------------Subject compon baseline contrast baseline contrast ~-~~~~~_--_----~-~~~~~~~_---~~--~~~~~-~~~-_---~~-~~~~~~~~~~~~~~ 83002
left right
50.9 58.3
54.7 0.3
61.5 70.8
91.7 9.4
83004
left right
22.6 19.1
4.5 0.4
26.42 24.6
32.72 7.1
83005
left right
31.8 35.3
4.4 0.1
-_-
-__
2457
left right
33.8 45.8
36.8 1.9
13.1 12.1
28.7 0.9
mean left 35.8 32.0 33.7 51.0 of 3 right 41.1 0.9 35.8 5.8 -------__________-__~~~~~~~~~~~~~~~~-~-~______-----------___----_
of
35
--_-________________~~~~~~~~~~--_--~~~~~--________--_---_-_--___~ TABLE
4 (continued)
Schedule -----___ 3-minute components ~~~~----_-~-----~-Subject compon baseline contrast ______-------------_-------------83002
left right
59.0 90.3
84.2 7.2
83004
left right
25.32 32.7
2J.a2 3.1
83005
left right
---
---
83006
left right
25.42 35.7
24.52 3.1
mean left 36.6 45.5 of 3 right 52.9 4.5 _____---~~----------~----~--------_-~-~--~_--------~~~~~~~~~~~~~~~ iBased on the mean of five earlier sessions. Based on the mean of only two sessions. Figure
3 presents
of component
the
size
subjects.
The
size
of contrast
Figure
The
data
presented
1.
of contrast
for each subject
duration
was
plotted
calculated
in Table
as a function
and for the mean
4 were
of all
as it was used
for
in these
calculations. A trends significant component pairs
test
linear
applied
to the points
relation
duration
between
(F(1,8) =
alsoshowedthatthe
components
was
contrast
for
components.
GENERAL
significantly
the
5-s
than
for
present
the
60-s
T-tests
(ptO.05) than and
30-s
significantly components
to study
procedure
offers
approximately
contrast.
two
The
the
use
As argued
benefits.
one-third
procedure.
support
the
for
matched
size
of
(t(2) = 28.43) smaller
(t(2) =
for the
3.75).
of the within-session earlier,
use
It can be conducted
of the time present
a
and
Procedure results
procedure
sessions
less
also
p
3 revealed
of contrast
DISCUSSION
The Within-Session The
was
size
of contrastforthelO-m
(t(2) = 4.91)
Contrast
components
26.89,
size
in Figure
the
required
experiments
by the required
of this in across600
3-m
36
sessions would per
per
have
subject
required
1800
Second,
subject.
fluctuations periods
to conduct.
rates
required
procedure
or approximately
six years,
sessions,
it reduces
in response
of time
An across-sessions
which
by the
30
the
confounding occur
effect
across
across-sessions
of
the
the long
procedure.
83002
Gi 5 z
‘O
f
t
;; Y
wo 20 n
ro
E -2
0
I
20
a
0
L 5
W N
v,
20
I
2457
83005
0
I
15
MEAN
k ,
J
, 5
* COMPO
ri5E NT DURATION INUTES)
(M
Fig. 3. The size of positive contrast as a function of ccmponent duration in Experiment 3. The figure has been plotted as in Figure 1. In spite would
of these
not be used
the across-sessions
unless
advantages, it produced
procedure.
the
within-session
results
The present
similar
procedure to those
experiments
stow
of that
37
it does. To begin
the within-session
with,
Forty-four
contrast. subjects
of the
in Figures
contrast
Rachlin,
case
produced
in which
component key
during
present
pecking
reported
are
longer
similarity
the
two
cases,
the
within-session
the
functions
in the
30-s
components
30-s
than
this
Finally, across
the
results
Only
than
further They
differences
between
experiments
presented
different
operandum. changes
those
have the
operanda.
reported test
some
of
might
for
larger
for the for the
experiments,
1 to 3 are
these
(1982) used
more
variable
(1982). differences
procedural the
of the multiple only
be eliminated
procedure.
longer
significant.
by other
also
in minute
the
5-s than
it was
why
per
but not in
For example,
the components
within-sessions
occurred
by McSweeney
can determine
McSweeney
per minute
present
in Figures
produced
two
differences.
statistically
reported
of contrast
for the
for the
but
research
can be size
20 responses
in the
studies.
further
40 responses
larger
not
been
The differences
in the
similar
(19821,
reported
the
experiments,
was
was
research
may
was
was
a strong
procedure
induction
components
difference
subjects
occurred.
5-s
(e.g.,
contrast
promising,
exceeded
Negative
in McSweeney
for the
to
reported
contrast
provides
produce than
it rarely
Contrast
(1982).
procedure
relating
in all of the present
McSweeney
although
did
larger
study.
are
basically
the procedures
present
components
results
were
(19821, but
to those
largest
similarity
the
McSweeney
present
sometimes
one
procedures.
before
was
only
extinction
size of contrast
and smaller
Although
procedures,
report
the
in trend
be done
Contrast
tables
the
the
duration
(e.g.,
component.
Although
to component
the
phase
the within-session
across-session
This
of the
for
between
contrast
during
similar
components
components.
accepted.
two
The
relating
In both
for short
the
that
greater
also
using
1982).
for
in the
the other
functions
duration
McSweeney,
discriminate
was
good
One prerequisite
must
discrimination.
individual
contrast.
produces
schedule
2 to 4 show
responding
than
The
must
Tables
1973).
component
for
multiple
produces for
positive
also
the components.
the subjects
of the
procedure
procedure
between
is that
components
plotted
1 to 3 represent
The within-session discrimination
procedure
58 points
present schedule
on
one by minor
For example,
only
time
38
signalled
the
the present transition
change
from
the baseline Adding
experiments. might
decrease
to the contrast
another
variability
stimulus
and
phase
to signal
increase
in this
the size
of
contrast. The
present
results
determine
the
source
procedure
has
some
and it produces the
of the Function
positive
key-peck
ways
from
striking between the
McSweeney similar
each
may
operanda
procedure
negative
treadle-press
peck
Staddon,
not
appear
The
of
in many but
with
which
major
size
changed
changes
changes
in the
the components
a one-operandum
when
A twoproduces
procedure Positive
1978).
the discriminative but not
in
of positive
et al., 1986).
between
operandum
the
shorter by
Herrnstein,
the absolute
contrast
patterns
the decrease
As predicted
change
eliminate
when
changed
variables
to produce
but
size
is a robust
is particularly
(e.g., McSweeney,
of similar that
from
1982;
by
studies.
results
of five
a timeout
larger
experiments,
(McSweeney
contrast,
does
emergence
contrast
procedure,
produced
stimuli they
key
are
are not
1975).
suggests
is robust.
which
with
may
(e.g., Schwartz, The
shown
the
contrast
in both
of these
a factor
on the response
studies
emerged
of reinforcement
a timeout
that
of contrast.
changeby
contrast
to those
components
of the procedural
and occurrence
rate
located
suggest
to measure
has been
baseline
peck
to
within-session
the across-sessions
(1982) to the present
contrast
without
used
similarity
because
The
results
for short
results
the studies
size
also
contrast
The overall
over
similar
results
The procedure
basically
it is worthwhile
procedure.
