NMEOUT ANQ PUNQSHMEN? Q? A $8NGLE, fiESZP‘GNSE WHHEH A SHGR? FE VED RAEC $CHEDULE Q? FGQQ Ezfiifiimfifififfiflé? That-ts {’09 Hm Dag?“ o? M. A. MECHIGAN STATE UNIVERSITY Raiph Wifiliam Richards 3969 aTHESIS 5" LIBRARY—LI Mirh.‘ 5311 State University JR?) W 0 ABSTRACT TIMEOUT AND PUNISHMENT OF A SINGLE RESPONSE WITHIN A SHORT FIXED RATIO SCHEDULE OF FOOD REINFORCEMENT By Ralph William Richards The purpose of the present study was to determine if pigeons would take timeouts from a fixed ratio 50 when the twenty-fifth response within this ratio was punished. A secondary purpose was to examine the location of these timeouts within the ratio. Four pigeons were reinforced on one key according to a fixed ratio 50 until a stable performance was attained. Shock was then delivered immediately after the twenty-fifth response of each ratio. A response to a second key produced a thirty second timeout. For two subjects the intensity was initially set at 12.0 ma.; however, this intensity produced severe response suppression and was later reduced to 8.0 ms. For one subject timeouts occurred only during the first session of the 8.0 ma. shock; this was the only shock session (except for the first session at 12.0 ma.) during which any ratios were completed and reinforce- ment obtained. The other subject maintained responding to the food key and took many timeouts. Although the Ralph William Richards timeouts decreased to near zero on further exposure to the shock, the location of the timeouts within the ratio changed as a function of the number of sessions under the shock contingency. Timeouts occurred after the shock on its initial introduction; however, on further exposure they occurred more frequently before shock. Eventually, their most frequent location was during the pause-after-reinforcement or early in the ratio. Complete response suppression on the food key was suggested as a possible cause for the absence of timeouts for one of the preceding subjects; therefore, the method of shock introduction was altered for the two remaining subjects. For these subjects the intensity was gradually increased to either 12.0 ma. or 17.0 ma. Only one of these subjects took timeouts. The location of timeouts again showed a change on further exposure to the shock. Initially timeouts occurred after shock; later they occurred before shock, most frequently during the pause-after-reinforcement or early in the ratio. The present results while not definitive suggest that shook intensity, its manner of introduction, and the number of shocks administered may have a profound influence in determining whether pigeons will initiate Ralph William Richards timeouts in a situation where responses to one key are both punished and rewarded according to short fixed ratios. It was also suggested that a thirty second timeout may not be sufficient to maintain escape from Shock 0 Approved: ‘ ’1: Date: $0}; [:7é 6/“ TIMEOUT AND PUNISHMENT OF A SINGLE RESPONSE WITHIN A SHORT FIXED RATIO SCHEDULE OF FOOD REINFORCEMENT By Ralph William Richards A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Psychology 1969 To Nancy 11 ACKNOWLEDGMENTS I would like to express my sincere thanks to Dr. Mark E. Rilling without whose guidance and understanding this study would not have been completed. His suggestions and criticisms were also instrumental in the writing of the thesis. My thanks are also extended to Dr. M. Ray Denny and Dr. Ralph E. Levine for their helpful suggestions and criticisms of the manuscript. 111 TABLE OF CONTENTS List of Tables v List of Figures vi List of Appendices vii Introduction 1 Method 5 Results 10 Discussion 29 List of References 33 Appendix 35 iv LIST OF TABLES Table 1. Summary of Procedure Figure Figure Figure Figure Figure Figure Figure 1. 2. 3. 7. LIST OF FIGURES Percent of session spent in timeout as a function of shock intensity for 8-832 and S-660. Percent of session spent in timeout as a function of shock conditions for Percent of timeouts occurring before shock for 8-832. During session 84-91, no shocks were delivered. Location of timeouts within the fixed ratio 50 schedule for 3-832. Each block was computed from five sessions of shock. Block v is not shown. Percent of timeouts occurring before shock for S-1055. Location of timeouts within the fixed ratio 50 schedule for 3-1055. Each block was computed from five sessions of shock. Sample cumulative records for 8-832. vi 16 18 20 22 24 26 28 Appendix A LIST OF APPENDICES vii 35 INTRODUCTION Punishment is defined as a reduction in the future probability of a response on delivery of a stimulus imme- diately following the response. The stimulus following the response is, by definition, a punishing stimulus (Azrin & Holz, 1966). One of the most