JOURNAL OF APPLIED BEHAVIOR ANALYSIS
2014, 47, 314–324
NUMBER
2 (SUMMER)
A LABORATORY COMPARISON OF TWO VARIATIONS OF
DIFFERENTIAL-REINFORCEMENT-OF-LOW-RATE PROCEDURES
JOSHUA JESSEL
AND JOHN
C. BORRERO
UNIVERSITY OF MARYLAND, BALTIMORE COUNTY
We compared 2 variations of differential-reinforcement-of-low-rate (DRL) procedures: spacedresponding DRL, in which a reinforcer was delivered contingent on each response if a specified
interval had passed since the last response, and full-session DRL, in which a reinforcer was presented
at the end of an interval if the response rate was below criterion within the specified interval. We used
a human-operant procedure and analyzed within-session responding to assess any similarities or
differences between procedures. Data revealed a positive contingency between responding and
reinforcement under the spaced-responding DRL schedule and a negative contingency under the
full-session DRL schedule. Furthermore, 60% of the participants discontinued responding by the
last full-session DRL session. Implications for the appropriate procedural and taxonomical usage of
both DRL schedules are discussed.
Key words: differential reinforcement of low rates, differential reinforcement of other behavior,
human operant, interresponse time, translational research
Differential-reinforcement-of-low-rate (DRL)
schedules are temporally based reinforcement
schedules that arrange the delivery of reinforcers
contingent on reduced rates of responding
(Ferster & Skinner, 1957). Two commonly
conducted iterations of DRL schedules differ
based on the unit of analysis; one programs a
contingent relation between each response,
whereas the other programs a contingent relation
between an overall count within a particular time
frame that can be intervals, sessions, or days
(Deitz & Repp, 1973).
In the first DRL variation, a response produces
a reinforcer only after a specified time of no
This study was conducted in partial fulfillment of the
requirements for the master’s degree of the first author at the
University of Maryland, Baltimore County. We thank Iser
G. DeLeon and SungWoo Kahng for their insightful
contributions to earlier versions of this manuscript. Joshua
Jessel is now affiliated with Western New England
University. Preparation of this article was supported by
Grant RO1HD049753 from the Eunice K. Shriver National
Institute of Child Health and Human Development
(NICHD). Its contents are solely the responsibility of the
authors and do not represent the official views of NICHD.
Address correspondence to J. C. Borrero, Department of
Psychology, University of Maryland, Baltimore County,
1000 Hilltop Road, Baltimore, Maryland 21250 (e-mail:
jborrero@umbc.edu).
doi: 10.1002/jaba.114
responding has elapsed (Catania, 2013). The
interval between each response is known as the
interresponse time (IRT), and this schedule is
sometimes called an IRT > t arrangement because
the time between responses (IRT) must be greater
than the programmed interval (t) to produce a
reinforcer. These temporally sensitive schedules
result in the subsequent reinforcement of both the
relative IRT and response, respectively, characterizing patterns of behavior that are referred to as
spaced responding. This schedule has been aptly
named a spaced-responding DRL in the applied
literature (Deitz, 1977).
The second DRL variation defines a contingency between responding and a reinforcer
following the elapse of a predetermined interval
as long as the overall rate within that interval is
below a predetermined criterion (Catania, 2013).
In some cases, the absence of responding is
considered an acceptable dimension, and the
characteristic features of maintaining low rates
during the DRL schedule are ignored (e.g., Bird,
Hepburn, Rhodes, & Moniz, 1991; Hagopian,
Kuhn, & Strother, 2009; Shaw & Simms, 2009;
Turner, Green, & Braunling-McMorrow, 1990).
For example, Deitz and Repp (1973) specifically
termed this variation a full-session DRL and arranged
criteria whereby reinforcers were presented, not
314
VARIATIONS OF DIFFERENTIAL REINFORCEMENT
contingent on a response, but at the end of the
interval if the target behavior (i.e., talking out
during class without permission) occurred at a
rate less than a specified criterion. This procedure
inherently permitted reinforcer delivery given no
responding. Because zero responding is sufficient to
produce reinforcer delivery, Deitz and Repp’s
example of a full-session DRL may have more in
common with a differential-reinforcement-of-otherbehavior (DRO) schedule in that (a) reinforcers are
presented on an interval basis, (b) there is only an
indirect relation between reinforcer deliveries and
IRTs, and (c) there is a possible negative contingency
between reinforcer presentation and the target
response. The combined schedule could instead
be described as an alternative DRO DRL schedule
(Ferster & Skinner, 1957). This description incorporates the two alternative contingencies available
for contacting reinforcement, in that the organism
can either not respond at all or respond x or fewer
times. To draw attention to this detail may seem
trivial, but the effects on behavior could be as
substantial. Specifically, the schedule arrangement
supports both low-rate and zero responding.
Therefore, the spaced-responding DRL and
full-session DRL might establish disparate
patterns of responding and might be properly
used in different contexts. Of these two methods
of programming DRL schedules, the majority
of basic research has been conducted using
spaced-responding DRL schedules that maintain responding. However, in application, the
spaced-responding DRL is used infrequently
and is supplanted by the full-session DRL that
might actually be more likely to eliminate
responding. In other words, basic and applied
research that involves DRL arrangements has
focused on different contingencies that may
promote very different response patterns. The
explanation for this difference may be elucidated
when considering why applied researchers
would select a DRL arrangement as a means
of clinical intervention.
Spaced-responding DRL schedules have often
been implemented to reduce response forms that
315
are acceptable or valued, but only when the
responses occur at a low to moderate rate. These
include responses such as hand raising in
classroom settings (Austin & Bevan, 2011) and
independent eating or drinking during mealtimes
(Anglesea, Hoch, & Taylor, 2008; Lennox,
Miltenberger, & Donnelly, 1987; Wright &
Vollmer, 2002). Thus, the goal is to reduce the
frequency of the target response but not to
completely extinguish responding. The spacedresponding DRL schedule lends itself to cases for
which high rates of the target behavior could be
hazardous (e.g., rapid eating could lead to an
increased chance of choking) but complete
elimination of the response would alternatively
lead to dangerous complications (e.g., starvation).
