JOURNAL OF THE EXPERIMENTAL ANALYSIS OF BEHAVIOR 2008, 90, 53–60 NUMBER 1 ( JULY)

VISUAL REINFORCEMENT IN THE FEMALE SIAMESE FIGHTING FISH, BETTA SPLENDENS

MIRARI ELCORO, STEPHANIE P. DA SILVA, AND KENNON A. LATTAL
WEST VIRGINIA UNIVERSITY

Operant conditioning with Betta splendens (Bettas) has been investigated extensively using males of the
species. Ethological studies of female Bettas have revealed aggressive interactions that qualitatively
parallel those between male Bettas. Given these similarities, four experiments were conducted with
female Bettas to examine the generality of a widely reported finding with males: mirror-image
reinforcement. Swimming through a ring was reinforced by a 10–s mirror presentation. As with males,
ring swimming was acquired and maintained when mirror presentations were immediate (Experiments
1, 2, and 3) and delayed (Experiment 4). The failure of conventional extinction (Experiments 1 and 2)
and response-independent mirror presentations (Experiment 3) to maintain responding confirmed the
reinforcing properties of mirror presentation. These results extend previous findings of mirror images
as reinforcers in males of the same species and illustrate a complementarity between behavioral ecology
and the experimental analysis of behavior.
Key words: mirror-image reinforcement, extinction, response-independent mirror presentation,
aggression, ring swimming, female Betta splendens

________________________________________

The aggressive display of male Siamese forcement also has been reported for other
fighting fish, Betta splendens (hereafter, Bettas), members of the Anabantoid family to which the
is reliably elicited by images of another male Betta belongs. In male paradise fish (Macro-
Betta. Visual reinforcement of the responding podus opercularis), for example, mirror images
of these males also has been demonstrated elicit aggressive displays (Francis, 1983) and
repeatedly when such images are presented strengthen and maintain operant responses
dependent on a specified response. In differ- (Melvin & Anson, 1970).
ent experiments, visual access to models of Female Bettas also display aggressively to-
other male Bettas (Thompson & Sturm, 1965), ward one another in a manner similar to that
a live male or female conspecific (Rnic, 1977), of males (Braddock & Braddock, 1955; Ro-
mirror presentations resulting in an image of bertson, 1979). Braddock and Braddock, for
the fish serving as the experimental subject example, described the female aggressive
(Lattal & Metzger, 1994; Thompson, 1963), display as follows:
and film clips of another Betta (Turnbough & Individual challenging is a stereotyped reac-
Lloyd, 1973) all are sufficient to develop and tion. The aggressor approaches the other fish,
maintain operant responding in male Bettas. usually at an angle, and stops at a distance of
In addition, the color of the fish presented as a less than 1 inch. It then erects its gill covers,
model (Thompson & Sturm), duration of holds its body rigidly in a flexed position,
mirror presentation (Wirth, Lattal, & Hopko, erects all its median fins, and rapidly vibrates
2003), delay of mirror presentation from the the pectoral fins. The color of the body and
response (Lattal & Metzger), and schedule of fins becomes more intense, and striping tends
to disappear. Individual challenges may be
reinforcement (Turnbough & Lloyd) each almost instantaneous or prolonged for 30 sec-
influence the probability of the operant onds or more. (p. 155)
response. Such responding for visual rein-
Even though Braddock and Braddock
(1955) observe that females can behave
Stephanie P. da Silva currently is affiliated with the aggressively with one another, females typically
Department of Psychology & Sociology at Columbus State
University, Columbus, GA, 31907 (dasilva_stephanie@ are housed communally without the intensity
colstate.edu). of aggressive interactions that preclude such
Address correspondence to: Mirari Elcoro, 53 Campus housing of males. On the one hand, this lack
Drive, 1124 Life Sciences Building, Department of of propensity for aggression that allows com-
Psychology, West Virginia University, Morgantown, WV
6506-6040. Ph # (304) 293-2001; Fax # (304) 293-6606 munal living suggests that visual access to a
(e-mail: Mirari.Elcoro@mail.wvu.edu). mirror image might not function with females
doi: 10.1901/jeab.2008.90-53 as it does with males. On the other hand, there

53
54 MIRARI ELCORO et al.