widely studied punishing stimuli has been electric shock. Although the immediate introduction of an intense shock after each response may result in complete suppression, responding has been maintained at extremely high shock intensities when the intensity is gradually increased (e.g. Azrin, Holz, & Hake, 1963). The effect of punishing each response is also partially dependent upon the schedule of reinforcement maintaining the response. When each response on a variable interval schedule is followed by shock, a stable, but reduced, response rate is maintained (Azrin, 1960). On a fixed ratio schedule, the effect is to selectively lengthen the pause-after-reinforcement, while leaving the local response rate unchanged (Azrin, 1959a). On a fixed inter- val schedule, a response reduction occurs in all portions of the interval, leaving the temporal discrimination unaffected (Azrin & Holz, 1961). Punishment of each response on a DRL schedule (differential reinforcement of low response rates) selectively reduces the shorter interresponse times (Holz, Azrin, & Ulrich, 1963). 1 Dardano & Sauerbrunn (1964), using pigeons, punished a single response within a fixed ratio 50 schedule of food reinforcement. Performance was differentially affected depending upon the position of the punished response within the ratio. If intense shock was delivered after the first response of the ratio, performance was characterized by a prolonged pause-after-reinforcement, breaks in responding after the intial responses, and unstable response rates at the beginning of the ratio. Punishment of the twenty-fifth response resulted in an unstable response rate in the first half of the schedule, with response rate in the second half being unaffected; the duration of the pause—after-reinforcement was found to be highly variable. Punishment of the fiftieth response had no such localized effect; rather, response breaks and changes in local response rates occurred within different portions of the ratio, and although the mean pause-after- reinforcement increased slightly, many of the ratios showed a short pause-after-reinforcement. Azrin, Hake, Holz, & Hutchinson (1965), in an attempt to determine if escape was motivated by punishment, condi- tioned pigeons to respond on one key under various schedules of food reinforcement. After responding on this key had become stable, shock was delivered after every response. A response on a second key removed the punish- ment contingency. Results showed that the subjects would escape from the punished to the unpunished situation and that this behavior could be brought under the control of both fixed interval and fixed ratio schedules of escape reinforcement. Self-imposed timeouts from positive reinforcement (i.e., a period of time during which positive reinforce— ment is not available have also been used to infer the existence of aversive stimuli (e.g., Azrin, 1961; Hearst & Sidman, 1961; Thompson, 196#; 1965). Azrin (1961) intro- duced a timeout key when pigeons were being reinforced under a fixed ratio schedule on another key. The first response to the timeout key put an extinction period, with changes in overhead illumination and key colors, into effect; a second response restored the original illumination, key colors, and the possibility of rein- forcement. Thus, the subjects wereable to initiate and terminate extinction periods. These self-imposed timeouts were usually located within the characteristic pause- after-reinforcement. Further, the amount of time spent in timeout was an increasing monotonic function of the number of responses required for food reinforcement. Azrin suggests the existence of aversive stimuli at certain stages within the fixed ratio schedule as a plausible explanation of the escape from a schedule of food reinforcement. Several other studies (Appel, 1963; Thompson, 196#; 1965; Zimmerman & Ferster, 1965) have also shown that 4 food reinforcement is delivered on a fixed ratio schedule. Hearst & Sidman (1961) employed a fixed duration time- out to determine if escape from a conflict situation was reinforcing. Rats were reinforced (with a mixture of sweetened milk and tap water) on a variable interval sched- ule for pressing one lever in the presence of a discrim- inative stimulus. Concurrently, responses on this lever were punished according to a fixed ratio schedule. Responses to a second lever terminated the discriminative stimulus and put a period of extinction into effect. Results suggested that rats will escape from a situation where responses are concurrently rewarded and punished. The primary purpose of the present study was to deter- mine if pigeons would escape from a fixed ratio 50 sched- ule of reinforcement when the twenty-fifth