In contrast, full-session DRL schedules, or the
alternative DRL DRO schedules, have been
implemented in cases in which the target behavior
is inappropriate and reduced rates are acceptable
but complete elimination is ideal. Common
examples of these behaviors include talking out
during class without permission (Deitz & Repp,
1973) or engaging in stereotypy (Singh, Dawson,
& Manning, 1981). The DRL component
accounts for the permissible (tolerable) rate of
inappropriate behavior to occur, whereas the
alternative DRO component schedules reinforcement during intervals of no responding.
Austin and Bevan (2011) used an amalgamation of procedures from both the spacedresponding and full-session DRL schedules to
decrease requests for help by three typically
developing elementary school students. The
authors reported considerable decreases in requests during the treatment component. However, some features of the results warrant
consideration. Although the ability to eliminate
the target response completely has often been
regarded as a strength of the DRO schedule, the
target response (requests for help) selected for this
study would likely not fall under that category.
The full-session DRL schedule has often been
preferred over the spaced-responding DRL
schedule because of the relative simplicity of
316
JOSHUA JESSEL and JOHN C. BORRERO
the nonresetting interval for teachers in classroom
settings (Deitz, 1977; Deitz & Repp, 1973);
however, this may not be clinically appropriate
when considering the possibility of extinguishing
an appropriate classroom response.
The purpose of the current study was to
compare the effects of the spaced-responding
DRL schedule and the full-session DRL schedule
in a preliminary human-operant investigation
using college students as participants. A DRObased schedule was specifically targeted because of
the frequency with which DRO procedures are
represented in the research literature related to
reducing problem behavior (Kahng, Iwata, &
Lewin, 2002). The study was designed to provide
laboratory (translational) evidence for possible
similarities or differences between the underlying
mechanisms of the procedures and whether or not
the change in a procedural taxonomy is warranted.
That is, if the full-session DRL schedule results in
maintained responding similar to that of the
spaced-responding DRL, then there would be
little evidence to suggest change, because the
implications for application will not differ for
practitioners and applied researchers.
METHOD
Participants
Sixteen university students (seven women,
nine men), with an age range of 18 to 29 years
old, were recruited for participation. All participants were sufficiently proficient in the manipulation of a computer mouse and had experience
using computers. Three participants served as
pilots during the initial stages of program
development. Two of the remaining 13 participants engaged in similar response rates during
the variable-ratio (VR) and extinction conditions, and one participant refused to wear the
headphones and could not hear when points
were being delivered. Therefore, 10 data sets
from the original 16 participants were produced;
one participant (P-8) completed two sessions.
(The reinforcement schedules and parameters for
each participant can be obtained from the
Supporting Information on the Wiley Online
Library.)
Apparatus and Settings
Participants were situated in a room (3 m by
3 m) with a desk (with laptop computer) and
chair. The participant was asked to be seated
while the instructions were read to him or her.
The participant was asked to read along with the
instructions on the computer screen, and to begin
the session when he or she was ready. The
instructions included the following statement:
Thank you for your participation in this
study. Your goal is to earn as many points as
possible before time is up. There is a
possibility of earning up to $50 (with other
monetary rewards for second and third place).
There are different ways to earn points.
Clicking on the colored buttons in different
patterns could add to your earnings, not affect
your earnings, or subtract from your earnings.
All of your earnings will be visible throughout
the experiment at the top of the screen, and a
tone will sound with each distribution. Your
time here will approximate 1 hour with a
minute break every 5 minutes. You are free to
leave at any point during this study; however,
you will only be eligible to win the monetary
prizes on completion. Remember, you are
trying to beat other participants so do your
best! Click the START button when you are
ready and good luck!
The program was created using Microsoft
Visual Basic and consisted of 24 colored squares
(100 by 100 pixels) in the center of the screen
with a text bar at the center top that displayed
real-time point accumulation. The colors of the
squares differed depending on the programmed
reinforcement schedule. The squares were stationary, and each click on a square made the
square disappear. All the squares that were clicked
regenerated after 6 s. Therefore all 24 squares
were visible every 6 s, and at no time was the
participant left without any squares to click.
VARIATIONS OF DIFFERENTIAL REINFORCEMENT
Design and Response Measurement
The computer program automatically recorded mouse clicks. Every 1 s, a preset automated timer recorded the current frequency of clicks
and points to a notepad file. Clicks were recorded
if they occurred on the squares. All other clicks on
the gray background produced no differential
consequences and were not recorded. The
primary dependent variable was rate of mouse
clicks expressed as responses per second.
ABA(C þ D) reversal designs were conducted
to assess possible effects between baseline and the
variations of the differential-reinforcement conditions. The initial reinforcer assessment (ABA)
consisted of three VR (A) blocks, followed by
three blocks of extinction (B), and a return to the
VR blocks. The imbedded multielement (C þ D)
design was implemented within treatment conditions to assess possible differentiating results of
the DRL IRT (C) and the DRO rate (D). The
design was extended to an ABA(C þ D)ACAD
for the one participant who opted to participate
for two sessions. However, all were given the
opportunity to participate for both sessions.
Procedure
Sessions lasted a minimum of 60 min and
consisted of three or more sequential 5-min
blocks, for a total of 12 to 15 blocks each session.
An optional 1-min break followed each block,
and a 5-min break followed every six blocks.
Points were delivered for mouse clicks based on
the relevant schedule of reinforcement in place.
After completion of the study, monetary rewards
of $50, $40, and $10 were awarded to the
participants who earned the most, second-most,
and third-most amount of points, respectively.
The initial nine blocks of each session
consisted of a reinforcer assessment. The reinforcer assessment was comprised of three blocks
of alternating extinction and VR schedules in an
ABA reversal design. The reinforcer assessment
was conducted to ensure that point delivery
increased responding.
317
Participants could not earn points during the
extinction phase. This arrangement included
blue squares that did not disappear when clicked
(the squares disappeared contingent on each click
during the reinforcement phases). We elected not
to remove squares contingent on clicks during
extinction, because three pilot participants
continued to click on the squares during sessions
in which points were never delivered.
During the VR phase, points were presented
on a VR 15 ( 5) schedule of reinforcement. The
scheduled mean number of responses was
selected based on pilot data that indicated that
responding would persist under a VR 15 ( 5).
The algorithm used to generate the VR schedule
was based on that provided by Dixon and MacLin
(2003). This schedule was correlated with green
squares. The VR schedule constituted the
reinforcement condition in the reinforcer assessment and the baseline condition in the comparative differential reinforcement assessment.