Fig. 1. Diagram of the apparatus: (a) Aquarium where the fish were housed; (b) Ring through which each fish swims;
(c) Plexiglas frame that holds the photobeam placed outside the narrow sides of the aquarium; (d) One part of the
photobeam; (e) Plexiglas rods; (f ) Box with the two light bulbs; (g) One-way mirror; and (h) Black panel to cover the
mirror before and after the session.

are qualitative similarities of the display of isolated from the control equipment. Ambient
female and male Bettas toward conspecifics. light in the room was programmed on a 12 hr
This latter observation combined with the light/dark cycle. During each session, the
susceptibility of males to visual reinforcement ambient light was on. Temperature and
as noted suggests that access to a mirror image humidity in the room were held constant. An
by a female Betta might function as a reinforcer example of the apparatus is diagrammed in
as it does in males. The present experiments Figure 1 (all letter references in this descrip-
therefore examined mirror presentation as a tion are to this figure). Each aquarium was
reinforcer of operant responding by female 20.5 cm high by 30 cm long by 14.5 cm wide
Bettas. and contained approximately 9.5 L of water
(a). The device used to record responses was a
METHOD plastic ring (b), with a 4.5 cm inner diameter,
and supported by a Plexiglas frame (cf. Wirth
Subjects et al., 2003) that fit over the two narrow sides
Fifteen experimentally naı̈ve female Bettas of the aquarium (c). Part of the frame
were obtained from local pet stores. Each was remained outside the aquarium and contained
housed individually in the aquariums described a light source and photoreceptor (d ). The
below. Each aquarium was arranged to ensure part of the frame that was placed inside the
visual isolation of each Betta. Water was main- aquarium consisted of two 1.4-cm-diameter
tained between 21 and 24uC and was replaced clear plastic rods (e) that terminated near the
regularly after first being chemically treated to center of the aquarium at the outer edge of
eliminate chlorine. Food was provided daily, the plastic ring. The Plexiglas rods allowed
20–30 min after the experimental session. transmission of infrared light from one side of
the aquarium to the other without being
Apparatus refracted by the water. The ring was placed
Each of four aquariums was housed in a in the aquarium such that its opening was
small room that was acoustically and visually parallel to the long sides. When the fish swam
 VISUAL REINFORCEMENT IN FEMALE BETTA SPLENDENS 55

Table 1
Number of Sessions and Mean (SD) Response Rate for Each Fish in Each Experiment.

Number of Sessions Mean (SD) Response Rate

Fish FR 1 EXT FR 1 FR 1 EXT FR 1
Experiment 1
F1 29 3 12.6 (0.08) 3.6 (0.02)
F2 29 3 17.4 (0.16) 6.0 (0.03)
F3 29 3 40.2 (0.19) 6.0 (0.03)
F4 27 3 15.6 (0.08) 6.0 (0.01)
Experiment 2
2A 34 13 16 16.8 (0.14) 2.4 (0.03) 9.6 (0.07)
4A 34 13 21 8.4 (0.06) 4.2 (0.05) 30.6 (0.25)
F1 12 8 11 9.6 (0.03) 2.4 (0.02) 12.0 (0.04)
F2 12 8 12 9.0 (0.05) 2.4 (0.03) 13.2 (0.09)
Experiment 3
F11 20 10 12 12.6 (0.08) 1.2 (0.02) 8.4 (0.04)
F12 26 10 17 22.2 (0.05) 0.0 (0.00) 13.2 (0.05)
F13 40 10 27 39.0 (0.04) 1.8 (0.04) 36.0 (0.13)
Tandem FR 1 DRO 10 s Tandem FR 1 DRO 10 s
Experiment 4
M1 16 13.2 (0.10)
M2 16 23.4 (0.14)
M3 10 0.0 (0.00)
M4 16 3.0 (0.59)
Note: Mean and SD of the response rates were calculated for the last six sessions of each condition, except for the
extinction condition of Experiment 1 for which the mean and SD are based on the three extinction sessions.