response within this ratio was punished. The escape response was peck- ing a second key which produced a 30 second timeout. If timeouts did occur, a secondary purpose was to examine their location within the ratio. METHOD Subjects Four experimentally naive female White Carneaux pigeons were used. All subjects were between 3 and 5 years of age and were maintained at approximately 75% of their free-feeding weight throughout the present experiment. Apparatus A standard operant conditioning chamber (Lehigh Valley Electronics, model 1519) with a three stimulus key intelli- gence panel was employed. The right and left keys were illuminated green and red, respectively. The center key was not illuminated and responses to it were not record- ed. A minimum force of approximately 20 grams was necessary to record a response to either key. During reinforcement, both the key lights were extinguished and a light within the food aperture was illuminated. Reinforce- ment was 3 second access to mixed grain, with the timing of this interval beginning when the subject placed its head through the aperture, thus interrupting a light beam focused on a photocell. An overhead light provided general illumination, except during periods of timeout. The shock source was of the constant current design. Line current was connected to the primary of a 460 volt AC step-up transformer, the secondary of which was connected to a series circuit consisting of two 12 K ohm 5 fixed resistors, the bird, a milliammeter, and a 250 K ohm variable resistor. In calibrating shock intensity, an 800 ohm resistor was substituted for the bird. Shock duration was .Oh5 seconds and was delivered through electrodes implanted in the subjects' back (Azrin, 1959b). Reinforcement and punishment contingencies were programmed by standard electro-mechanical equipment in an adjacent room. A cumulative recorder and three-channel printout counter were also used. Procedure Each subject had one session a day, except on holidays and as necessary to maintain deprivation level. After the initial shaping, sessions usually terminated on the delivery of 50 reinforcements or six hours, whichever occurred first. The chamber was completely dark for a variable duration, before and after each session. All subjects were magazine trained and shaped to peck the left key (red) within two sessions. The shaping session terminated on the delivery of 50 reinforcements under a schedule of continuous reinforcement. On all days thereafter, a response on the right key (green) produced a 30 second timeout. During timeouts, the chamber was completely dark, and all responses were nonfunctional. After a second session of continuous reinforcement, the response requirement for food rein- forcement was gradually increased until an FR 50 (fixed ratio 50) schedule was established. Subjects were given a minimum of 25 sessions on the FR 50 without shook. Shock was then delivered immediatedly following the twenty-fifth response of each ratio. The shock intensities used and the number of sessions at each are shown in Table 1 for each subject. For 5-832, the intensity was initially set at 12.0 ma. with sessions terminating at the end of four hours or 50 reinforcements, whichever occurred first. In an attempt to increase responding to the food key, maximum session length was increased to seven hours for one session, and set at six hours for all days thereafter. In a further attempt to increase responding, the intensity was reduced to 8.0 ma. after 12 sessions under the 12.0 ma. shock. On the completion of 400 ratios at 8.0 ma., the shock contingency was removed and 8 sessions of FR 50 without shock programmed. Then, the 12.0 ma. shock was reintroduced following the twenty-fifth response until 400 ratios were completed. For 3-660, shock intensity was set at 12.0 ma. for 22 sessions. As responding to the food key was completely suppressed at this intensity, it was necessary to reshape the key peck response and deliver food reinforcement under the FR 50 without shock for 16 sessions. Shock was then presented for 9 sessions at an 8.0 ma. intensity. In a further attempt to prevent complete suppression of responding, a different procedure for introducing the shock was employed for S-hAO and 801055. Shock intensity was initially set at 2.0 ma. and increased in 2.0 ma. steps following the completion of three ratios at that intensity until the 12.0 ma. intensity was reached. For S-440 the 12.0 ma. intensity remained in effect for 25 sessions. For 8-1055 the intensity was further increased to 15.0 ma. and 17.0 ma. on the following two days, and was maintained at 17.0 ma. for 25 sessions. For s-uuo and S-1055 an apparatus failure occurred during two sessions of the FR 50 without shock. During these two sessions, the food magazine did not raise up completely, thus permitting only restricted access to the mixed grain reinforcement. Throughout all of the above, the number of timeouts and their location within the ratio were recorded. Table 1 Summary of Procedure SUBJECT CONDITION NUMBER OF SESSIONS S-832 no shock #0 sessions 12 ma. 12 sessions 8 ma.