The Form ---The
present
that
differences.
advantages
basically
across-sessions
finding.
suggest
of these
to the
several 19701,
of results
in the
size
longer
theories it is not
size of contrast,
across
the
of positive
component
key-
durations
(e.g., Ettinger altered but
and
by variables
which
do not
altogether.
The Generality
of Contrast The present results also
different treadle contrast
function pressing.
of component All
to component
suggest
that
duration
contrast
changes
for key pecking
and
of the
functions
relating
key-peck
duration
in this
experiment,
and
in
as a
39
(19821, differ
McSweeney contrast
(McSweeney,
negative
induction but
pressing, components
but
pressing,
or small
large
were
The
the
largest
contrast
for treadle
component
durations
for key
between
the
attributed measure
functions
between
Claiming
that
two
of the variables and
controlled
as close
procedure that
this
measured
procedures
positive
contrast
or negative hand,
the
pecking obtained
treadle the
pressing and
differences subjects, the
when
forms
cannot
in the
different
differences
they
or acrosssizeable
smaller
contrast
On the
other
(1986) for key though
Therefore, for key
functions
both
pecking
to the difference
were
the and
between
procedures.
all of these
studies that
the
of day at which laboratory,
even
functions
be attributed
between
et al.
procedures.
of the
and
long.
it is
of whether
reported
short were
by McSweeney
it is unlikely
occurred
climate
were
were
the
A
an across-
of the
regardless
components
within-sessions
the time
contrast.
However,
forms
All
come
in
3, but
experiments)
1982).
components when
across-sessions
because
laboratory,
similar
all
occurred
contrast
produced
the
(the present
pressing
across-
difference
that
studies
difference
in Experiment
earlier,
out.
possible.
for treadle-pressing.
and treadle
Second, same
used
used
to
identified
the present
is currently
used
require
been
can be
differences
ruled
would have
procedural
(McSweeney,
in the
be completely
treadle-press
reported
differences
studies
to measure
was
functions
using
the
procedural
used
basically
induction
minor
differences pressing
key-peck
by within-
sessions
the
to measure
As argued were
that
However, as
procedural
functions.
for
at one-minute
used
was
occurred
at 5 or 30-second
and treadle
behavior
identifiable
that
but
are identical
influence
procedure
for key pecking were
one
3 and
within-session
in the
never
studies.
Experiment
unlikely
can
identity
only
occurred
between
that
across
the procedure
sessions
differences
to achieving
First, between
pecking
procedures which
for treadle
induction
it unlikely
for key
studies
occurred
also
short,
When
for key pecking.
pressing,
The possibility
occurred
were
for treadle
pecking.
make
to procedural
them.
occurred
or negative
durations
results
components
contrast
contrast
for treadle-press
When
occurred
sizeable
component
The present
function
contrast
contrast
long,
little
key pecking.
from
et al., 1986).
were
other,
studies. the
The
studies
the physical
conducted
less
in the
obvious,
handling were
of the
conducted,
arrangement
of the
40
experimental were
held
enclosures,
as similar
Third,
the
to
function
relating
duration
is not
the form
of the functionis
form
The
present
two
that
responses.
might that these
provide future
key-peck
by major
basic
contrast
procedural
robust,
differences
it
etc.
the studies.
3 showthatthe
experiments
ruled
for the different
pressing.
evidence
across
feedings,
of the
to component Given
changes.
that
unlikelythatother
seems
would
form
radically
change
the
function.
explanations treadle
positive changed
procedural
of this
of post-session
as possible
Experiments1
unidentified
time
Therefore,
different However, a single
experiments
functions they
factors
might
the
possible
other
pecking
strongest
positive
explanation provide
possible
for key
provide
produce
it remains unified
out several
that
contrast future
for these
and current for the theories
functions
explanations
or
for
data.
REFERENCES Charman, L. and Davison, M., 1982. On the effects of component durations and component reinforcement rates in multiple schedules. J. Exp. Anal. Behav., 37: 417-439. The effects of different Davison, M. and Ferguson, A. 1978. component response requirements in multiple and concurrent J. Exp. Anal. Behav., 29: 283-295. schedules. Ettinger, R. H. and Staddon, J. E. R., 1982. Behavioral component duration and multiple schedule competition, contrast. Behav. Anal. Letters, 2: 31-38. R. H. and Staddon, J. E. R., 1983. The operant Ettinger, A static analysis. Behavioral regulation of feeding: Neuroscience, 97: 639-653. Fleshler, M. and Hoffman, H. S., 1962. A progression for generating variable-interval schedules. J. Exp. Anal. Behav., 5: 529-530. Herrnstein, R. J., 1970. On the law of effect. J. Exp. Anal. Behav., 13: 243-266. Heyman, G. M., 1983. A parametric evaluation of the hedonic and pimozide and amphetamine. J. motoric effects of drugs: Exp. Anal. Behav., 40: 113-122. Behavioral Hinson, J. M. and Staddon, J. E. R., 1978. competition: A mechanism for schedule interactions. Science, 202: 432-434. Kodera, T. L. and Rilling, M., 1976. Procedural antecedents of A re-examinination of errorless behavioral contrast: learning. J. Exp. Anal. Behav., 25: 27-42. Negative behavioral contrast on multiple McSweeney, F. K., 1978. treadle-press schedules. J. Exp. Anal. Behav., 29: 463-473. Positive and negative contrast as a McSweeney, F. K., 1982. function of component duration for key pecking and treadle pressing. J. Exp. Anal. Behav., 37: 281-293. McSweeney, F. K., Dougan, J. D, Higa, J. and Farmer, V. A., 1986. Behavioral contrast as a function of comoonent duration and
41
baseline rate of reinforcement. Anim. Learn. Behav., 14: 173-183. Rachlin, H., 1973. Contrast and matching. Psychol. Rev., 80: 217-234. Schwartz, B., 1975. Discriminative stimulus location as a determinant of positive and negative behavioral contrast in J. Exp. Anal. Behav., 23: 167-176. the pigeon. K. L., 1971. Matching to relative Shimp, C. P. and Wheatley, reinforcement frequency in multiple schedules with a short J. Exp. Anal. Behav., 15: 205-210. component duration. Interactions in multiple schedules: The Spealman, R. D., 1976. role of the stimulus-reinforcer contingency. J. Exp. Anal. Behav., 26: 79-93. L. R., 1974. Spealman, R. D. and Gollub, Behavioral interactions in multiple variable-interval schedules. J. Exp. Anal. Behav., 22: 471-481. Behavioral competition, contrast and Staddon, J. E. R., 1982. matching. In: M. L. Commons, R. J. HerrnSteiT. and H. Analyses of Behavior, Rachlin (Editors), Quantitative Matching and Maximizing Accounts (Vol. 2). Ballinger Publishing Company, Cambridge, Massachusetts, pp. 243-262. Williams, B. A., 1983. Another look at contrast in multiple J. Exp. Anal. Behav., 39: 345-384. schedules.
