Efectos de la duración de la señal sobre la adquisición de la respuesta con programas de reforzamiento demorado definidos temporalmente

Marco Antonio Pulido Rull¹, Mariana Rubí Gonzalez and Cornelie Backer Hoekstra

Laboratorio de Condicionamiento Operante, Universidad Intercontinental

¹ Please send all correspondence related with this paper to the first author at:
Avenida Universidad 1330, Edificio A, Interior 1102, Col. del Carmen Coyoacán, ]]> C.P. 04100, México, DF. Email: mpulido@uic.edu.mx

Received: 23 April, 2008
Reviewed: 21 May, 2008
Accepted: 25 August 2008

Abstract

This study assessed the effects of signal duration and t^d duration on response acquisition of lever–pressing by rats. Thirty naïve male Wistar rats were exposed to one of ten different 32–s temporally defined schedules of delayed, signaled reinforcement. Reinforcement cycles were divided into two temporal time windows, t^d and t delta; responses emitted during t^d produced reinforcement at the end of the cycle; responses emitted during t delta had no programmed consequences. t^d was located at the beginning of the reinforcement cycle and was fixed at one of two different values (4–s or 8–s). The first response emitted during td produced an audible tone of different duration; tone duration was varied across groups Results showed variability, however evidence of response acquisition was more apparent with the long t^d duration and with the longer signals. This effect could probably be attributed to the interaction of Pavlovian, and mnemic variables with stimulus discrimination processes.

Key words: Delayed reinforcement, lever–pressing, response acquisition, signal duration, temporally defined schedules.

Se evaluaron los efectos de la duración de la señal y la duración de t^d sobre la adquisición del palanqueo en ratas. Treinta ratas macho de cepa Wistar fueron expuestas a uno de diez programas de reforzamiento demorado señalado, definidos temporalmente de 32–s. Los programas se dividieron en dos porciones, t^d y t delta. La primera respuesta durante t^d producía alimento al Analizar el ciclo; las respuestas durante t delta no tuvieron consecuencias programadas. t^d se colocó al inicio del ciclo y se fijó en uno de dos valores (4–s u 8–s). La primera respuesta emitida durante t^d produjo un tono auditivo; la duración del tono varió entre los grupos. Los resultados mostraron que la adquisición de la respuesta fue más aparente cuando t^d tuvo un valor de 8–s y con las señales más largas. El efecto podría atribuirse a la interacción de variables Pavlovianas, y mnémicas así como a procesos de discriminación de estímulos.

Palabras clave: Reforzamiento demorado, palanqueo, adquisición de la respuesta, duración de la señal, programas definidos temporalmente.

Studies have shown that lever pressing by rats and key pecking by pigeons may be established in the absence of explicit shaping and under conditions of delayed unsignaled reinforcement (Lattal & Gleeson, 1990; Van Haaren, 1992, Dickinson, Watt & Griffith, 1992). Response acquisition with delayed reinforcement may occur under different delay of reinforcement procedures (Wilkenfield, Nickel, Blakely & Poling, 1992) and under intermittent reinforcement schedules (Bruner, Avila & Gallardo, 1994). Critchfield & Lattal, (1993) and Lattal & Metzger (1994) have found evidence of response acquisition under delayed reinforcement of spatially defined operants by rats and Siamese fighting fish.

The previous studies suggests some research has focused on the description of the conditions under which response acquisition of free–operant responding may occur, other studies however have centered on the description of variables that may facilitate the development of new behavior repertories. For instance, Critchfield & Lattal (1993) exposed rats to tandem FR1, RDO 30–s schedules where the response identified for reinforcement was the interruption of an infrared light beam located at the top of a conditioning chamber. The authors compared conditions where the emission of the response produced a brief auditory signal with conditions in which the response produced no programmed signal. Results showed response frequency in both groups increased gradually and reached very similar rates. Results also showed that the response–reinforcer ratio was higher in the signaled group than in the unsignaled group .