During the spaced-responding DRL phase,
points were presented contingent on the first
instance of a mouse click that followed the
completion of a preset interval in which no
responding occurred. The initial IRT interval was
calculated as twice the mean IRT interval during
the last VR phase. If clicks occurred at any point
before the interval elapsed, the automated timer
reset to the original IRT and no points were
delivered. Points were not presented following
the interval if no response occurred and were
withheld, without restarting the timer, until the
target response was emitted. This schedule was
correlated with yellow squares.
Both the spaced-responding DRL IRT and the
full-session DRL interval were calculated from
responding during the VR phase. During the fullsession DRL phase, the interval duration was
calculated as four times the average IRT during
the VR phase. Tolerance for the full-session DRL
was defined as the maximum frequency of
responding that could occur without resetting
the reinforcer-delivery interval. Tolerance was
calculated as half the mean response rate of the
318
JOSHUA JESSEL and JOHN C. BORRERO
target response during the VR condition. For
example, if the mean IRT during the VR phase
was 2 s, the spaced-responding DRL would be
calculated as 4 s (i.e., IRT > 4 s), the full-session
DRL interval would be set as 8 s, and tolerance
would be set as one response. Therefore, the
scheduled probability of reinforcer delivery
during the full-session DRL sessions remained
proportional to the spaced-responding DRL
condition. Although this method resulted in an
interval substantially longer than DRO intervals
typically conducted in applied contexts (see
Vollmer & Iwata, 1992), without this modification comparative results between the spacedresponding DRL and full-session DRL procedures would not be mutually interpretable.
Decreased response rates during the spacedresponding DRL are a function of increased IRTs
and longer intervals, relative to some baseline
response rate. On the other hand, short intervals
are preferred during DRO arrangements to
reduce the negative side effects of extinction
and increase contact with the scheduled reinforcement. Thus, we elected to increase the fullsession DRL interval rather than decrease the
spaced-responding IRT because decreases in
spaced-responding DRL durations relative to a
full-session DRL interval would have contraindicative or no effects on responding that was
already occurring slower than the imposed rate.
For example, if a participant is already responding
at a pace of one response every 10 s, a minimum
IRT of 5 s would not likely affect behavior. In
addition, in comparison to the full-session DRL
intervals commonly conducted in applied settings, the currently calculated intervals were
relatively small. The intervals from Dietz and
Repp (1973) consisted of entire class periods of
50 min. The negative effects often associated with
long intervals may be avoided by the addition of
tolerance.
During the full-session DRL condition, points
were delivered following the elapse of the interval,
whether or not any clicks occurred at any time
during the session block. However, if the
frequency of clicks exceeded that of the set
tolerance within the interval, points were not
delivered and the interval was restarted immediately after the violation of tolerance. This
condition was correlated with red squares.
Data Analysis
Data from a cumulative record were analyzed
across 300 1-s bins per session to determine
comparative optimal and allowable response rates
during the last sessions of the spaced-responding
DRL and full-session DRL conditions, respectively. The optimal response rate refers to the rate
of responding during the spaced-responding
DRL condition in which the most reinforcers
can be produced within the allotted session time.
The allowable response rate refers to the rate of
responding during the full-session DRL condition in which all reinforcers can be delivered
within the allotted session time without penalty
of point loss. The slope of the cumulative data
during the last session of each condition (i.e., VR,
spaced-responding DRL, full-session DRL) was
calculated by taking the response frequency at
10 s (Y1), subtracting it from the response
frequency at 290 s (Y2), and dividing the
difference by 280 (X2–X1).
A contingency strength analysis (Luczynski &
Hanley, 2009, 2010) was conducted to provide a
quantifiable value of the contingent relations
during the spaced-responding DRL phase and
the full-session DRL phase. The contingency
value was defined as the difference of two
disparate conditional probabilities: response
conditional probability and point conditional
probability. A positive contingency value from
the contingency strength analysis supports a
correlation between a response and a reinforcer;
the higher the positive value the stronger the
correlation. Therefore, responses with positive
contingencies are more likely to be strengthened,
whereas negative contingencies are more likely to
be weakened. By definition, for example, DRO
arrangements promote a negative contingency
between responding and reinforcer presentation
VARIATIONS OF DIFFERENTIAL REINFORCEMENT
(Vollmer, Borrero, Wright, Van Camp, &
Lalli, 2001).
The response conditional probability was the
quotient of the number of responses followed by
a point (1-s window) divided by the total number
of responses. A 1-s window was selected for two
reasons: (a) Previous research that has compared
window sizes suggests that short window sizes
(i.e., 2 s to 5 s) provide a more conservative
measure (Luczynski & Hanley, 2009), and (b)
little variance was observed between 1 s and 3 s
for the current data. The point conditional
probability was defined as the quotient of the
number of points not preceded by a response (1-s
window) divided by the total number of points
delivered. Therefore, a continuous reinforcement
schedule will result in a positive contingency
value of 1 because the response will always
precede reinforcer delivery. In contrast, a DRO
schedule without tolerance will result in a
contingency value of 1 because the numerator
of the response conditional probability will
always be 0 (i.e., zero responses will be followed
by a point within 1 s). The contingency strength
analysis was conducted to determine whether or
not the full-session DRL schedule resulted in
positive contingencies similar to most differential
reinforcement schedules or in a negative contingency, as is the case with the DRO schedule.
RESULTS
Figures 1 through 3 depict the results of the
reinforcer assessment and the comparison of
response-reducing differential reinforcement
techniques for all participants as responses per
second. The solid horizontal line represents
the optimal performance during the spacedresponding DRL condition, and the dashed
horizontal line represents the allowable performance during the full-session DRL condition.
Both differential reinforcement techniques reduced responding compared to baseline for all
participants. Furthermore, all participants continued to respond during the spaced-responding
319
DRL condition, whereas 60% of participants
ceased responding during the final blocks of the
full-session DRL condition.
Seven of the 10 participants produced response
rates below optimal performance during the
spaced-responding DRL conditions and did not
maximize the total possible points that could have
been earned (mean difference from optimal ¼
36.7%; range, 13.3% to 85.6%). However,
P-19 (Figure 2, bottom left) exhibited near
optimal responding (0.74 responses per second)
during the final spaced-responding DRL session
(difference from optimal ¼ 2.8%). Three participants’ rate of responding was above optimal
(mean difference from optimal ¼ 40.8%; range,
14.6% to 63.5%) with P-12 (Figure 1, bottom
right) exhibiting near optimal responding (0.7
responses per second) during the final spacedresponding DRL session (difference from
optimal ¼ 1.3%).