CONTRAST
WITHIN-SESSION
FRANCES
AS A FUNCTION
K. MCSWEENEY
Accepted
OF COMPONENT
DURATION
USING
A
PROCEDURE
AND
CAM L. MELVILLE
Department of Psychology, WA 99164-4830 , U.S.A.
(
21
10 November
1987
Washington
State
University,
Pullman,
)
ABSTRACT
McSweeney, F. K. and Melville, C. function of component duration using Behav. Process., 16 : 21-41
L. Positive contrast as a a within-session procedure.
Pigeons pecked keys for food reinforcers delivered by multiple schedules. Three experiments examined changes in the size of positive contrast with changes in component duration. They used a within-session procedure which measured rates of responding during the baseline and contrast phases within the same session. Contrast was largest for shorter components, a result similar to that found using traditional procedures. Therefore, the present results provide preliminary support for the use of the withinsession procedure to study contrast. They also showed that the inverse relation between key-peck contrast and component duration is a robust finding which survives major changes in procedure. Finally, they suggest that different functions relate the size of contrast to component duration for key pecking and treadle presssing. INTRODUCTION The yielded
study few
for this whether
failure
may
contrast
circumstances. failure
of multiple-schedule
generally
Such
to observe
have used
the more contrast
studies
studies
few
under
approach
to independent would
little
examine
only
particular because
to many
information. of examining
variables.
constrain
has
of the reason
conclusions
can be attributed
productive
contrast Part
frequently
to occur
yield
it provides
studies
behavioral
conclusions.
or fails
contrast
Therefore,
such parametric
be that
occurs
factors.
relating
accepted
different Few
studies
the functions
Information
theorizing
the
from
to a greater
degree.
---------Preparation of this manuscript was partially supported by NIMH Thedata werepresented at the1987 meeting of grantMH42466. the Psychonomic Society. The authors wish to thank Dr. John Hinson for his comments on an earlier version of this paper.
0376-6357/88/$03.50 0 1988Elsevier Science Publishers
B.V. (Biomedical
Division)
22
The time
required
to measure
to the failure
to conduct
is an increase
in responding
multiple
schedule
reinforcement decrease
in the other
contrast,
contrast
schedule
in responding
represented
contrast
of contrast,
and
fluctuations
study.
in response
rates
example,
subjects
may double
schedule
to its recovery
Spealman
and Gollub,
To encourage experiments problems. measures
a new
The within-session
schedules
1983;
Heyman,
equipment to
studies
measure
used
1983).
over
for a five-
time.
rates
by
For
from
one
et al., 1986;
phases
has been
similar
used
procedures
reduces
these
procedure
which
single
sessions.
within because
other
to study
(e.g., Ettinger
The replacement control
the present
which
use a within-session
of reinforcement
by computer
measurement
is confounded
of contrast,
of contrast
and contrast
approach
have successfully other
to 50
schedule.
for each
(e. g., McSweeney,
The experiments the baseline
for each
occur
First,
Thirty
1974).
parametric
test
drawbacks.
response
over
1986).
are required
which
their
that phase
in responding
the measure
a
of
the conditions
to conduct.
are required
Second,
Then
to the contrast
has two
subject
provides
to ensure
and Farmer,
450 to 750 sessions
parametric
and varies
fluctuations
sessions
90 to 150 sessions
which
components.
is recovered
Higa,
To
the conditions
the baseline
for each
is a
with
is conducted
holds
procedure
many
are conducted
Therefore,
point
requires
of a
of
contrast
component
in both
constant
than
contrast
of reinforcement.
which
Dougan,
This across-sessions
Negative
schedule
from
rather
(e.g., McSweeney,
sessions
of the conditions
the baseline
the changes
the procedure
Positive
of reinforcement
Finally,
contributed
component
conditions
in one component
in the other.
have
studies.
a constant
is conducted
may
a constant
component.
a baseline
conditions
reinforcement
during
during
in the other
the same
time
parametric
a worsening
in responding
improvements measure
with
contrast
authors
responding
and
on
Staddon,
of electromechanical
also makes
such procedures
easier
conduct. The present
experiments
the within-session whether produces component Positive
this
procedure.
procedure
the same duration contrast
offer
The experiments
produces
function
a relatively
contrast,
relating
for short
test of
determine
not only
but also whether
positive
as the traditional is largest
strong
contrast
across-sessions components
when
it
to procedure. pigeons
23
peck
keys
for
food
Kodera
and Rilling,
1971;
Spealman,
replicate
this
reinforcers
1976). finding
The present using
Three
experiments
procedures.
Different
replications
of the results
They are also between
and component that contrast
changes
responses
will
more
of Herrnstein's
for shorter
components
should
produce
observed been
more
for short
observed
Therefore, pecking,
peck
experiments
theories
might
larger
components. also
be
The
larger
m parameters
contrast
press
is not
treadles,
(McSweeney,
replicated
that
and terminal
and larger
it has
et al., 1986).
these
successfully
of
components
(1982) argued
should
pigeons
keys
theories when
for shorter
Although when
procedure.