* 16 sessions no shock 8 sessions 12 ma.* 10 sessions 8-660 no shock 40 sessions 12 ma. 22 sessions no shock 16 sessions 8 ma. 9 sessions 8-440 no shock 25 sessions 2-912 ma.** 1 session 12 ma. 25 sessions 3—1055 no shock 25 sessions 2-912 ma.** 1 session 15 ma. 1 session 17 ma. 25 sessions *to the completion of 400 ratios **intensity increased in steps of 2 ma. RESULTS Figures 1 and 2 show the percent of each session spent in timeout for 3-832, 3-660, and S-1055. s-uuo took no timeouts during the 25 sessions of the 12.0 ma. shock. It is clear from these figures that only S-832 and 3-1055 spent any appreciable time in timeout. For S-832, on the initial introduction of the 12.0 ms. shock (during session 56), timeouts increased. However, on further exposure to this intensity timeouts decreased; during sessions 65-67, timeouts accounted for only about 8% of the sessions' length. On the reduction of intensity to 8.0 ma. (during session 68), timeouts increased and later decreased, until less than 1% of the sessions' length was spent in timeout. On the removal of shock (sessions 8h-91), no timeouts occurred except during session 84. It should be noted that the timeouts during session 84 occurred before the completion of any ratios without shock; after the completion of one ratio, no timeouts were taken. On the reintroduction of the 12.0 ma. shock (during session 92) timeouts increased, although to a lower level than had previously occurred at either the 12.0 ma. or the 8.0 ma. intensity. Again, further exposure to the shock resulted in decreased timeouts. Although the decrease in timeouts at 8.0 ma. and the smaller number of timeouts and their subsequent decline on the reintroduction of the 12.0 ms. intensity may have 10 11 been a recovery phenomenon, such does not appear to be the case during the initial presentation of the 12.0 ma. shock. Rather, here the decrease in timeouts was accom- panied by severe response reduction (see appendix A), and further, timeouts did increase on the reduction of the intensity to 8.0 ma. This decrease in responding to both the food and time- out key may explain the absence of timeouts for 8-660. The lower portion of figure 1 shows that S-66O took few timeouts, except during the first session of the 8.0 ms. intensity, during which 19.7% of the session was spent in timeout. With the exception of the first session at 12.0 ma., this was the only session during which any ratios were completed and reinforcement obtained (see appendix A). ' As the complete suppression of responding on the food key was suggested as a possible cause for the absence of timeouts for S-660, the method of shock introduction was altered for S-hho and S-1055. For these subjects, the intensity was gradually increased as previously described. s-uuo took no timeouts during the 25 sessions of the 12.0 ma. shock, and although the responding was reduced, it was not completely suppressed. 3-1055 did initiate some time- outs as shown in figure 2, with the greatest amount of any shock session spent in timeout being 19.3$ of the total session length. As previously mentioned, an apparatus failure occurred 12 during two sessions of the FR 50 without shock for s-uuo and S-1055. The effect of this restricted access to reinforcement is shown in figure 2 (sessions 15 and 16) for S-1055. A slight increase in timeouts occurred for S-th (see appendix A), while for S-1055 the increase is readily apparent, accounting for 37.3% and 21.8% of the session length. Session length also increased for both subjects. On repair of this failure, timeouts and session length rapidly declined. location of timeouts within.£hg‘£gtig S-832 took 39 timeouts during the 40 sessions of FR 50 without shock. All of these timeouts occurred during the pause-after-reinforcement. As can be seen in figure 3, the initial introduction of shock resulted in most timeouts occurring after the shock. However, on further exposure to shock, timeouts occurred more frequently before shock, than after shock. Only timeouts from completed ratios were used in forming figures 3-6. If no ratios were completed or less than three timeouts taken, no point was plotted for that session in figures 3 or 5. Figure # is a series of histograms showing the location of timeouts within the completed ratios. Each histogram was computed from five sessions of shock. Part I