through the ring, breaking the photobeam, a Thus, the subjects were never removed from
response was recorded. their aquariums. Each session lasted 120 min
A 25.5 cm high by 35.5 cm long by 13.5 cm in the first three experiments and 90 min in
wide box housed two 40-W 110-V light bulbs Experiment 4. The sequence of conditions
behind a clear plastic front panel ( f ). This and number of sessions per condition for each
panel was located in front of one of the long experiment is shown in Table 1. Conditions
sides of the aquarium. A one-way mirror were changed when there were no systematic
(25.5 cm by 35.5 cm) was placed between the trends in response rates as judged by visual
clear panel and the aquarium (g). When the inspection in Experiments 1–3, except as
light bulbs were on, no image of the fish was noted below. In Experiment 4, the number
displayed, and when the light bulbs were of sessions was fixed for each subject. Because
turned off, the reflection of the fish was there were only four aquariums, 4 fish were
displayed on the mirror thus created. The studied at a given time.
other three sides of the aquariums were The following procedures were common to
covered with black panels to ensure visual all four experiments: There was no explicit
isolation. At the end of each session, the training of the response. Immediately before
mirror also was covered by a black panel (h). each session, the frame was emplaced and the
The photocells were interfaced to a personal lights were illuminated behind the one-way
computer, located in an adjacent room, mirror. The mirror then was uncovered and
through a MedAssociatesH interface. Program- the session began. During each session, rein-
ming and data collection were arranged using forcement consisted of a 10-s period during
MedPCH software. which the lights behind the mirror were
extinguished, making the mirror reflective. A
Procedure 10-s period of mirror access was selected based
All sessions were conducted in the home on Wirth et al.’s (2003) findings that durations
aquariums during the light-on cycle, 6–7 days between 3 and 15 s were effective in maintain-
per week at approximately the same time. ing responding of male Bettas (cf. Turnbough
56 MIRARI ELCORO et al.

& Lloyd, 1973). The unique features of each presentations occurred independently of the
experiment were as follows. responses of the fish. The values of each VT
Experiment 1. Acquisition and maintenance schedule were yoked to the mean interrein-
of ring swimming were studied by exposing 4 forcer intervals (IRI) of the last 6 sessions of
fish to a fixed-ratio (FR) 1 schedule of the preceding FR 1 condition to equate the
reinforcement. When the fish swam through number of reinforcers in the two conditions.
the ring, the lights behind the mirror were This mean value was calculated for each fish
extinguished for 10 s, thereby making the separately. Values for the VT schedule were
mirror reflective. After 10 s, the lights again generated according to a Fleshler and Hoff-
were illuminated, removing the mirror’s re- man (1962) progression. After 10 such yoked
flection. This sequence repeated following VT sessions for each fish, the FR 1 schedule
each such response. was reinstated for 12–27 sessions.
After 27–29 sessions of exposure to the FR 1 Experiment 4. In the first three experiments,
schedule, extinction (EXT), defined as the during the FR 1 schedules, a mirror presenta-
removal of the mirror presentation following a tion always immediately followed the ring-
response (i.e., the lights behind the mirror swimming response. Lattal and Metzger
were on continuously), was in effect for three (1994) showed that responding of male Bettas
consecutive sessions. was acquired even if the mirror presentation
Experiment 2. To examine further how did not immediately follow the response. This
responding is maintained by the mirror experiment examined delayed reinforcement
presentation, the procedures of Experiment with females. Four experimentally naı̈ve fe-
1 were replicated and extended, with a reversal males were exposed to a tandem FR 1 differen-
following extinction. Four experimentally na- tial-reinforcement-of-other-behavior (DRO) 10-s
ı̈ve fish were exposed to the FR 1 schedule schedule. Thus, once the session started, when
described in Experiment 1. Of these 4 fish, the fish swam through the ring, an unsignaled
Fish 2A and 4A acquired ring swimming. resetting 10-s delay was initiated (cf. Lattal &
Therefore, Fish F1 and F2, previously used in Metzger, 1994). During this delay, each re-
Experiment 1, were retrained on the FR 1 sponse restarted the 10-s interval. Following a
schedule and used in this experiment. After 10-s period without a response, the mirror was
12–34 sessions of exposure to this FR 1 presented for 10 s and then removed as
schedule, EXT, as described in the first described in Experiment 1. This sequence
experiment, was in effect for 8–13 sessions. repeated following each response. This proce-
Finally, the FR 1 schedule was reinstated for dure was in effect for 16 sessions.
11–21 sessions. During the first FR 1 condition,
for Fish 2A EXT was introduced even though RESULTS
responding did not appear stable because the
responding was on an upward trajectory and Response rates were calculated by dividing
the subsequent extinction condition was ex- the number of ring swims by the session time,
pected to decrease responding. During the last minus the mirror presentation time.
FR 1 condition for Fish 4A, the experiment was Experiment 1. Figure 2 shows responses per
terminated after 21 sessions. hr during each session for each fish. Fish F1,
Experiment 3. To examine the role of the F2, and F3 responded for the first time during
dependency between the ring-swimming re- the third session. Fish F4 first responded
sponse and mirror presentation on response during the eighth session. Of the average
maintenance, responding was compared when response rates over the last six sessions, Fish F3
reinforcers were delivered under an FR 1 showed the highest (40.2 responses per hr)
schedule of mirror presentation to those and Fish F1 the lowest (12.6 responses per hr).
delivered according to a yoked variable-time For each fish, response rates decreased sub-
(VT) schedule. Three experimentally naı̈ve stantially during EXT. Fish F4 ceased respond-
fish were exposed to an FR 1 schedule of ing.
mirror presentation as described in Experi- Experiment 2. Figure 3 shows responses per
ment 1. After 20–40 sessions on the FR 1 hr for each fish. Fish 2A first responded during
schedule, a VT schedule was in effect for 10 the third session and Fish 4A during the sixth
sessions. During the VT schedule, mirror session. Responding occurred earlier in Fish 1
 VISUAL REINFORCEMENT IN FEMALE BETTA SPLENDENS 57