Several
of interim
than for longer,
components
if future
different
of the relation
be largest
(1970) theory
pigeons
procedure.
using
and Staddon
contrast
to
to provide
duration. should
1982;
and Wheatley, attempt
the robustness
contrast.
when
these
are used
Ettinger
m parameter
experiments
within-session
in the competition produce
Shimp
of the within-sessions
to determine
For example,
and Staddon,
1982;
this function
procedures
predict
short.
dynamic
the
examine
contrast
contrast are
used
(Ettinger
McSweeney,
1976;
results
describe
for key
key-peck
contrast. Different procedural contrast
Positive
inversely
with
were
duration
the different
duration
contrast
duration
whether
functions
for key pecking
and negative
component
component
also used to determine
produced
to component
pressing.
with
procedures
differences
generally
for key pecking,
for treadle-pressing
relating
and treadle vary but directly
(McSweeney
et al.,
1986). The different warrant
investigation
theories
which
responses.
a simple
failure might
between
low
rates
occurs
functions
contrast
to an insensitivity
can easily
functions
1982),
for treadle explain
of component
why
pigeons
press changes
(e.g., Davison
of treadle or to poor pressing. contrast
duration
simple
For example,
in detecting
pressing
to
for all
out many
when
pressing
damaging
responses.
to difficulty
of treadle
(e.g., Staddon,
similarly
rules
between
to find positive
the components
explanations different
different
and treadle
they are particularly
contrast
be attributed
1978),
reinforcement
that
for key pecking
for the differences
treadles
in generally
because
argue
Finding
explanations
Ferguson,
functions
pressing
and to
discrimination
But, none should
of these
change
for the two
as
responses.
24
Therefore, differ This
the differences
in important
would
contrast
question
occurs
Herrnstein,
pressing
contrast several study
ways
from
and
operandum
for key
components Finally,
interval
contrast.
than
pressing,
interval that
reinforcers
for key
15
used
presented
pressing
but
timeout
but not
stimuli
The
to study
2-mext).
were
A three-second
pecking.
for treadle
schedule
in
of reinforcement
variable
for treadle
positive
to differ
2-mVI2-mandmultipleVI
operanda
treadle
in the
had
rate
extinction)
1983). and
to produce
key-peck
a higher
that
(e.g.,
Williams,
In order
15 second
the discriminative
produced
1978;
may
pressing.
for key pecking
to study
15-s
contrast
assume
responses
the experiment
used
VI
all
which
by differences
them.
of the multiple
response
for
that
and treadle
of contrast
Staddon,
used
(multipleV1
different
pecking
produced
those
suggest
functions
pressing,
multiple
The components
which
been
pressing
variable
key pecking
and
to measure
for treadle
(multiple
ways
Hinson
have
used
of treadle
second,
for key theories
the different
might
experiments
ways many
in similar
1970;
However,
in functions
on two on the same
separated
the
for key pecking.
appeared pecking
on the operandum but
not for treadle
pressing. Because
of the theoretical
is important
to determine
differences
produced
responses.
The present
contrast
Experiment
rate
of reinforcement
stimuli
by McSweeney
EXPERIMENT This McSweeney
away
Experiment
reinforcers.
(e.g., multiple
Experiment
et al.
from
positive
to those
key-peck
between
1 but also
contrast
uses
a high
VI
15-s
VI
moves
the
the operanda
which
produce
replicates
(1986) to study
treadle
used
the components.
2 but also
3 essentially
it
the
study
similar
1 studies
a timeout
Experiment
3 replicates
discriminative
with
between They
do this.
results,
procedural
differences
procedures
Experiment
procedure
2 replicates
Experiment
used
functional using
of these
any of these
experiments
pressing.
a two-key
baseline
the
for key pecking
for treadle using
importance
whether
the
15-s).
procedure
pressing.
1 experiment
uses
(1982) to study
of component
duration
a procedure postive
except
that
similar
key-peck
to that
contrast
the present
used
by
as a function
procedure
uses
a
25
within-session
different
procedure, and a timeout
components,
between
operanda the
for the two
components.
Method
Subjects: their
Four
free-feeding
experience
pecking
number
model
were
3124A-300,
the
evenly
walls
and
The
hopper
hopper
was
The
key
the chamber.
centered
the
Experimental
located
All
subjects
outlined
in the order
(baseline)
key; pecking
multiple on the key
were
was
conducted a multiple
key.
The
VI l-m ext schedule. left
key,
extinction
illuminated
when
were
placed
VI were
are
houselight
noises
from by a
l-m
VI
part
The
for food
the
conditions
was
The l-m
were
were
key.
into
part
schedule.
obtained
component
was
divided
first
by pecking
(contrast)
right
keys
into
listed.
obtained
its component
key. 5-W,
programmed
session
The VI l-m on the
The
pecked
1.
component
second
center
directly
in Table they
the
successively.
for the other
the right
was
room.
the experimental
for one component
reinforcers
key
and
corner.
previously
were
in which
condition,
presented
Reinforcers
from was
with
center
A fan masked
events
had
conditions
which
key
in diameter,
right
in another
experimental
parts
The
illuminated
A 4 cm
sesion.
they
two
not
in the upper
Therefore,
In each
keys
6.5 cm
center
illuminated
light.
10.5 cm under
reinforcers.
conducted
green
model
All
used.
located The
left key was
with
4.5 by 5.5 cm.
microcomputer
were
floor.
It was
throughout
Procedure:
the
pigeon
no effect. was
located
illuminated
outside
from
keys
was
had
by a forceof
operated
side
experiment.
measured
houselight
two
them.
right
Grason-Stadler
had
in a Grason-Stadler,
chamber,
and were The
22.5 cm
on it had
A food
enclosed
85% of
All
as subjects.
three-key,
E6446C,
sound-attenuating
in this
responses
SYM
reinforcement.
between
light.
not used
was
for food
0.1 N.
spaced
white
The
keys
2.5 cm in diameter
side
at approximately
served
number
approximately
maintained
weights,
A standard,
Apparatus: station,
pigeons,
body
The
the
left
by
presented was
appropriate
available.
a
scheduled
26
TABLE
1
The conditions conducted they were conducted
in Experiment
1 in the order
in which
--______---__--___-_~~--~~~~~~__---____--__----__----___~~-_----_ Component duration Number of components per --______---______--_-~-------~_-~~~~----~__---~---~~~-~~--~~~~~~60 30 16 5 3
20 40 2 240 8
seconds seconds minutes seconds minutes
Component
duration
listed
in Table
during
which
in Table
been
some
Reinforcers mixed
responses
were
were
collected
presentation
Each
in each
becoming
extremely
of 5-s access lights
during
experimental
number
was
stable
responding
chosen
were
parts
Reinforcers
which
changed
of the
to the magazine
were
All
reinforcers in Fleshler
scheduled
held
which
and no
outlined
were
to
long.
extinguished
procedure
listed
conditions
reinforcement.
to the
a component
were
ofthetwo
from
Key
the
of components
experimental
according
of that
Sessions
the number
as
timeout,
separated
across
consisted
before
when
conditions,
a three-second
varied
grain.
(1962).
experimental
illuminated,
presented
recorded
scheduled
and Hoffman
ended
sessions
contained
across
conditions,
key was
The number
prevent
varied
In all
Sessions
lhad
session.
1.
neither
components.
part
over
but
not
for the
next
component.
conducted condition because when
daily, was
it is more
pigeons
five
to six times
conducted
peck
for
than keys
per
week. This
40 sessions.
sufficient on multiple
to produce schedules
1982).