The full-session DRL schedule resulted in all
but P-4 (Figure 1, top left) producing mean
response rates (M ¼ 0.11 responses per second;
SD ¼ 0.13) below allowable performance (mean
difference from allowable ¼ 62.7%; range,
100% to 86.5%). Furthermore, there was a
20.4% difference from the allowable performance for the only participant (P-4) who
continued to respond near allowable (0.51
responses per second).
Results of the contingency strength analyses for
all 10 participants were positive (M ¼ 0.71;
SD ¼ 0.26) during the spaced-responding DRL
condition, and greater than that of the VR schedule
value (0.07). The top panel of Figure 4 shows data
for P-21, a representative example of the contingency strength analysis. That is, during the spacedresponding DRL condition, the response conditional probability exceeded the point conditional
probability. In addition, negative contingency
values (M ¼ 0.82; SD ¼ 0.28) were obtained
for 9 of 10 participants during the full-session DRL
condition. The bottom panel of Figure 4 shows
data for P-4, who was the only participant to
continue to respond at near allowable rates during
JOSHUA JESSEL and JOHN C. BORRERO
320
CLICKS (RPS)
3
VR 15(±5)
3
VR 15( ±5)
2
2
1
1
0
P-4
0
5
3
CLICKS (RPS)
EXT
10
15
2
1
1
P-7
0
5
10
BLOCKS
15
EXT
VR 15( ±5)
Optimal
Allowable
DRL-s
DRL-f
P-11
5
3
2
0
VR 15(±5)
10
15
10
15
P-12
5
BLOCKS
Figure 1. Response rates during the variable ratio (VR), extinction (EXT), spaced-responding DRL (DRL-s), and fullsession DRL (DRL-f ) conditions across sessions for P-4, P-11, P-7, and P-12. The solid horizontal line depicts the optimal
response rate for producing the most points during the spaced-responding DRL condition. The dashed horizontal line
depicts the allowable response rate without resetting point delivery during the full-session DRL condition. RPS ¼ responses
per second.
the full-session DRL condition, and who showed
the highest mean negative contingency (M ¼
0.11; SD ¼ 0.16) with the last session a positive
value (0.06) comparable to that of the positive VR
value (0.07).
We also evaluated responding during the VR,
spaced-responding, and DRL conditions as a
proportion of responding during each participant’s
extinction condition. The greatest proportional
increase in responding compared to extinction
occurred in the VR for all participants. In addition,
90% of the participants displayed higher rates of
proportional responding during the spaced-responding DRL condition (M ¼ 2.96; SD ¼ 1.72) compared to extinction. Notably, 30% of the
participants displayed higher rates of proportional
responding to extinction during the full-session
DRL condition (M ¼ 0.54; SD ¼ 0.51).
DISCUSSION
These findings replicate previous research on
DRL and DRO in that low-rate spaced-response
patterns and the elimination of responding were
obtained in the spaced-responding DRL and fullsession DRL conditions, respectively (Deitz &
Repp, 1973). Although the full-session DRL
schedule permitted a predetermined response rate
derived from the participants’ baseline performances, many participants stopped responding. We
also observed an overall negative contingency value
consistent with a DRO-based schedule during the
full-session DRL. Thus, the two schedules reduced
responding compared to the VR schedule of
reinforcement, with the full-session DRL arrangement resulting in the larger reduction.
Although previous research has supported the
use of full-session DRL schedules over spaced-
VARIATIONS OF DIFFERENTIAL REINFORCEMENT
CLICKS (RPS)
3
VR 15(±5)
3
VR 15( ±5)
2
2
1
1
0
P-15
0
5
3
CLICKS (RPS)
EXT
10
15
2
1
1
P-19
0
5
10
BLOCKS
15
EXT
VR 15( ±5)
Optimal
Allowable
DRL-s
DRL-f
P-17
5
3
2
0
VR 15(±5)
321
10
15
10
15
P-21
5
BLOCKS
Figure 2. Response rates during the variable ratio (VR), extinction (EXT), spaced-responding DRL (DRL-s), and fullsession DRL (DRL-f ) conditions across sessions for P-15, P-17, P-19, and P-21. The solid horizontal line depicts the optimal
response rate for producing the most points during the spaced-responding DRL condition. The dashed horizontal line
depicts the allowable response rate without resetting point delivery during the full-session DRL condition. RPS ¼ responses
per second.
responding DRL schedules due to the ease of
implementation (Deitz, 1977; Deitz &
Repp, 1973), results of the present study suggest
that functional differences in these procedures are
not simply a matter of structural semantics.
Evaluation and implementation of spaced-responding DRL and full-session DRL should be
considered in the context of clinical or research
goals. The more effortful approach (spacedresponding DRL) may be required when the goal
is to sustain responding, albeit at rates lower than
those produced under baseline conditions (e.g.,
applications to rapid eating, excessive hand
raising, or tattling). In addition, the limitations
of using spaced-responding DRL schedules (e.g.,
requires individual timers for each student to be
reset following responding below the pro-
grammed IRT) may be assuaged when considering the possibility of the use of widespread
technology such as tablets in classroom settings.
Applications could be created on tablets with
which teachers could select the IRT for each
student immediately before the start of classroom
activities. The teacher would only have to click on
the child’s name after each response to determine
whether or not he or she met the criteria for
reinforcer presentation. Future research could
assess the use of programmatic schedules of
reinforcement on handheld devices to reduce
work-related effort needed to create individualized interventions for classwide implementation.