Schlinger, & Blakely (1994), exposed rats to tandem RF1, RDO 4 and 10–s where light beam interruption produced reinforcement after the delay interval. Results showed that light beam interruption rate consistently increased in both the signaled and the unsignaled groups, however subjects exposed to the condition where the response was followed by a brief tone, produced higher response rates, in fewer sessions, than subjects that did not receive the cue. The effect of the tone was consistent across both delay durations.

Williams, (1999) exposed groups of naive rats to tandem FR1, FT 30–s schedules where subjects could be exposed to and unsignalled delay condition or to signaled conditions that differed in cue temporal location during the delay interval. For some subjects a 5–s change in illumination occurred immediately after the response; other subjects received the cue 5–s before reinforcement delivery. Results showed signal location had opposite effects, when the signal immediately followed the response, it substantially facilitated learning, in contrast, when the signal occurred just before reinforcement delivery, response acquisition did not develop. Both reinforcement and response rates produced by the unsignaled delay condition were intermediate between the two signaled conditions.

Using temporally defined schedules of reinforcement (Schoenfeld, & Cole, 1972), Pulido, Lanzagorta & Lopez (2005) exposed experimentally naive rats to 32–s reinforcement cycles (T) where a response located at the beginning of the cycle produced reinforcement at the end of T. In some groups a response produced a 4–s tone, in other groups the response produced an 8–s tone, in still other groups the response identified for reinforcement was unsignaled. In general, lever pressing in signaled reinforcement conditions was higher than lever pressing frequency in the unsignaled group. Lever pressing rates produced by the two signaled conditions were similar, however response variability was higher in the group with the shorter signal.

In general, the studies presented in this paper suggest response acquisition may be enhanced by the presentation of exteroceptive stimuli, immediately after the response identified for reinforcement occurs. Results produced so far also suggest signal duration may module the response enhancing properties of the signal. For instance in the previously described study by Critchfield & Lattal, response rates produced by groups exposed to briefly signaled and unsignaled conditions did not differ. In contrast, results produced in both the previously described Pulido et al. and Williams studies, showed that longer signals following the response, may produce response rates that are considerably higher than those produced under unsignaled conditions.

With the exception of the Pulido et al (2005) study, the rest of the studies featured in this paper have used experimental procedures that allow programmed reinforcement rate to co vary with change in delay duration (Lattal, 1987). Because programmed reinforcement rate has been found to have important effects on response rate (Clark, 1958) a second purpose of the present study was to assess the effects of signal duration on response acquisition using temporally defined schedules of delayed signal reinforcement (Schoenfeld, Lang & Mankoff, 1973) to program reinforcement contingencies. Weil suggested this type of temporally defined schedule could avoid confusion between programmed reinforcement rate and delay duration because interreinforcer interval is fixed at a predetermined value and delay duration is varied by means of systematically changing the duration of the opportunity to respond (t^d). Experiments conducted using this metric have been previously used to study unsignaled delay of reinforcement effects on response acquisition (Bruner, Pulido & Escobar, 1999; 2000) and the effects of signaled delay of reinforcement on the same dependent variable (Pulido, Lopez & Lanzagorta 2005; Pulido & Lopez, 2006).

METHOD

Subjects

Thirty experimentally naive male Wistar Lewis rats were used as subjects. All subjects were approximately Ave months old at the beginning of the study. Each subject's weight was registered on five consecutive days under free–feeding conditions to determine ad libitum body weight; access to food was then restricted until all subjects reached 80% of their free–feeding weight. Subjects were kept at their prescribed body weights throughout the experiment by means of supplementary feeding following each experimental session. Subjects were kept in the Laboratory vivarium under constant temperature conditions and a twelve–hour light–dark cycle (lights on a 7:00 a.m.). All experimental subjects were kept in individual cages with free access to water.