(e.g., McSweeney,
Results Table
2 presents
each
component
calculated component The time all
the
of each
by dividing
of responding
for which
calculations. emitted
multiple
for which
the magazine The
results
during
the
in pecks
all subjects
schedule.
the number
by the time
of the rates
rates
for the meanof
foreachsubjectand
Response
of pecks that
was
presented
presented last
emitted
component
five
per minute
responding
was was
in Table sessions
rates during
on
were a
available. excluded 2 are
from
the
for which
means each
27
schedule
TABLE
was
conducted.
2
Response rates in responses per minute during each multiple schedule conducted in Experiment
each 1
component
of
Schedule ----_--30-second components l-minute components --~~~~-~~~~-~~--_---_----_----~~~--c~~_contrast contrast baseline baseline Subject compon _-____----_----~~~--~~-~-~~~--~~---~~-------_--___~~---~~~-~~~-1
left right
48.6 39.4
65.8 14.7
50.6 57.2
88.7 23.3
11
left right
70.6 145.2
63.5 85.0
105.1 141.1
156.2 85.6
12
left right
103.4 44.0
112.5 29.8
86.7 105.4
105.3 34.5
13
left right
139.3 190.2
166.3 84.2
170.9 116.8
163.0 38.5
mean
left right
90.5 104.7
102.1 53.4
103.3 105.1
128.3 45.5
Schedule ------_16-minute components 5-second components _-_--___-__-___--___ ------------------c Subject compon baseline contrast baseline .contrast _--____________________~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1
left right
54.0 46.8
77.6 5.9
89.1 12.6
100.4 4.6
11
left right
103.6 116.1
100.6 24.2
155.0 92.3
191.8 36.9
12
left right
108.5 107.5
65.8 23.1
131.2 130.8
160.8 35.8
13
left right
135.2 91.2
113.4 23.5
165.9 47.1
181.2 23.0
mean
left right
100.3 90.4
89.4 19.2
135.3 70.7
158.6 25-l--
28 --_--____-__________~~--~~~~~-~~~~_--~~~-----------~_---~~---~~~_ TABLE 2 (contiued) Schedule ---_---3-minute component --_-----_--__-_-_Subject compon baseline contrast ___---_----~~---~~~---~~~-~~~~~-~~~ 1
left right
52.1 58.4
83.4 11.7
11
left right
106.5 157.3
122.9 57.4
12
left right
117.0 129.0
88.1 53.9
13
left right
148.2 137.8
179.7 52.8
mean
left right
106.0 120.6
118.5 44.0
.-- ~~-----___----____-___--------~~The rates baseline which
rates
differences over
Each
probably
of all
of responding ext,
reported
on the
fall
negative
induction,
defined
the
But,
left
of contrast
key
size
of responding
during
the
were
based
subject
was
on the left key
schedule,
the horizontal
which
fall
from
contrast,
the
multiple
on the data
a constant
component
a trends
the
in the
with
at 0 represent line rate
decreases
represent of in the
rate
component.
the variable
of contrast
line
below
as a decrease
in the other
1 reveals
measurement
al., 1986).
The
VI l-m VI l-m
above
Points
Figure
for an individual
rate
per
in minutes.
results
Calculations
contrast.
during
the
duration
1982).
2.
which
of reinforcement
the
two-
Davison,
in responses
of component
subjects.
positive
responding
of contrast
operandum
when
and
the
These
for one appear
(e.g., Charman
multiple
schedule.
in Table
Points
usually
of the
provided
components.
a preference
size
represents
the baseline,
VI l-m
the
components
2, or in the tables
schedules
in the two
used
by subtracting
two
the baseline
as a function
or for the mean
rate
are
the
in Table
differences
1 presents
plotted
determined
equal
represent
Such
set of axes
during
though
procedures
Figure minute
during
not
of reinforcement
the other.
operanda
were
even
follow,
same
of responding
schedules
results
which
(e.g., McSweeney, test
applied
to the
are typical
1982; points
of
McSweeney in Figure
et 1
29
revealed
a significant
contrast
and
test
matched
for
smaller
the
30-s,
relation
duration
pairs
significantly shorter,
linear
component
also
showed
for the
ones
between
the size of
(F(1,12) = 5.26, that
longest,
contrast 16-m,
ptO.05).
A t-
was
component
than
for
(t(3) = 2.84, p
II
a fk 5
15 COMPONENT
DURATION
(MINUTES)
contrast as a function of component duration in Fig. 1. Positive by Experiment 1. The size of contrast has been measured subtracting the rate of responding during the left component of the multiple VI l-m VI l-m baseline from the rate of responding in the left component during the multiple VI l-m extinction The size of contrast is reported in responses per schedule. duration in minutes. minute, component
30
EXPERIMENT
2
Experiment higher
rates
study
positive
2 replicates
Experiment
of reinforcement contrast
for
used
1 except
by McSweeney
treadle
that
it uses
et al.
the
(1986) to
pressing.
Method Subiects: They
here. feeding
The
were
body
used
in Experiment
at approximately
and Procedure:
in Experiment
Experiment
subjects
1 were
used
05% of their
free-
was
one
weights.
Apparatus used
same
maintained
1.
1 except
The
that
15-sVI15-sand the
The
apparatus
procedure
was
the baseline
also
schedule
contrastschedulewas
the same
identical was
to
a multiple
a multiple
VI
VI 15-s
ext.
Results Table component meanof
3 presents of each
multiple
all subjects.
Table
2.