Fig. 2. Responses per hr across sessions for each fish in Experiment 1. Conditions (FR 1 & EXT) are separated by a
dashed vertical line.

and 2, presumably because of their previous EXT condition, relative to those during the
exposure to the FR 1 schedule. Across fish, preceding FR 1 schedule. After several subse-
average response rates for the last six sessions quent sessions of the return to the FR 1
during both FR 1 conditions ranged from 8.4– schedule, however, response rates were con-
30.6 responses per hr. For Fish 2A, F1, and F2, siderably higher for each fish than during the
responding decreased substantially during preceding EXT condition.
EXT. For these same fish, responding recov- Experiment 3. Figure 4 shows that, for each
ered when the FR 1 schedule was reinstated. fish, response rates were substantially lower
Fish 4A responded at lower rates during the FR during the VT schedule relative to those
1 schedule than did the other fish. During during the FR 1 schedule. Response rates of
EXT, response rates for Fish 4A decreased each fish recovered when the FR 1 was
slightly, and primarily toward the end of the reinstated. The average response rate for the

Fig. 3. Responses per hr across sessions for each fish in Experiment 2. Conditions (FR 1, EXT, & FR 1) are separated
by dashed vertical lines.
58 MIRARI ELCORO et al.

observations revealed displays that were con-

sistent with those observed by Braddock and
Braddock (1955) when females were placed
together and allowed to interact.

DISCUSSION
Braddock and Braddock (1955) concluded
that ‘‘the female pattern [of aggression] …
resembles closely that of the male [Bettas]’’
(p. 170). The results of the four present
experiments taken together show that re-
sponse-dependent mirror presentation main-
tains responding in female Bettas in a similar
manner to that of males of the same species
under similar conditions (cf. Lattal & Metz-
ger, 1994; Wirth et al., 2003). The congru-
ence between the aggressive behavior of male

Fig. 4. Responses per hr over the final six sessions of

the initial FR 1 condition and across all sessions of the
yoked VT and the final FR 1 conditions for each fish in
Experiment 3. Conditions (FR 1, VT, & FR 1) are
separated by dashed vertical lines.

last six sessions during both FR 1 conditions

(see Table 1) was highest for Fish F13 (39.0
and 36.0 responses per hr) and lowest for Fish
F11 (12.6 and 8.4 responses per hr).
Experiment 4. Of the 4 fish, 3 acquired ring
swimming even though reinforcement always
was delayed by 10 s from a response. Figure 5
shows that response rates of Fish M1 and M2
increased systematically across sessions, but
Fish M4 responded more erratically and at
much lower rates during the experiment.
As with most other studies of the reinforcing
efficacy of mirror presentations for Bettas,
systematic observations of the visual displays
during mirror presentations were not made
during these experiments. The female Bettas, Fig. 5. Responses per hr across sessions for the 3 fish in
however, were informally observed from time Experiment 4 that acquired responding. Note the differ-
to time during mirror presentations. These ences in the scale of the y axis for Fish M4.
 VISUAL REINFORCEMENT IN FEMALE BETTA SPLENDENS 59