Apparatus:

Sessions were conducted in a Med Associates (ENV 008) rodent operant conditioning chamber made of stainless steel and transparent Plexiglas. The space in which the subjects were studied measured 21 cm in height by 30.5 cm length by 24.1 cm depth. A 3 cm stainless steel rolled lever was placed on the left panel of front wall of the chamber. The lever was placed 3.0 cm above the floor and 3.0 cm apart from a trough type pellet receptacle located on the center panel of the front wall. The pellet receptacle was placed 1 cm above the floor and consisted of a 5 cm square opening 2cm in depth. A force of at least 30.5 grams was required to register the response. A depression of the lever produced an audible click and was counted as a response. A .45 mg pellet dispenser delivered 2 .20 mg pellets in each emission. Pellets were produced by means of remolding pulverized Purina Nutri Cubes. Two 1.1 W, 28 Vdc lights were used to illuminate the experimental chamber. One light was placed on the center panel of back wall of the chamber and was used as a house light. The second light (a pilot light with a white glass translucent cover) was placed 5 cm above the lever and was used to present visual stimuli to the subjects. A sonalert, which delivered a 74.13 dB auditory signal, was attached to the external back wall of the experimental chamber, 5 cm to the right of the house light. The conditioning chamber was placed inside a sound–attenuating larger wooden box equipped with a ventilating fan. Experimental events were programmed and recorded using an IBM compatible 386 microcomputer equipped with an industrial automation card (Advantech, PC–Labcard 725) coupled to a relay rack.

Procedure:

During the first session, with the lever absent from the chamber, each rat was exposed to a magazine training procedure. Magazine training consisted of thirty consecutive response–independent food deliveries using an FT 30–s schedule. All experimental subjects consumed the food in the tray after just one exposure to the schedule. On the second session (and on thirty additional consecutive sessions) the lever was inside the experimental chamber and all subjects were exposed to temporally defined schedules of reinforcement (Schoenfeld & Cole, 1972). The schedules consisted of a repetitive time cycle of fixed duration (T). Two different components alternated within the reinforcement cycle (t^d and t delta). The first response emitted during t^d produced reinforcement at the end of the 32–s cycle; responses during t delta were recorded but had no programmed consequences. t^d was always placed at the beginning of the cycle in order to separate the response identified for reinforcement from food delivery. Because responses could occur during t delta, (or after one response had occurred during t^d) the experimental procedure used in the study may be characterized as a variable delay of reinforcement procedure (Lattal, 1987; Schoenfeld, Cole, Lang & Mankoff, 1973). This experiment can also be conceptualized as a between subjects factorial design with two factors: 1) t^d duration (4–s or 8–s). t^d was varied in order to assess at least two different response–reinforcer temporal relations. For the shorter t^d value, programmed reinforcer delay could reach a minimum duration of 28–s; for the longer t^d value programmed reinforcement delay could reach a minimum duration of 24–s. 2) Signal duration. For those subjects exposed to t^d=4–s signal duration could be fixed at either 0, .5, 4, 8 or 32–s; for those subjects exposed to t^d=8–s conditions, signal duration could be fixed at either 0, .5, 8, 16 or 32.s. It may be objected that this design does not allow the comparison of the same signal durations across the two t^d values, however it allows a perfect proportionality between t^d duration and signal duration. Without pertinent data to make an empirically supported design decision, proportionality was chosen over comparability.

Three animals were assigned to each experimental condition. Sessions were conducted 6 days per week at approximately the same time each day. Each session lasted one hour or the time necessary to obtain thirty reinforcers, whichever occurred first.

RESULTS

Figure 2 shows mean number of responses emitted per minute for all three animals in each session for all sessions and experimental conditions. Response rate is shown as a function of exposure to the different reinforcement schedules used in the study. To facilitate the description of the data, short t^d conditions are placed on the upper part of the figure; long t^d conditions are placed on the lower part of the figure. The first column in the figure shows the unsignaled t^d condition. For both t^d conditions, signal duration increases from left to right.