The results of the contingency strength
analysis also provide insight into the possible
quantitative differences between the reductions
JOSHUA JESSEL and JOHN C. BORRERO
322
VR 15(±5)
CLICKS (RPS)
3
VR 15(±5)
EXT
3
VR 15( ±5)
Optimal
Allowable
2
0
2
DRL-s
1
1
DRL-f
P-18
0
5
10
BLOCKS
EXT
15
P-8
5
15
25
BLOCKS
Figure 3. Response rates during the variable ratio (VR), extinction (EXT), spaced-responding DRL (DRL-s), and fullsession DRL (DRL-f ) conditions across sessions for P-18 and P-8. The solid horizontal line depicts the optimal response rate
for producing the most points during the spaced-responding DRL condition. The dashed horizontal line depicts the
allowable response rate without resetting point delivery during the full-session DRL condition. RPS ¼ responses per second.
in responding under the spaced-responding and
full-session DRL schedules. The VR and spacedresponding DRL schedules both maintained
higher rates than those observed during extinction. This should not be surprising, considering
that a positive contingency between responding
and point delivery existed for both VR and
spaced-responding DRL schedules. Reductions
were observed during the spaced-responding
DRL sessions when mean rates were analyzed;
however, the contingency value was actually
greater than the comparative VR reinforcement
schedule from a within-session perspective. An
artificial reduction in response rates was created
in spaced-responding DRL schedules by reinforcing larger IRTs relative to baseline response rates.
In contrast, the full-session DRL schedule
produced a negative contingency between responding and point delivery.
A change in procedural taxonomy may be
warranted. Currently, little distinction is made
between the two variations of DRL schedules
(e.g., Cooper, Heron, & Heward, 2007) although
it is evident that they differ in process and
application. Assuming the two DRL schedules to
be equivalent could result in misapplication (i.e.,
inadvertently maintaining inappropriate behavior
with the spaced-responding DRL or inadvertently
eliminating appropriate behavior with the fullsession DRL). We suggest that the full-session
DRL could be expressed as a variation of a DRO
schedule and that the term DRO-with-tolerance
schedules could be considered when reducing
inappropriate behavior. In a very literal sense, a
predetermined tolerated level of inappropriate
behavior is overlaid on top of a DRO schedule.
Not only does the definition of DRO with
tolerance incorporate both components in the
basic arrangement (alternative DRO DRL), but it
also implicitly suggests targeting undesirable
behaviors. This simplifies the categorical usage
for practitioners as DRL schedules for appropriate
behavior and DRO schedules for inappropriate
behavior. Furthermore, modifying its classification to a DRO schedule (rather than the
previously reported DRL schedule) may improve
the selection of interventions by clarifying its
effects, by highlighting the units of analysis, and
by underscoring the behavioral mechanisms that
are responsible for effects obtained therein.
VARIATIONS OF DIFFERENTIAL REINFORCEMENT
CONTINGENCY VALUE
1.0
P-21
DRL-s
0.5
0.0
-0.5
DRL-f
-1.0
CONTINGENCY VALUE
1.0
P-4
0.5
0.0
-0.5
-1.0
2
BLOCKSKS
4
6
Figure 4. Data in the top panel illustrate a strong
positive contingency between responding and point delivery
(DRL-s) and a strong negative contingency between
responding and point delivery (DRL-f ). These are
representative data for 9 of 10 participants. Data in the
bottom panel are those from P-4, the only participant who
continued to respond at near allowable rates during the fullsession DRL condition.
Our brief experimental arrangement also could
be used to examine two previously reported
explanations for the suppressant effects of DRO
and whether similar outcomes would be observed
with DRO schedules implemented with tolerance.
One explanation is that “other behavior” that
occurs contiguously with the presentation of the
reinforcer displaces target responding (Ecott &
Crtichfield, 2004; Poling & Ryan, 1982). The
second explanation conceptualizes the DRO
schedule as a form of negative punishment because
responding delays reinforcer onset (Lattal, 2013).
323
The brevity of our procedures and emphasis on
temporal properties of responding also lend
themselves to evaluations for which mechanisms
of timing are thought to be operative or
important. Dube and McIlvane (2002) used
this type of assessment approach to evaluate
sensitivity to concurrent reinforcement schedules
by individuals with intellectual disabilities. Interestingly, most of the participants’ responding in
the current assessment was above optimal performance, resulting in mean pauses that were longer
than the set IRT. This is directly juxtaposed with
previous basic research with rats, in which subjects
were more likely to engage in below optimal
performances, resulting in mean pauses that were
shorter than the set IRT (Mazur, 1994). Although
little can be responsibly deduced from these
differences due to the procedural variance
between the present human operant arrangement
and basic research with nonhumans, it may still be
important to note for those interested in studying
temporal dimensions of behavior.
One limitation of the present study was the lack
of multiple tolerance settings. It is therefore
unknown if an increased ratio of allowable
responses to the same interval duration would
result in responding more similar to that under the
spaced-responding DRL. The tolerance may have
been too stringent and resulted in increased contact
with the DRO-related contingency of the fullsession DRL schedule. However, many participants who discontinued responding reported the
ability to continue without resetting the timer and
in some cases reported the approximate full-session
DRL interval and tolerance setting. Therefore,
even though the participants could tact the
contingencies with surprising accuracy, they chose
not to respond at all. Of further interest was the
fact that many of the studies that have implemented the full-session DRL schedule used
arbitrary reinforcers intended to compete with
multiple unknown sources of qualitatively dissimilar reinforcement (e.g., Austin & Bevan, 2011;
Deitz & Repp, 1973). The current study used only
one source of reinforcement and one value of
324
JOSHUA JESSEL and JOHN C. BORRERO
tolerance for each participant, but future research
could extend the generality of these findings to
multiple sources of reinforcement and multiple
values of tolerance.
REFERENCES
Anglesea, M. M., Hoch, H., & Taylor, B. A. (2008).
Reducing rapid eating in teenagers with autism: Use of a
pager prompt. Journal of Applied Behavior Analysis, 41,
107–111. doi: 10.1901/jaba.2008.41-107
Austin, J. L., & Bevan, D. (2011). Using differential
reinforcement of low rates to reduce children’s requests
for teacher attention. Journal of Applied Behavior
Analysis, 44, 451–461. doi: 10.1901/jaba.2011.44-451
Bird, F., Hepburn, S., Rhodes, K., & Moniz, D. (1991).
Multiple reinforcement contingencies to reduce aggression, self-injury, and dysfunctional verbal behaviors in
an adult who is sensory impaired. Behavioral Interventions, 6, 367–383. doi: 10.1002/bin.2360060506
Catania, A. C. (2013). Learning (5th ed.). Upper Saddle
River, NJ: Prentice Hall.
Cooper, J. O., Heron, T. E., & Heward, W. L. (2007).
Applied behavior analysis (2nd ed.). Upper Saddle River,
NJ: Prentice Hall.