TABLE
3
the
rates
of responding
schedule
All data
for all
emitted subjects
during and
for
werecalculatedastheywere
Response rates in responses per minute during each multiple schedule conducted in Experiment
each 2
each the for
component
of
____c_______________~~~~~~~~--------------_--_------_----____--_-
Schedule -____--_ 30-second components l-minute components ------------------_----------------___ contrast baseline contrast baseline Subject compon _______________--_----------~---------_________--_-------_-__-___ 69.0
2.8
55.5 52.1
74.3 6.1
93.2 208.8
85.1 51.1
76.9 148.5
72.2 35.6
left right
125.3 106.9
123.7 29.1
119.4 109.4
163.7 30.0
13
left right
161.7 156.7
189.1 35.9
121.3 101.5
148.9 23.9
mean
left right
107.6 132.0
116.7 29.7
93.3 102.9
114.8 23.9
left right
50.1 55.7
11
left right
12
1
31
_____----------------------------~--------~-_~~-~--~~~~--------~ 3 (continued)
TABLE
Schedule _-------
16-minute components ---_-----_---__---_contrast Subject compon baseline __----------_-----~~~~~~~~~~~~~~~~~~~~~
5-second components ---_---_------~~~~~ baseline contrast ----_--------------~~~~
1
left right
46.7 40.7
46.8 1.8
111.6 56.3
149.2 8.6
11
left right
71.8 86.4
67.1 3.5
66.1 120.4
92.0 51.6
12
left right
70.6 115.8
70.5 0.3
142.6 134.0
177.6 27.2
13
left right
34.9 57.7
37.5 6.8
154.8 120.0
129.8 15.9
mean
left right
56.0 75.2
55.5 3.1
118.8 107.7
137.2 25.8
_____~________---_-_------------------------------~_~~~~~~~~~~~~ Schedule -------3-minute components ----------_--~~---~ Subject compon baseline contrast --------------_------~~~~--------1
left right
83.6 63.6
96.3 6.3
11
left right
59.4 133.9
80.7 30.3
12
left right
119.8 104.1
168.9 30.8
13
left right
124.4 98.2
134.2 13.2
mean
left 96.8 120.0 right 100.0 20.2 ____________________~~~~~--~~--~-_--______-__________--------~~~ Figure
2 presents
component
duration
subjects.
Points
A trends reveal
t-tests
and
test
less
size
showed
(ptO.05) the
for 3-m
as a function
for the mean
as they
relationship
duration
pairs
and
to the points
linear
(t(3) = 2.41) and
of contrast
subject
calculated
applied
component
for matched
significantly 30-s
were
a significant
contrast
the
for each
were
in Figure between
(F(1,12) = that the
the
16-m
for
1.39,
size
Figure 2 did
the
size
p>O.O5).
of contrast
components
Of
of all
than
(t(3) = 2.54) components.
1. not of But, was for the
32
13
I
8 w
N
5 2o
5
0
Is
15
MEAN
15
5
COMPONENT
DURATION
(MINUTES) Fig. 2. The size of positive contrast as a function of component duration in Experiment 2. The figure has been plotted as in Figure 1. EXPERIMENT
3
Experiment discriminative
3 was
identical
stimuli
multiple
schedule
operanda
which
which
appeared
produced
to Experiment
signalled
2 except
the components
on the center
key
instead
that
the
of the of on the
reinforcers.
Method Subjects:
Four
pigeons
maintained
at approximately
85% of
33
their had
free-feeding
pecked
placed
keys
directly
into
Apparatus one
used
and
to that
discriminative and
right
key
It was
produced
white
The sixteen-minute
relatively
small.
reinforcers
multiple
reinforcers
alleviated
Pecks
center left
red light left and
during
also
when
right
both was
also
of them
failed
delivered
during
long
requires
the
that
problem
the
from
longest
Failure
all
on this very
to collect
all
The
of reinforcement
the baseline
components
to some
to collect
impossible.
rate
by a
were
components
of contrast
on the
were
replaced
some
component
pecks
components.
Therefore,
schedule.
was
produced
keys
experiment
15-s
on the
key key
in this
in one
this
The
to the
the
key.
on the
duration
the measurement
Shortening
phase.
light
component
of contrast
constant
with The
was
The subjects
15-s VI
made
definition held
VI
subjects were
identical
that
center
pecks
illuminated
reinforcers.
duration.
procedure
on the
when
was
2 except
reinforcers.
light
with
and therefore
apparatus The
2.
appeared
produced
ten-minute
rich
The
in Experiment
green
All
as subjects.
experiment.
1 and
used
illuminated
of the
served
experiment,
Procedure:
keys
with
reinforcers.
both
the
stimuli
illuminated
right
weights
in Experiments
identical
left
body
in a previous
be
to the contrast
to ten minutes
extent.
Results Table component mean Table
4 presents of each
of all
subjects.
2 with
subjects
did
the
following always
this
rich
in the data 4 are not When
were
analysis. always
the
mean
sessions,
five
sessions
then
the reported
in which
these
As discussed,
the
this
reinforcers happened
the data
of the
from
the
reinforcement. conditions Subject
83005
last
data
all reinforcers
suchsessionsdidnotoccur,
full
and were
all of their
in which
each
as they
calculated
Therefore,
during subject
last
for the for
from
could
presented
five
not
be
in
sessions.
a subjectdidnotobtainallreinforcersinthelast
five
sessions,
for each
exceptions.
collect
Sessions
schedule.
of responding
schedule
Rates
not
used Table
the rates
multiple
five,
Footnotes
are the mean were
the data
are the mean
in which
the subject
in the
table
of the
obtained.
indicate
last
When
five
of two did
obtain
whenever
occurred. also
failed
to obtain
all
reinforcers
in any
34
session indicate all at
for two when
subjects all
TABLE
component
that are
durations.
occurred.
the
mean
The
Blanks
data
for the three
in the table
reported
for the mean
subjects
which
durations.
4
Response rates in responses per minute during each multiple schedule conducted in Experiment
each 3
component
~~~~~~~----------~-~~~~~-----_--~~~~~~~~--_----~__-----_~------Schedule ____-___ l-minute components 30-second components ------_-_--_-_----_----_--------~----baseline contrast contrast Subject compon baseline ~~~~~~~---_-----_-------_______---------___________------_____ left right
48.8 60.3
77.0 6.9
51.9 77.6
77.9
83004
left right
29.0 21.2
37.4 6.6
43.11 34.1
48.5l 9.1
83005
left right
19.0 27.3
39.3 4.6
31.31 49.8
49.11 4.5
2457
left right
51.2 60.8
60.0 10.2
34.5 24.7
58.3 23.8
mean of 3
left right
43.0 47.4
58.4 7.9
43.2 45.5
61.6 17.2
83002
of
responded
18.8
____---_____-__-_------_______----------___------~~~~~~~~-~~~~~~Schedule ------__ lo-minute components 5-second components ------_------------_-----------------Subject compon baseline contrast baseline contrast ~-~~~~~_--_----~-~~~~~~~_---~~--~~~~~-~~~-_---~~-~~~~~~~~~~~~~~ 83002
left right
50.9 58.3
54.7 0.3
61.5 70.8
91.7 9.4
83004
left right
22.6 19.1
4.5 0.4
26.42 24.6
32.72 7.1
83005
left right
31.8 35.3
4.4 0.1
-_-
-__
2457
left right
33.8 45.8
36.8 1.9
13.1 12.1
28.7 0.9
mean left 35.8 32.0 33.7 51.0 of 3 right 41.1 0.9 35.8 5.8 -------__________-__~~~~~~~~~~~~~~~~-~-~______-----------___----_
of
35
--_-________________~~~~~~~~~~--_--~~~~~--________--_---_-_--___~ TABLE
4 (continued)
Schedule -----___ 3-minute components ~~~~----_-~-----~-Subject compon baseline contrast ______-------------_-------------83002
left right
59.0 90.3
84.2 7.2
83004
left right
25.32 32.7
2J.a2 3.1
83005
left right
---
---
83006
left right
25.42 35.7
24.52 3.1
mean left 36.6 45.5 of 3 right 52.9 4.5 _____---~~----------~----~--------_-~-~--~_--------~~~~~~~~~~~~~~~ iBased on the mean of five earlier sessions. Based on the mean of only two sessions. Figure
3 presents
of component
the
size
subjects.