In general, results showed response rates were lower in unsignaled conditions than in those conditions where a cue occurred during delay interval. Response rate variability differed among the two t^d conditions; in general responding was more consistent in the long t^d conditions. Under long t^d conditions, response rates were low in the unsignaled and .5–s signal conditions and comparatively higher with longer signal durations (reaching its maximum height with subject J8 in the 32–s signal condition). Under short t^d conditions responding in the unsignaled, .5–s and 4–s conditions was relatively low and generally circumscribed to one subject (K16 in the .5–s signal condition) or the later experimental sessions (as was the case of subjects I15 and I17 in the 4–s signal condition). Five subjects reached high response rates under 8–s and 32–s signal conditions, particularly J4 (in the 8–s signal condition) and I10 (in the 32–s signal condition).

Due to the high response variability shown by figure 2, and in order to assess the effects of the independent variables in the first and last experimental sessions, figure 3 was designed to show the mean response rate of each subject for the two t^d conditions in the first and last 5 experimental sessions.

Figure 3 confirms the pattern showed in figure 2, regarding response variability; in general individual means were more closely "bunched together" under the long t^d duration than in the short condition Also response rates in the first and last sessions differed less under the long t^d condition than in the short condition. In relation with the first experimental sessions, response rates reached higher numbers under signaled conditions when t^d=8–s; with the shorter t^d value this was only true with .5–s and 8–s signals. Regarding the last experimental sessions, most subjects exposed to signaled conditions emitted more responses than those subjects exposed to unsignaled conditions. Additionally the highest response rates produced in each condition are a direct function of signal duration. A slope produced by the highest response rates produced under the long t^d value is less steep than that produced when t^d=4–s, however response variability is considerably smaller in the former than in the later condition. Evaluation of the slopes produced by the highest subject in each condition is only and indirect was of assessing the effects of signal duration on response rate; in order to more thoroughly weight the effects of this independent variable, the regression equation between signal duration and response rate produced in the last Ave experimental sessions was calculated for both t^d durations. The regression equation for t^d=4–s may be described as follows: y = (.107)x + 1.812. The slope of the equation attained statistical significance (t(74) = 4.203, p =.000). The regression equation when t^d=8–s may be described as follows y = (.091)x + 1.990. The slope of the equation attained statistical significance (t(74) = 5.071, p =.000). In order to further assess the effects of signal duration, and also assess the effects of t^d duration in the last Ave experimental sessions, an overall two–way analysis of variance was also calculated. The analysis showed signal duration had a significant effect on response rate (F(5/140)=22.849, p=.000) however t^d duration did not (F(1/150)=.059, p=.809); the interaction between td duration and signal duration did not attain significance (F(3/140)=1.014, p=.389).

Figure 4 is in general agreement with the data presented in both figures 2 and 3; response rates in the signaled conditions are higher than those produced in the unsignaled conditions. Under the long t^d duration response distribution across the reinforcement cycle is relatively flat, the exception is the 32–s signal condition were response rate decreases during the second bin and then gradually increases again. Under t^d=4–s conditions, response rates are once again flat under unsignaled and brief signal conditions but at least one subject in the 8–s signal condition and two subjects in the 32–s signal condition show the distinctive "V" pattern previously described for the t^d=4–s groups. Data from figure 4 should be analyzed carefully as no provision was made, in the computer program, to separate cycles where a signal had occurred, from those cycles where a signal did not occur. A recent study by Pulido, Backer & Rubi (2003) suggests the distinctive "V" pattern previously described may be composed of signaled cycles where high response rates occur during the onset of the signal (bin 1) and are then followed by a sharp decline in responding once the signal has commenced (bin 2). The increase in response rate observed in bins 3 and 4 may occur in unsignaled cycles where the subject gradually increases response rate following reinforcement delivery. Data produced by Pulido, Paz & Sosa (2008) also suggest that once the response targeted for reinforcement has occurred, local response rate drops abruptly during the delay interval.