Deitz, S. M. (1977). An analysis of programming DRL
schedules in educational settings. Behaviour Research
and Therapy, 15, 103–111. doi: 10.1016/0005-7967
(77)90093-6
Dietz,[sic], S. M., & Repp, A. C. (1973). Decreasing
classroom misbehavior through the use of DRL
schedules of reinforcement. Journal of Applied Behavior
Analysis, 6, 457–463. doi: 10.1901/jaba.1973.6-457
Dixon, M. R., & MacLin, O. H. (2003). Visual basic for
behavioral psychologists. Reno, NV: Context Press.
Dube, W. V., & McIlvane, W. J. (2002). Quantitative
assessments of sensitivity to reinforcement contingencies in mental retardation. American Journal of Mental
Retardation, 107, 136–145.
Ecott, C. L., & Critchfield, T. S. (2004). Noncontingent
reinforcement, alternative reinforcement, and the
matching law: A laboratory demonstration. Journal of
Applied Behavior Analysis, 37, 249–265. doi: 10.1901/
jaba.2004.37-249
Ferster, C. B., & Skinner, B. F. (1957). Schedules of
reinforcement. East Norwalk, CT: Appleton-CenturyCrofts, doi: 10.1037/10627-000
Hagopian, L. P., Kuhn, D. E., & Strother, G. E. (2009).
Targeting social skills deficits in an adolescent with pervasive
developmental disorder. Journal of Applied Behavior Analysis,
42, 907–911. doi: 10.1901/jaba. 2009.42-907
Kahng, S., Iwata, B. A., & Lewin, A. B. (2002). Behavioral
treatment of self-injury 1964 to 2000. American Journal
on Mental Retardation, 107, 212–221.
Lattal, K. A. (2013). The five pillars of the experimental
analysis of behavior. In G. J. Madden, W. V. Dube,
T. D. Hackenberg, G. P. Hanley, & K. A. Lattal (Eds.),
APA handbook of behavior analysis (pp. 33–63).
Washington, DC: American Psychological Association.
doi: 10.1037/13937-002
Lennox, D. B., Miltenberger, R. G., & Donnelly, D. R.
(1987). Response interruption and DRL for the
reduction of rapid eating. Journal of Applied Behavior
Analysis, 20, 279–284. doi: 10.1901/jaba.1987.20-279
Luczynski, K. C., & Hanley, G. P. (2009). Do children prefer
contingencies? An evaluation of the efficacy of and
preference for contingent versus noncontingent social
reinforcement during play. Journal of Applied Behavior
Analysis, 42, 511–525. doi: 10.1901/jaba.2009.42-511
Luczynski, K. C., & Hanley, G. P. (2010). Examining the
generality of children’s preference for contingent
reinforcement via extension to different responses,
reinforcers, and schedules. Journal of Applied Behavior
Analysis, 43, 397–409. doi: 10.1901/jaba.2010.43-397
Mazur, J. E. (1994). Learning and behavior. Englewood
Cliffs, NJ: Prentice Hall.
Poling, A., & Ryan, C. (1982). Differential-reinforcementof-other-behavior schedules. Behavior Modification, 6,
3–21. doi: 10.1177/01454455820061001
Shaw, R., & Simms, T. (2009). Reducing attentionmaintained behavior through the use of positive
punishment, differential reinforcement of low rates,
and response marking. Behavioral Interventions, 24,
249–263. doi: 10.1002/bin.287
Singh, N. N., Dawson, M. J., & Manning, P. (1981). Effects
of spaced responding DRL on the stereotyped behavior of
profoundly retarded persons. Journal of Applied Behavior
Analysis, 14, 521–526. doi: 10.1901/jaba.1981.14-521
Turner, J. M., Green, G., & Braunling-McMorrow, D.
(1990). Differential reinforcement of low rates of
responding (DRL) to reduce dysfunctional social
behaviors of a head injured man. Behavioral Interventions, 5, 15–27. doi: 10.1002/bin.2360050103
Vollmer, T. R., Borrero, J. C., Wright, C. S., Van Camp, C.,
& Lalli, J. S. (2001). Identifying possible contingencies
during descriptive analyses of severe behavior disorders.
Journal of Applied Behavior Analysis, 34, 269–287. doi:
10.1901/jaba.2001.34-269
Vollmer, T. R., & Iwata, B. A. (1992). Differential
reinforcement as treatment for behavior disorders:
Procedural and functional variations. Research in Developmental Disabilities, 13, 393–417. doi: 10.1016/08914222(92)90013-V
Wright, C. S., & Vollmer, T. R. (2002). Evaluation of a
treatment package to reduce rapid eating. Journal of
Applied Behavior Analysis, 35, 89–93. doi: 10.1901/
jaba.2002.35-89
Received June 15, 2012
Final acceptance December 12, 2013
Action Editor, Michael Kelley
Copyright of Journal of Applied Behavior Analysis is the property of Wiley-Blackwell and its
content may not be copied or emailed to multiple sites or posted to a listserv without the
copyright holder's express written permission. However, users may print, download, or email
articles for individual use.
to do so," to keep students from prematurely reading the
next page. The second page gives one of three sets of the
instructions: "You will be shown a series of cards. Try to
remember the (words on) (numbers on). (colors
of) the cards
.
in the order they are shdwn."
Each sheet can be !4 of a regular sized sheet of paper. The
three data collection pages and the cover page can all by
typed on one 8% X 11 in. sheet of paper. The three different
sets of instructions can be printed on another page. Each
data collection sheet contains six lines, numbered 1 to 6,
and instructions to remember the (words on) (numbers on)
(colors of) the cards in the order they were presented. For
those instructed to "remember the word," data collection
sheets should be arranged as follows: (a) recall words, (b)
recall numbers, and (c) recall colors. For the "remember the
number" group, the order is: (a) recall numbers, (b) recall
words, and (c) recall colors. Finally, for those told to "remember the colors" the order is: (a) recall colors, (b) recall
words, and (c) recall numbers.
Take one of each type of booklet, randomly arrange them
in a stack; take another three (one of each type), randomly
arrange them and add them to the stack; and so on. This
random ordering of the booklets and students' selection of
where thev will sit determine their assignment to the three
conditions. You assure identical treatment of the groups
because you show them the cards in the same way and at the
same time.
u
Conducting the Demonstration
When all students have their booklets, ask t h e ~ nto open
the booklet and read only the first page. Next, show the six
cards for about 5 s each. Be sure to show each card for the
same length of time by counting silently to the same number
at a regular rate. After showing the cards, tell students to
turn to the next page, follow the instructions, and continue
through the booklet until they are finished. After the students have written as many items as they can recall, show
them the cards again and have them score their own responses. Ask them to write on the cover page the number of
words, numbers, and colors they got correct. Then ask them
to assemble in groups according to what they were instructed
to remember and to find the group mean for each of the three
kinds of information.