The
size
of contrast
Figure
The
data
presented
1.
of contrast
for each subject
duration
was
plotted
calculated
in Table
as a function
and for the mean
4 were
of all
as it was used
for
in these
calculations. A trends significant component pairs
test
linear
applied
to the points
relation
duration
between
(F(1,8) =
alsoshowedthatthe
components
was
contrast
for
components.
GENERAL
significantly
the
5-s
than
for
present
the
60-s
T-tests
(ptO.05) than and
30-s
significantly components
to study
procedure
offers
approximately
contrast.
two
The
the
use
As argued
benefits.
one-third
procedure.
support
the
for
matched
size
of
(t(2) = 28.43) smaller
(t(2) =
for the
3.75).
of the within-session earlier,
use
It can be conducted
of the time present
a
and
Procedure results
procedure
sessions
less
also
p
3 revealed
of contrast
DISCUSSION
The Within-Session The
was
size
of contrastforthelO-m
(t(2) = 4.91)
Contrast
components
26.89,
size
in Figure
the
required
experiments
by the required
of this in across600
3-m
36
sessions would per
per
have
subject
required
1800
Second,
subject.
fluctuations periods
to conduct.
rates
required
procedure
or approximately
six years,
sessions,
it reduces
in response
of time
An across-sessions
which
by the
30
the
confounding occur
effect
across
across-sessions
of
the
the long
procedure.
83002
Gi 5 z
‘O
f
t
;; Y
wo 20 n
ro
E -2
0
I
20
a
0
L 5
W N
v,
20
I
2457
83005
0
I
15
MEAN
k ,
J
, 5
* COMPO
ri5E NT DURATION INUTES)
(M
Fig. 3. The size of positive contrast as a function of ccmponent duration in Experiment 3. The figure has been plotted as in Figure 1. In spite would
of these
not be used
the across-sessions
unless
advantages, it produced
procedure.
the
within-session
results
The present
similar
procedure to those
experiments
stow
of that
37
it does. To begin
the within-session
with,
Forty-four
contrast. subjects
of the
in Figures
contrast
Rachlin,
case
produced
in which
component key
during
present
pecking
reported
are
longer
similarity
the
two
cases,
the
within-session
the
functions
in the
30-s
components
30-s
than
this
Finally, across
the
results
Only
than
further They
differences
between
experiments
presented
different
operandum. changes
those
have the
operanda.
reported test
some
of
might
for
larger
for the for the
experiments,
1 to 3 are
these
(1982) used
more
variable
(1982). differences
procedural the
of the multiple only
be eliminated
procedure.
longer
significant.
by other
also
in minute
the
5-s than
it was
why
per
but not in
For example,
the components
within-sessions
occurred
by McSweeney
can determine
McSweeney
per minute
present
in Figures
produced
two
differences.
statistically
reported
of contrast
for the
for the
but
research
can be size
20 responses
in the
studies.
further
40 responses
larger
not
been
The differences
in the
similar
(19821,
reported
the
experiments,
was
was
research
may
was
was
a strong
procedure
induction
components
difference
subjects
occurred.
5-s
(e.g.,
contrast
promising,
exceeded
Negative
in McSweeney
for the
to
reported
contrast
provides
produce than
it rarely
Contrast
(1982).
procedure
relating
in all of the present
McSweeney
although
did
larger
study.
are
basically
the procedures
present
components
results
were
(19821, but
to those
largest
similarity
the
McSweeney
present
sometimes
one
procedures.
before
was
only
extinction
size of contrast
and smaller
Although
procedures,
report
the
in trend
be done
Contrast
tables
the
the
duration
(e.g.,
component.
Although
to component
the
phase
the within-session
across-session
This
of the
for
between
contrast
during
similar
components
components.
accepted.
two
The
relating
In both
for short
the
that
greater
also
using
1982).
for
in the
the other
functions
duration
McSweeney,
discriminate
was
good
One prerequisite
must
discrimination.
individual
contrast.
produces
schedule
2 to 4 show
responding
than
The
must
Tables
1973).
component
for
multiple
produces for
positive
also
the components.
the subjects
of the
procedure
procedure
between
is that
components
plotted
1 to 3 represent
The within-session discrimination
procedure
58 points
present schedule
on
one by minor
For example,
only
time
38
signalled
the
the present transition
change
from
the baseline Adding
experiments. might
decrease
to the contrast
another
variability
stimulus
and
phase
to signal
increase
in this
the size
of
contrast. The
present
results
determine
the
source
procedure
has
some
and it produces the
of the Function
positive
key-peck
ways
from
striking between the
McSweeney similar
each
may
operanda
procedure
negative
treadle-press
peck
Staddon,
not
appear
The
of
in many but
with
which
major
size
changed
changes
changes
in the
the components
a one-operandum
when
A twoproduces
procedure Positive
1978).
the discriminative but not
in
of positive
et al., 1986).
between
operandum
the
shorter by
Herrnstein,
the absolute
contrast
patterns
the decrease
As predicted
change
eliminate
when
changed
variables
to produce
but
size
is a robust
is particularly
(e.g., McSweeney,
of similar that
from
1982;
by
studies.
results
of five
a timeout
larger
experiments,
(McSweeney
contrast,
does
emergence
contrast
procedure,
produced
stimuli they
key
are
are not
1975).
suggests
is robust.
which
with
may
(e.g., Schwartz, The
shown
the
contrast
in both
of these
a factor
on the response
studies
emerged
of reinforcement
a timeout
that
of contrast.
changeby
contrast
to those
components
of the procedural
and occurrence
rate
located
suggest
to measure
has been
baseline
peck
to
within-session
the across-sessions
(1982) to the present
contrast
without
used
similarity
because
The
results
for short
results
the studies
size
also
contrast
The overall
over
similar
results
The procedure
basically
it is worthwhile
procedure.
The Form ---The
present
that
differences.
advantages
basically
across-sessions
finding.
suggest
of these
to the
several 19701,
of results
in the
size
longer
theories it is not
size of contrast,
across
the
of positive
component
key-
durations
(e.g., Ettinger altered but
and
by variables
which
do not
altogether.