The issue regarding the identification of criteria that may be used to determine if response acquisition has taken place is far form solved. In the present study response acquisition was arbitrarily defined as obtaining at least 10 reinforcers at one of six different moments during the course of the experiment (5, 10, 15, 20, 25 or 30 sessions). Although any definition of response acquisition is arbitrary, the selection of 10 reinforcers as "the cut point" was preferred because the probability of earning at least 10 reinforcers given that no acquisition has occurred (zero mean), in a normal distribution is .0001. The decision to keep 10 reinforcers as response acquisition criterion was also supported by the fact that the frequency analyses carried on the data presented on table 1 did not change substantially with more stringent cut points.

Table 1 shows the experimental label of the subjects that had earned 10 reinforcers by the time a certain number of experimental sessions had elapsed. In the table the session number is presented in the top horizontal frame and the experimental conditions are presented in the left column.

In general table 1 shows that response acquisition criteria was met more frequently in the last three columns (sessions 20, 25, 30) than in the first three (sessions 5, 10, 15) (if subjects are counted only once). In table 1 the criterion was met 22 times in the last columns and only 17 times in the first columns.

Table 1 was also used to assess the effects of t^d duration on response acquisition. Results from the table showed that the response acquisition criterion was met more frequently in the long t^d conditions (14 times) than in the short t^d conditions (9 times).

Table 1 showed that the response acquisition criterion was met more frequently with the long signals than with the short ones. The table shows that in the absence of the signal, or in 0.5–s signal conditions, the response acquisition criterion was met only 7 times; in contrast response acquisition criterion was met 11 times with the longer signal durations (8–s and 32–s under t^d=4–s conditions and 16–s and 32–s under t^d=8–s conditions).

An important number of studies have shown that interreinforcer interval duration is strongly related to signal effects in delayed reinforcement studies (Schaal, Schuh &Branch, 1992; Schaal, Odum and Shahan, 2000). Programmed interreinforcer interval is always constant under temporally defined schedules of reinforcement, however obtained cycle duration may be highly variable (Schoenfeld y Cole, 1972). To determine how interreinforcer interval duration could have affected the present results, table 2 shows mean reinforcement rate for each subject for the last Ave experimental sessions. Table 2 was also elaborated as an indirect way of assessing signal presentation effects. As the experimental procedure only presents signals when the subject emits a response during t^d, and only responses presented during t^d produce reinforcement, obtained reinforcement rates are also a measure of obtained signal presentation rates.

The lowest reinforcement rates in the table occur in the short t^d condition and in the absence of the signal. Reinforcement rate for the long t^d conditions is significantly different than that produced by the short td conditions (t(28)=2.59, p<.05) (and thus the two t^d durations also differ in obtained signal presentation rate).

DISCUSSION

The results from the present study suggest that the presentation of an exteroceptive stimulus during delay interval may considerably enhance evidence suggesting acquisition of free operant behavior. This finding is in general agreement with results produced by Critchfield & Lattal, (1993) Schlinger & Blakely, (1994) and Pulido, Lanzagorta & López, (2005). The results from the present study also showed longer signals are probably associated with higher response rates than shorter ones (at least under the long t^d duration). The effects of signal duration were however variable and lack the clear impact found in the Schaal & Branch studies (1988; 1990). The comparison of the present results with the Schaal and Branch experiments, suggest that the effects of independent variables may be more variable when response acquisition is used as a dependent variable, relative to those studies where steady state behavior is used as dependent variable. This finding had already been reported by Pulido, López & Lanzagorta (2005) and Pulido & López (2006) who exposed rats to temporally defined schedules and varied the type of signal the animals received during t^d. Results showed response rates were low when the subjects received a non–contingent signal and comparably higher when the signal was response–produced. The result was the same regardless of the dependent variable used, however response rate variability was considerably higher when response acquisition was the dependent variable. The authors acknowledge that lack of systematic effects and behavior variability could also be attributed to the experimental procedure, specifically to the subject's possibility of changing signal duration by responding early or late during t^d.