The main point of the activity is to demonstrate that, on
the average, each group will have remembered more items in
the category they were asked to remember. Only those asked
to remember the colors of the cards get many of the colors
correct. The same basic effect occurs in the other two
groups, but to a lesser degree. Those with the "number"
instructions usuallv remember more numbers than the other
groups, and those with the "word" instructions tend to remember more words than the other groups. I conducted a 3
x 3 (Instructions x Recall Category) analysis of variance on
the data from 30 students who participated in this demonstration. There was no main effect for instructions, F(2, 27)
< 1, and no main effect for recall category, F(2, 54) = 2.08,
ns, but there was a significant interaction, F(2, 54) = 17.70,
p .< .01. The means and standard deviations for this sample
are shown in Table 1.
Table 1. Means and Standard Deviations of ltems
Recalled Out of Six ltems in Each Category
by Groups With Different lnstructionsa
Items Recalled
Color
Word
Number
Instructions
M
SD
M
SD
M
SD
Remember color
Remember word
Remember number
5.1
1.7
0.7
1.9
1.3
1.3
2.3
4.1
3.4
1.8
1.9
1.5
1.6
3.2
4.4
2.0
2.0
1.6
an = 10 per condition.
What Does it Show?
Some important elements of conducting an experiment
are illustrated by the procedure. Subjects were randomly
assigned to conditions, and all other variables were kept
constant.
A discussion of what the students were expecting can
yield interesting information. Usually a couple of people
figure out the experiment and remember the colors even
though they were not instructed to do so. This fact can be
used to illustrate that subjects in experiments often act as
problem solvers.
As the statistical analysis shows, students usually recall
about the same number of items regardless of the instructions, and no one category is more memorable than any
other. However, each group remembers best the category
they were told to remember. I use the fact that attention
increases memory to emphasize the importance of looking at
chapter outlines or lists of learning objectives in the textbook before studying. Pay attention!
Note
Requests for reprints should be sent to Janet D. Larsen, Department
of Psychology, John Carroll University, University Heights, OH
441 18.
Demonstrating Differential Reinforcement
by Shaping Classroom Participation
Gordon K. Hodge
Nancy H. Nelson
University of New Mexico
A classroom demonstration using differential reinforcement was
devised to shape classroom participation of 14 students in an
introductory psychology lab. Based on our observations and student comments, the technque was useful for illustrating how
reinforcers shape behavior. The demonstration facilitated students' understanding of operant conditioning procedures and
seemed to encourage a more equitable distribution of classroom
participation for all students.
Principles of operant conditioning are easily presented in
Vol. 18, No. 4, December 1991
Downloaded from top.sagepub.com at Apollo Group - UOP on September 12, 2015
classroom demonstrations. For example, descriptions of students operantly shaping their instructors' behaviors have
been reported (Chrisler, 1988; Melvin, 1988). As a variation on this theme, we devised a demonstration in which the
instructor shapes the students' level of class participation
using a differential reinforcement procedure.
In our experience, uneven distribution of student participation in the classroom is common. In large lecture sections of 400 to 600 students, opportunities for participation
are sometimes limited by time constraints and the intimidating atmosphere. But in small classes, such as seminars, labs,
or discussion sections, the ideal scenario is one in which all
students contribute, and discussions dominated by the assertive few are minimized.
We strive to foster creative exchange and discussion of
ideas in the introductory psychology labs. As stated in the
syllabus, students receive a grade for class participation.
They are encouraged and, presumably, motivated to take
part in discussions; still, many students do not participate.
In one lab section, we noted that three students overly
participated in discussions at the expense of other students
who rarely spoke. We believed it would be advantageous to
foster more equitable interactions. We also saw an opportunity to implement an educationally valuable demonstration that would enhance previously learned class material.
Method
Subjects
Subjects were 8 women and 6 men enrolled in an introductory psychology lab at the University of New Mexico.
Materials
Two weeks before the actual demonstration, students circled the value on a 7-point scale that best indicated their
own level of class participation. The scale ranged from I
never participate ( 1 ) to 1 always participate ( 7 ) . These ratings
provided a baseline level of self-perceived participation for
each student.
After the demonstration and before debriefing, each student completed an anonymous questionnaire consisting of
the following three items:
1. In your own words, describe the demonstration implemented during today's class. Include whether or
not you believe the demonstration influenced your
level of class participation.
2. Was the demonstration useful in illustrating how
reinforcers may be used in an operant conditioning
procedure? Explain how the demonstration was
useful and possible ways it may be changed andlor
improved.
3. Any additional thoughts concerning this demonstration.
Design and Procedure
One week after the lecture on learning, the lab instructor
distributed the scale assessing each student's perceived level
of class participation. T o avoid biasing the demonstration,
students were not told the reason for filling out the questionnaire. The scale was used to assist the instructor in determining the appropriate differential reinforcement schedule to be
implemented during the demonstration.
The demonstration using differential reinforcement to
shape classroom participation was implemented 2 weeks
after the rating scales were distributed. Each student's initials were placed on the top of the chalkboard at the front of
the class. The reinforcer consisted of a plus mark placed
underneath a student's name whenever the desired behavior
(increased or decreased participation) was emitted. The instructor determined before the demonstration which students would receive a reinforcer for either participating or
not participating, based on the rating scales and familiarity
with class dynamics.
Among the 14 students, 3 were judged as overparticipators; they were reinforced only when they did not participate or interrupt or when specifically called on by the
instructor after raising their hands. The 5 students who
rarely participated were reinforced for making even the
slightest effort to participate; for example, hand raising,
saying anything (correct or not) when called on, and, in one
instance, making eye contact with the instructor. These
students were then reinforced less frequently as they began
to emit more responses according to general shaping procedures (Gordon, 1989). The remaining 6 students normally participated in a moderate fash~onand were reinforced
on a variable ratio schedule t~ maintain their active
participation.