Bettas exposed directly to one another and response rates of 4 males varied between about
the reinforcing properties of mirror presen- 30 and 50 responses per hr toward the end of
tation with these fish is well established. The that condition. In the present Experiment 4,
present results in concert with the naturalistic the response rates of the females ranged
observations of Braddock and Braddock show between 3.0 and 23.4 responses per hr toward
a similar congruence between aggressive the end (last six sessions) of the experiment.
behavior and visual reinforcement for fe- Two variables that may have contributed to the
males. low response rates in all of these studies were
Several features of the present findings can the absence of shaping of the response,
be compared to the results obtained with male discussed above, and the nature of the operant
Bettas in previous experiments. Both Lattal and response of swimming through a ring. This
Metzger (1994) and Wirth et al. (2003) used latter response involves movement of the
the same apparatus employed in the present entire body through a liquid medium that
study, but with male Bettas. Across the four offers considerable resistance, as opposed to
present experiments, the acquisition of ring the bar press studied in rats or the key peck
swimming with immediate or delayed (for characteristic of pigeon operant conditioning
10 s) reinforcement was obtained in 12 out experiments. In addition, there may be incom-
15 female fish. These numbers are similar to patible responses occurring concurrently with
results reported in the two aforementioned the operant response, such as persisting at
experiments. There were no obvious differ- mirror approach following termination of the
ences in appearance or behavior outside the mirror presentation.
experimental sessions between the 12 fish that Braddock and Braddock’s (1955) seminal
acquired ring swimming and the 3 that did not laboratory-based research in an ethological
(2 in Experiment 2 and 1 in Experiment 4). tradition was the first to document both
Some of the variability in acquisition may topographical characteristics and controlling
relate to the absence of shaping of the variables of aggressive behavior of female
response. Although response shaping is the Bettas. The previously noted congruence be-
sine qua non of developing new responses in tween the Braddock and Braddock observa-
operant conditioning procedures, several ex- tions and the reinforcing efficacy of a mirror
periments have shown that shaping is unnec- presentation for both males and females
essary in the development of sustained oper- invites questions about the ecological signifi-
ant responding. Lattal and Gleeson (1990), for cance of such a reinforcer, in terms of its
example, showed that responding of both rats origins and possible role in survival. Skinner
and pigeons was acquired without any re- (1966) described how susceptibility to certain
sponse training even though the operant kinds of reinforcers might come about
response produced a reinforcer only after an through phylogenic contingencies. One func-
unsignaled delay period that commenced with tion of aggressive behavior by females was
the response (Lattal & Williams, 1991; Wilken- suggested by Braddock and Braddock to be
field, Nickel, Blakely, & Poling, 1992). Shaping that of maintaining territory. If this is the case
also was unnecessary for response acquisition in natural settings, female Bettas could be
with either male or, in these experiments, thought of as actively seeking out potential
female Bettas. intruders in the same way that a hungry rat
Under the present FR 1 schedules of forages for food. If this is the case, then the
immediate reinforcement with females (Ex- sequence of events can be likened to those
periments 1, 2, and 3), response rates varied proposed by Abarca and Fantino (1982),
between 0 and 40.2 responses per hr. Re- Collier (1980), Lea (1979), and others. First,
sponse rates under FR 1 schedules also were searching occurs in the form of the operant
variable across individual male Bettas in both response (e.g., Fantino, 1991) followed by
Lattal and Metzger (1994) and Wirth et al. procuring, in the form of approaching the
(2003), ranging between near-zero and almost mirror, and then handling, in the form of the
400 responses per hr, depending on the aggressive display, which Braddock and Brad-
conditions in effect (Wirth et al. manipulated dock have shown to be a necessary first step in
reinforcer duration). With the unsignaled an actual attack, which would, in this model,
resetting 10-s delays in Lattal and Metzger, constitute the final consumption.
60 MIRARI ELCORO et al.