Schaal and Branch's 1990 study showed response rate during the delay interval was relatively unaffected by signal duration, a finding that generally agrees with the relatively flat functions showed in figure 4. As classical conditioning studies have shown that the CR varies with delay duration in trace conditioning procedures (Ellison & Konorski, 1964), the results of the present study appear to give support to the Schaal & Branch's suggestion that a respondent–conditioning account of signal effects in delayed signaled reinforcement schedules, appears insufficient (although this argument would be more compelling if the present study had been capable of presenting a separate analysis of response distribution during signaled and unsignaled reinforcement cycles). Reinforcement rates produced by the two different t^d durations showed reinforcement was more frequent under the long t^d; signal effects were also more noticeable with the longer t^d duration. This last finding is also difficult to reconcile with a respondent–conditioning account of the present data, because most studies regarding the effects of interreinforcer interval on classical conditioning have shown that signal effects are more potent under longer interreinforcer intervals (Gibbon, Baldock, Locurto, Gold & Terrace, 1977; Kaplan, 1984). It may be argued however that subjects assigned to the short t^d duration earned fewer reinforcers and were thus less exposed to the particularities of their corresponding reinforcement schedules. Perhaps if the present study had equated schedule exposure for both t^d durations based on a performance criterion (instead of simply equating them in terms of the number of experimental sessions) results could make a stronger case against (or in favor) of a respondent–conditioning interpretation of signaled delayed reinforcement.

If a respondent–conditioning interpretation is insufficient to adequately account for the present data, how can they be accounted for? Richards (1981) and Schaal & Branch, (1990) have suggested behavior different from the one selected for reinforcement may be superstitiously reinforced during the delay interval. In unsignaled delay conditions these behaviors may generalize to all components of the schedule and compete with lever pressing (thus lowering overall response rate). In the presence of a signal (specially a long signal) superstitious behaviors may come under discriminative control of the cue and thus their generalization to other components of the schedule may be minimized (which in turn increases overall response rate). As was the case of the respondent–conditioning hypothesis, a superstitious behavior generalization account of the present data could better be assessed had data from reinforced and unreinforced cycles been separated, (in any case however figure 4 suggests some subjects exposed to the longer signals may be behaving in the way predicted by the superstitious behavior hypothesis). As it was mentioned before, a study by Pulido, Backer & Rubi (2003) suggests that the distinctive "V" patterns that appeared in figure 4 may in fact consist of the added effects of two different conditions. The high response rates produced in the first bin and the subsequent decline observed in the second bin may in fact reflect what occurs in the signaled reinforcement cycles. Apparently the onset of the signal elicits high response rates which are subsequently followed by an abrupt decline in responding. In contrast responding during unreinforced cycles is relatively low following reinforcement delivery and gradually increases reaching a peak during the third and fourth bin. In summary, at least with the longer signals some subjects display behavior that may be consistent with a superstition hypothesis (that is response rates are low during the signal because competing behavior may suppress lever pressing; in contrast, response rates remain high during unsignaled cycles because superstitious behavior is under control by the cue and thus does not disrupt lever–pressing behavior).

But how can signal effects be explained in the vast majority of the subjects where response rates remained more or less homogeneous throughout the reinforcement cycle? One possibility could derive from an analysis of obtained delays. Figure 4 suggests responding during the reinforcement cycle was continuous throughout the reinforcement cycle and thus obtained delays were considerably shorter than programmed delays. Comparably lower response rates produced in unsignaled conditions could probably favor longer obtained delays which would in turn further diminish response rate.