Following the demonstration and before debriefing, a
short questionnaire was given to assess whether the students
caught on to the demonstration and to get their feedback
and suggestions. Debriefing consisted of discussing the items
on the questionnaire in a classroom forum and reemphasizing the principles of operant conditioning, shaping, and
differential reinforcement.
Results
Student responses on the rating scale reflected the instructor's perceptions of participation levels. Only one student, rated by the instructor as a 1 (I never participate),
placed a rating of 2 on his scale. The scales, therefore,
appeared to complement the ratings made independently hv
the instructor, providing a reasonably reliable tool for devising a differential reinforcement schedule for each student.
Based on the instructor's subjective observation, class participation seemed noticeably more balanced after the technique was implemented. Overparticipators contributed
much less; underparticipators contributed more often and
enthusiastically. Only one of the identified underparticipators remained reticent. This balance in participation
was noticed by 5 of the students on their questionnaires.
One wrote that "people like an upbeat situation and are
encouraged to join in," and another student noted that "btudents that [sic] don't normally contribute hegan to
participate. "
Based on the responses to the first item on the questionnaire, 12 of 14 students identified the demonstration as an
example of operant conditioning or the use of ,I shaping
Teaching of Psychology
Downloaded from top.sagepub.com at Apollo Group - UOP on September 12, 2015
procedure or both. One student identified it as a motivational study (the demonstration was implemented during
a lecture on motivation), and one student did not answer the
question.
In response to the second part of Item 1 (whether or not
the demonstration influenced their level of class participation-either increasing or decreasing it), 10 of 14 students
believed that the demonstration affected their class participation. One student stated "I think some people talked a
lot more than usual," and another stated "many students
gave input that usually do not." The other 4 students did not
believe the demonstration affected their participation.
In response to Item 2, 12 of the 14 students found the
demonstration useful in illustrating the role of reinforcers in
an operant conditioning procedure, noting that the demonstration was "closely related to our computer assignment"
(which dealt with shaping) and better "integrated our lecture notes." Four students mentioned that they would prefer
a more potent reinforcer, such as extra credit, rather than
"just a plus mark."
Discussion
The demonstration was a valuable and worthwhile way to
illustrate differential reinforcement in shaping classroom
participation. The usefulness of the demonstration, based
on responses to the questionnaire, is evident, although there
are obvious limitations. The demonstration is limited to a
small class size (approximately 20 students) in order for the
instructor to implement an effective differential reinforcement procedure tailored for each student. With larger
groups, it would be difficult to keep track of the target responses and dispense reinforcers in a timely fashion.
Of some concern was how students, particularly overparticipators, perceived the fairness of the procedure. For example, overparticipators might have wondered why others received plus marks and they did not. On the questionnaires,
two students expressed frustration for not receiving reinforcement of their active participation. One student wrote
that the activity was "quite frustrating, considering that I
didn't get a plus until the latter portion of the demonstration
after participating considerably." That overparticipators experienced frustration when their normally high participation levels went unrewarded was not surprising. In general,
overparticipators seem frustrated whenever the instructor
fails to call on them or limits their comments. The concern
is whether frustration elicited by the demonstration was notably different from feelings ordinarily elicited during routine classroom management.
l ~ to initial mispercepSome frustration was ~ r o b a b due
tions that the goal was to increase everyone's participation
equally, rather than differentially. As the demonstration
progressed, however, overparticipators responded by curtailing their behavior in order to earn plus marks. One student's
strategy provided insight into the process: "I initially increased participation to ascertain whether I would receive
points. This attempt was to no avail so I proceeded to lessen
my degree or amount of participation" (for which the student was then reinforced). Although frustration occurred,
there were no indications from questionnaire responses or
discussions during debriefing that overparticipators thought
the experience was unfair. Both students who expressed ini-
tial frustration reported that the demonstration was interesting and useful. Nevertheless, instructors should be alert for
signs of unusual discomfort or frustration and be ready to end
the demonstration and initiate debriefing as necessary.
In discussing the demonstration with students, it is worthwhile to point out that the changes in participation frequencies probably reflected a real-life application of an operant
conditioning procedure. Moreover, even though students
became aware of the purpose of the demonstration while it
was ongoing, their behavior nevertheless changed in response to the procedures (cf. Blanchard & Johnson, 1982).
Discussion could then focus on students' ideas of how similar
procedures could be applied in other situations (e.g., encouraging more balanced communication in a personal
relationship).
Althoughnot quantified, some positive effectson class participation appeared to remain throughout the semester. This
pleasant residual effect may have occurred because students
and instructor were now more aware of their class behavior.
Students who normally did not participate may have felt
more comfortable about contributing after their first experience in speaking, or, possibly, the class dynamics became
less threatening and more comfortable to these students.
Ways to improve this demonstration include developing a
more objective, less intrusive assessment of class participation in addition to the questionnaire. One possibility would
be to use a hidden tape recorder or video camera to record a
baseline session before the demonstration, record the demonstration, and then have an objective third party score the
tape(s).
The demonstration is a useful technique for illustrating
differential reinforcement and for encouraging more equitable participation in small classes.
References
Blanchard, K., & Johnson, S. (1982). The one minute manager.
New York: Berkley.
Chrisler, J. C. (1988). Conditioning the instructor's behavior: A
class project in psychology of learning. Teaching of Psychology,
15, 135-137.
Gordon, W. C. (1989). Learning and memory. Pacific Grove, CA:
BrooksICole.
Melvin, K. B. (1988). Rating class participation: The proflpeer
method. Teaching of Psychology, 15, 137-139.
Notes
1. We thank Frank A. Logan, Charles L. Brewer, and the anonymous reviewers for their helpful comments.
2. Requests for reprints should be sent to Gordon K. Hodge, Department of Psychology, University of New Mexico, Albuquerque, NM 87131.
Demonstrating Personality Scale
Validation Procedures
Robert C. Reinehr
Southwestern University
A technique is described for demonstrating personality scale walidation techniques to students in introductory psychology classes.
Vol. 18, No. 4, December 1991
Downloaded from top.sagepub.com at Apollo Group - UOP on September 12, 2015
Copyright of American Journal of Psychotherapy is the property of Association for the Advancement of
Psychotherapy and its content may not be copied or emailed to multiple sites or posted to a listserv without the
copyright holder's express written permission. However, users may print, download, or email articles for
individual use.
Purchase answer to see full
attachment