The Generality
of Contrast The present results also
different treadle contrast
function pressing.
of component All
to component
suggest
that
duration
contrast
changes
for key pecking
and
of the
functions
relating
key-peck
duration
in this
experiment,
and
in
as a
39
(19821, differ
McSweeney contrast
(McSweeney,
negative
induction but
pressing, components
but
pressing,
or small
large
were
The
the
largest
contrast
for treadle
component
durations
for key
between
the
attributed measure
functions
between
Claiming
that
two
of the variables and
controlled
as close
procedure that
this
measured
procedures
positive
contrast
or negative hand,
the
pecking obtained
treadle the
pressing and
differences subjects, the
when
forms
cannot
in the
different
differences
they
or acrosssizeable
smaller
contrast
On the
other
(1986) for key though
Therefore, for key
functions
both
pecking
to the difference
were
the and
between
procedures.
all of these
studies that
the
of day at which laboratory,
even
functions
be attributed
between
et al.
procedures.
of the
and
long.
it is
of whether
reported
short were
by McSweeney
it is unlikely
occurred
climate
were
were
the
A
an across-
of the
regardless
components
within-sessions
the time
contrast.
However,
forms
All
come
in
3, but
experiments)
1982).
components when
across-sessions
because
laboratory,
similar
all
occurred
contrast
produced
the
(the present
pressing
across-
difference
that
studies
difference
in Experiment
earlier,
out.
possible.
for treadle-pressing.
and treadle
Second, same
used
used
to
identified
the present
is currently
used
require
been
can be
differences
ruled
would have
procedural
(McSweeney,
in the
be completely
treadle-press
reported
differences
studies
to measure
was
functions
using
the
procedural
used
basically
induction
minor
differences pressing
key-peck
by within-
sessions
the
to measure
As argued were
that
However, as
procedural
functions.
for
at one-minute
used
was
occurred
at 5 or 30-second
and treadle
behavior
identifiable
that
but
are identical
influence
procedure
for key pecking were
one
3 and
within-session
in the
never
studies.
Experiment
unlikely
can
identity
only
occurred
between
that
across
the procedure
sessions
differences
to achieving
First, between
pecking
procedures which
for treadle
induction
it unlikely
for key
studies
occurred
also
short,
When
for key pecking.
pressing,
The possibility
occurred
were
for treadle
pecking.
make
to procedural
them.
occurred
or negative
durations
results
components
contrast
contrast
for treadle-press
When
occurred
sizeable
component
The present
function
contrast
contrast
long,
little
key pecking.
from
et al., 1986).
were
other,
studies. the
The
studies
the physical
conducted
less
in the
obvious,
handling were
of the
conducted,
arrangement
of the
40
experimental were
held
enclosures,
as similar
Third,
the
to
function
relating
duration
is not
the form
of the functionis
form
The
present
two
that
responses.
might that these
provide future
key-peck
by major
basic
contrast
procedural
robust,
differences
it
etc.
the studies.
3 showthatthe
experiments
ruled
for the different
pressing.
evidence
across
feedings,
of the
to component Given
changes.
that
unlikelythatother
seems
would
form
radically
change
the
function.
explanations treadle
positive changed
procedural
of this
of post-session
as possible
Experiments1
unidentified
time
Therefore,
different However, a single
experiments
functions they
factors
might
the
possible
other
pecking
strongest
positive
explanation provide
possible
for key
provide
produce
it remains unified
out several
that
contrast future
for these
and current for the theories
functions
explanations
or
for
data.
REFERENCES Charman, L. and Davison, M., 1982. On the effects of component durations and component reinforcement rates in multiple schedules. J. Exp. Anal. Behav., 37: 417-439. The effects of different Davison, M. and Ferguson, A. 1978. component response requirements in multiple and concurrent J. Exp. Anal. Behav., 29: 283-295. schedules. Ettinger, R. H. and Staddon, J. E. R., 1982. Behavioral component duration and multiple schedule competition, contrast. Behav. Anal. Letters, 2: 31-38. R. H. and Staddon, J. E. R., 1983. The operant Ettinger, A static analysis. Behavioral regulation of feeding: Neuroscience, 97: 639-653. Fleshler, M. and Hoffman, H. S., 1962. A progression for generating variable-interval schedules. J. Exp. Anal. Behav., 5: 529-530. Herrnstein, R. J., 1970. On the law of effect. J. Exp. Anal. Behav., 13: 243-266. Heyman, G. M., 1983. A parametric evaluation of the hedonic and pimozide and amphetamine. J. motoric effects of drugs: Exp. Anal. Behav., 40: 113-122. Behavioral Hinson, J. M. and Staddon, J. E. R., 1978. competition: A mechanism for schedule interactions. Science, 202: 432-434. Kodera, T. L. and Rilling, M., 1976. Procedural antecedents of A re-examinination of errorless behavioral contrast: learning. J. Exp. Anal. Behav., 25: 27-42. Negative behavioral contrast on multiple McSweeney, F. K., 1978. treadle-press schedules. J. Exp. Anal. Behav., 29: 463-473. Positive and negative contrast as a McSweeney, F. K., 1982. function of component duration for key pecking and treadle pressing. J. Exp. Anal. Behav., 37: 281-293. McSweeney, F. K., Dougan, J. D, Higa, J. and Farmer, V. A., 1986. Behavioral contrast as a function of comoonent duration and
41
baseline rate of reinforcement. Anim. Learn. Behav., 14: 173-183. Rachlin, H., 1973. Contrast and matching. Psychol. Rev., 80: 217-234. Schwartz, B., 1975. Discriminative stimulus location as a determinant of positive and negative behavioral contrast in J. Exp. Anal. Behav., 23: 167-176. the pigeon. K. L., 1971. Matching to relative Shimp, C. P. and Wheatley, reinforcement frequency in multiple schedules with a short J. Exp. Anal. Behav., 15: 205-210. component duration. Interactions in multiple schedules: The Spealman, R. D., 1976. role of the stimulus-reinforcer contingency. J. Exp. Anal. Behav., 26: 79-93. L. R., 1974. Spealman, R. D. and Gollub, Behavioral interactions in multiple variable-interval schedules. J. Exp. Anal. Behav., 22: 471-481. Behavioral competition, contrast and Staddon, J. E. R., 1982. matching. In: M. L. Commons, R. J. HerrnSteiT. and H. Analyses of Behavior, Rachlin (Editors), Quantitative Matching and Maximizing Accounts (Vol. 2). Ballinger Publishing Company, Cambridge, Massachusetts, pp. 243-262. Williams, B. A., 1983. Another look at contrast in multiple J. Exp. Anal. Behav., 39: 345-384. schedules.