The differences in response rate between the unsignaled and briefly signaled conditions could also be understood in terms of a marking hypothesis. Lieberman, McIntosh & Thomas (1979) suggested brief signals may increase operant behavior under conditions of delayed reinforcement because the response that produces the signal is "marked" in the subject's memory and this "mnemic salience" makes it easier to associate the response with its consequence. In order to properly assess a marking hypothesis both responses during t^d and responses occurring outside t^d should have been followed by a signal (although the later should not produce food). In the absence of the appropriate control condition data from the present study can not confirm (or disconfirm) a marking account of the present data.

ACKNOWLEDGEMENTS

¹ This research was supported by the Facultad de Psicología and the Instituto de Posgrado e Investigación of the Universidad Intercontinental. The present authors wish to thank Christopher Podlesnik from Utah State University and Marc Branch from Florida University for their valuable comments on a previous draft of the present study. The authors also wish to thank Nuria Lanzagorta Rodrigo Sosa and Guillermo Martínez for their assistance in designing some of the figures and graphs presented in this paper. The authors are also in debt with the anonymous reviewers for their helpful comments. The study was conducted as a part of the seminary on operant behavior acquisition, directed by the first author and attended by the second and third author.

REFERENCES

Bruner, C., Avila, R. & Gallardo, L. (1994). Acquisition of lever pressing in rats under an intermittent Schedule of delayed reinforcement. Revista Mexicana de Análisis de la Conducta, 20, 119–129.         [ Links ]

Bruner, C., Pulido, M.. & Escobar, R. (1999). Response acquisition and maintenance with a temporally defined schedule of delayed reinforcement. Revista Mexicana de Análisis de la Conducta, 25, 379–391.         [ Links ]

Bruner, C., Pulido, M.. & Escobar, R. (2000). La adquisición del palanqueo con programas temporales de reforzamiento demorado. Revista Mexicana de Análisis de la Conducta, 26, 91–103.         [ Links ]

Critchfield, T.S. & Lattal, K.A. (1993). Acquisition of a spatially defined operant with delayed reinforcement. Journal of the Experimental Analysis of Behavior, 59, 373–387.         [ Links ]

Dickinson, A., Watt, A. & Griffiths, W.J.H. (1992). Free–operant acquisition with delayed reinforcement. The Quarterly Journal of Experimental Psychology, 45B, 241–258.         [ Links ]

Ellison, G.D. & Konorsky, J. (1964). Separation of the salivary and motor responses in instrumental conditioning. Science, 146, 1071–1072.         [ Links ]

Clark, F.C. (1958). The effect of deprivation and frequency of reinforcement on variable–interval responding. Journal of the Experimental Analysis of Behavior, 1, 221–228.         [ Links ]

Gibbon, J., Baldock, M.D., Locurto, C. Gold, L. & Terrace, H.S. (1977) Trial and inter–trial durations in autoshaping. Journal of Experimental Psychology: Animal Behavior processes. 3, 264–284.         [ Links ]

Kaplan, P.S. (1984). Importance of relative temporal parameters in trace autoshaping: From excitation to inhibition. Journal of Experimental Psychology: Animal Behavior processes, 10, 113–126.         [ Links ]

Lattal, K. A. (1987). The effect of delay and of intervening events on reinforcement value. in Commons, M., Mazur, J.E., Nevin, J.A. & Rachlin, H. (Eds.) Quantitative Analyses of Behavior. Vol 5. New Jersey: Lawrence, Erlbaum Associates Publisher.         [ Links ]

Lattal, K.A. & Gleeson, S. (1990). Response acquisition with delayed reinforcement. Journal of Experimental Psychology: Animal Behavior Processes, 16, 27–39.         [ Links ]

Lattal, K.A. & Metzger, B. (1994). Response acquisition by Siamese fighting fish. Journal of the Experimental Analysis of Behavior, 61, 35–44.         [ Links ]