At this point, there are insufficient etholog- Lattal, K. A., & Metzger, B. (1994). Response acquisition by
ical data to either substantiate or refute the Siamese fighting fish (Betta splendens) with delayed
visual reinforcement. Journal of the Experimental Anal-
territorial function of aggression in female ysis of Behavior, 61, 35–44.
Bettas. Regardless of its survival significance, Lattal, K. A., & Williams, A. M. (1991). Body weight and
the aggressive behavior of female Bettas, response acquisition with delayed reinforcement.
including the susceptibility to reinforcement Journal of the Experimental Analysis of Behavior, 67,
of operant responses leading to mirror-image 131–143.
Lea, S. E. G. (1979). Foraging and reinforcement
presentations, has been shown in the present schedules in the pigeon: Optimal and non-optimal
experiments to be a reliable finding in many aspects of choice. Animal Behaviour, 27, 875–886.
members of the species. With respect to Melvin, K. B., & Anson, J. E. (1970). Image-induced
survival significance, Braddock and Braddock aggressive display: Reinforcement in paradise fish.
(1955) suggest that ‘‘the student of animal Psychological Record, 20, 225–228.
Pitts, R. C., & Malagodi, E. F. (1996). Effects of
behavior … is still interested in what various reinforcement amount on attack induced under a
species are capable of doing. It is the range of fixed-interval schedule in pigeon. Journal of the
capability that is of primary importance rather Experimental Analysis of Behavior, 65, 93–110.
than what does or does not happen in nature’’ Rnic, A. (1977). Male versus female Siamese fighting fish
(Braddock & Braddock, 1955, p. 171). The as reinforcing stimuli for conspecific males in single-
and two-choice operant situations. Learning & Moti-
present results not only suggest dimensions of vation, 8, 263–274.
operant responding in female Bettas, but also Robertson, C. M. (1979). Aspects of sexual discrimination
complement findings with other species (e.g., by female Siamese fighting fish (Betta splendens
fighting cocks, Thompson, 1964; other ana- Regan). Behaviour, 70, 323–336.
bantoid fishes, Melvin & Anson, 1970; and Skinner, B. F. (1966, September 9). The phylogeny and
ontogeny of behavior. Science, 153, 1205–1213.
pigeons, Pitts & Malagodi, 1996) to show the Thompson, T. (1963, July 5). Visual reinforcement in
generality of mirror-image reinforcement. Siamese fighting fish. Science, 141, 55–57.
Thompson, T. (1964). Visual reinforcement in fighting
cocks. Journal of the Experimental Analysis of Behavior, 7,
REFERENCES 45–49.
Abarca, N., & Fantino, E. (1982). Choice and foraging. Thompson, T., & Sturm, T. (1965). Visual-reinforcer color,
Journal of the Experimental Analysis of Behavior, 38, and operant behavior in Siamese fighting fish. Journal
117–123. of the Experimental Analysis of Behavior, 8, 341–344.
Braddock, J. C., & Braddock, Z. I. (1955). Aggressive Turnbough, P. D., & Lloyd, K. E. (1973). Operant
behavior among females of the Siamese fighting fish, responding in Siamese fighting fish (Betta splendens)
Betta splendens. Physiological Zoology, 2, 152–172. as a function of schedule of reinforcement and visual
Collier, G. (1980). An ecological analysis of motivation. In reinforcement. Journal of the Experimental Analysis of
F. M. Toates, & T. R. Halliday (Eds.), Analysis of Behavior, 20, 355–362.
motivational processes (pp. 125–151). London: Academ- Wilkenfield, J., Nickel, M., Blakely, E., & Poling, A. (1992).
ic Press. Acquisition of lever press responding in rats with
Fantino, E. (1991). Behavioral ecology. In I. H. Iversen, & delayed reinforcement: A comparison of three proce-
K. A. Lattal (Eds.), Experimental analysis of behavior Part dures. Journal of the Experimental Analysis of Behavior, 58,
2 (pp. 117–153). New York: Elsevier. 431–443.
Fleshler, M., & Hoffman, H. S. (1962). A progression for Wirth, O., Lattal, K. A., & Hopko, S. (2003). Using visual
generating variable-interval schedules. Journal of the reinforcement to establish stimulus control of re-
Experimental Analysis of Behavior, 5, 529–530. sponding of Siamese fighting fish (Betta splendens).
Francis, R. C. (1983). Experiential effects on agonistic Journal of Comparative Psychology, 117, 111–118.
behavior in the paradise fish, Macropodus opercularis.
Behaviour, 85, 292–313.
Lattal, K. A., & Gleeson, S. (1990). Response acquisition
with delayed reinforcement. Journal of Experimental Received: February 20, 2007
Psychology: Animal Behavior Processes, 16, 27–39. Final acceptance: February 8, 2008