Lieberman, D.A., McIntosh, D.C. & Thomas, G.V. (1979). Learning when reward is delayed: A marking hypothesis. Journal of Experimental Psychology: Animal Behavior processes, 5:224–242.         [ Links ]

Newlin, R.J. & Lolordo, V.M. (1976). A comparison of pecking by serial, delay and trace autoshaping procedures. Journal of the Experimental Analysis of Behavior, 25, 227–241.         [ Links ]

Pulido, M. Backer, C. & Rubí, M. (2003). Response acquisition with signaled delay of reinforcement: Effect of signal duration. Paper presented at the XXIX Annual Convention of the Association for Behavior Analysis.         [ Links ]

Pulido, M., Lanzagorta, N., & Lopez, L., (2005) Effects of contingent and non–contingent signals during delay interval on response acquisition by rats. Revista Mexicana de Análisis de la Conducta, 31, 227–244.         [ Links ]

Pulido, M & López, L. (2006). Delay of reinforcement effects under temporally defined schedules of reinforcement. Revista Mexicana de Análisis de la Conducta, 32, 39–53.         [ Links ]

Pulido, M, Paz, M & Sosa, R. (2008). The effects of behavioral history on response acquisition with delayed reinforcement: A parametric analysis. Sent for review to Revista Mexicana de Análisis de la Conducta, 34, 43–56.         [ Links ]

Richards, R.W. (1981). A comparison of signaled and unsignaled delay of reinforcement. Journal of the Experimental Analysis of Behavior, 35, 145–152.         [ Links ]

Schaal, D.W. & Branch, M.N. (1988). Responding of pigeons under variable–interval schedules of unsignaled, briefly signaled, and completely signaled delays to reinforcement. Journal of the Experimental Analysis of Behavior, 50, 33–54.         [ Links ]

Schaal, D.W. & Branch, M.N. (1990) Responding of pigeons under variable–interval schedules of signaled–delayed reinforcement: Effects of delay–signal duration. Journal of the Experimental Analysis of Behavior, 53, 103–121.         [ Links ]

Schaal, D.W., Schuh, K.J. & Branch, M.N. (1992). Key pecking of pigeons under variable interval schedules of briefly signaled delayed reinforcement: Effects of variable interval value. Journal of the Experimental Analysis of Behavior, 58, 277–286.         [ Links ]

Schaal, D.W., Odum, A.L. & Shahan, T.A. (2000). Pigeons may not remember the stimuli that reinforced their recent behavior. Journal of the Experimental Analysis of Behavior, 73, 125–139.         [ Links ]

Schlinger, H.D. & Blakely, E. (1994). The effects of delayed reinforcement and a response–produced auditory stimulus on the acquisition of operant behavior in rats. Psychological record, 44, 391–409.         [ Links ]

Schoenfeld, W.N., & Cole, B.K. (1972). Stimulus schedules: the t–t systems. New York, Harper and Row.         [ Links ]

Schoenfeld, W.N., Cole, B.K., Lang, J. & Mankoff, R. (1973). "Contingency" in behavior theory. In F.J. McGuigan and D.B. Lumsden (Eds.) contemporary approaches to conditioning and learning (151–172). Washington, DC: Winston and sons.         [ Links ]

Wilkenfield, J., Nickel, M., Blakely, E. & Poling, A. (1992). Acquisition of lever–press responding in rats with delayed reinforcement: A comparison of three procedures. Journal of the Experimental Analysis of Behavior, 58, 431–443.         [ Links ]

Williams, B. (1999). Associative competition in operant conditioning: Blocking the res–ponse–reinforcer association. Psychonomic Bulletin & Review, 6, 618–623.         [ Links ]

Van Haaren, F. (1992). Response acquisition with fixed and variable resetting delays of reinforcement in male and female Wistar rats. Physiology and Behavior, 552, 769–772.         [ Links ]