of this figure shows that on the initial exposure to shock most timeouts occurred immediately after the shock. On further exposure, timeouts 13 began to occur frequently at all locations within the ratio and the percent occurring immediately after the shock decreased (see part III). Eventually, timeouts rarely occurred after shock and their most frequent location was during the pause-after-reinforcement or early in the ratio run (see parts VI and VIII). S-1055 took 189 timeouts during the sessions of FR 50 without shock (except sessions 15 and 16); 1%? of these were during the pause-after-reinforcement. For the last 20 sessions on FR 50 without shock (except sessions 15 and 16), 123 of the 129 timeouts occurred during the pause- after-reinforcement. Figure 5 shows that most timeouts occurred after shock on the shock's initial introduction. On further exposure to the shock, timeouts occurred more frequently before shock, except during sessions h3-h6. Figure 6 is a series of histograms representing the per- cent of timeouts at each location within the ratio. Again, each histogram was computed from five sessions of shock. Although the initial introduction of shock did not produce many timeouts after shock nor a wide spread in timeout locations, it is evident that during block V timeouts occurred early in the ratio. Portions of the cumulative records for S-832 are shown in figure 7. The upper pen returned to baseline on the delivery of reinforcement, and the ”pips" represent the delivery of shock; timeouts are indicated by deviations from the lower line. The record for day 5b is one of the 1“ last sessions without shock and is typical of short fixed ratio performance (Forster & Skinner, 1957). Day 59 shows disrupted responding throughout the ratios with timeouts occurring after the shock. Days 6h and 69 also show dis- rupted responding throughout the ratio, but with timeouts occurring at all locations within the ratio. By day 7“ timeouts were occurring mainly before the shock. Day 80 shows what appears to be the acquisition of differential responding, i.e. broken and unstable response rates before shock and high stable response rates after shock. Time- outs have decreased to near zero (see figure 1). Days 93 and 98 clearly show differential responding within the ratios; this differential responding is in agreement with that found by Dardano and Sauerbrunn (196M). On day 93 timeouts occurred in ”groups” after the rapid completion of several ratios. Cumulative records for s-uuo and 8-1055 showed similar differential responding during the two halves of the ratio.. 15 Figure 1. Percent of session spent in timeout as a function of shock intensity for S-832 and S-660. 16 mzommmm op. mm Lbs be me 8 mm . . .1 air; A u m -9 _ n " ION A _ _ . 68w " n " -mw .oZm ".520" .02 N_ m .520 m8 ngmmmm 8._ mm pm mm ob ms ow me 8 mm . u A “d u" a d u — d .1 #JTIO _ . n u _ n [O— . _ u " ION _ mam-m_ H u m u -00 _ . IO? .0: N: .020. dam “ .02 N. " 620:8 _ .OE .LN3dS NOlSSBS JO lNBOHBd iflOBWll Nl 17 Figure 2. Percent of session spent in timeout as a function of shock conditions for S-1055. 18 N OE mzo_mmwm mm Om me. 0? mm. Om. mm ON 9 0. d u H .0 mm H u -0. mm . _ ION ll " u -on Nb. is _ u -8 we .th . . .OEO IOm Much-m HO iN @872 19 Figure 3. Percent of timeouts occurring before shock for S-832. During sessions 84-91, no shocks were delivered. o 92 20 S -832 FIG. 3 0000080000 mCDNLOLo rON“ XDOHS 380338 9NltltlflCDOO SiflOI—JNIJ. 50 iNBOHEd ; 80 l 7'5 SESSIONS 9'5 71% 6'5 21 Figure h. Location of timeouts within the fixed ratio 50 schedule for S-832. Each block was computed from five sessions of shock. Block V is not shown. 9.2m z_I.:>> ZOE-4004 22 Viol tnvci EZZII tv EZZII t-V EEZZII asses ass-mus ego-£3 .V .II II- 0-L I0 .0 om.- Iom m oeJ low M Owl I00 I. O 00.1 I00 .3 81 HE S E 18 u I Io w o I. — all... .- m 0N] ION mu owl low owl loo owl Iom Nan/mum 23 Figure 5. Percent of timeouts occurring before shock for S-1055. 21+ LO LO £2 I U) 1) Ln -. “ho ”0 L0 m \ Z ' 9. m "to at”.- m __o q. __LO ro \\ IIIIFIIIIII O O o o o oo o o o o 9 m 00 w LO Loe- r) (\J - MDOHS 380.138 SNIHHHDOO SiflOEIWIJ. JO .LNBOHBd FIG. 5 25 Figure 6. Location of timeouts within the fixed ratio 50 schedule for 3-1055. Each block was computed from five sessions of shock. 26 0. Ham 2. It 3 20. F4004 0 .0E has. .0. c. _o. with WWfiflRWGMGt 00 l loo owl ION owl low ool lo o owl Io m 00. I H I I09 00 I ON I owl SI m mo_ m I o o owl lo 0 oo_l m H 00. SLOOEIINIJ. :IO .LNEIOHBd 27 Figure 7. Sample cumulative records for S-832. wig-\lmfiéliflfijslsxlé-Il , lard-Fl? If??? lYIKlI: v KIT. T. NI S all N IRRIISII .1 I T. l sIH IlllllllllkiilfillllllII IKIIIH III Illkx an >40 - - .222 o. m no dig; 5, cm lag DISCUSSION The purpose of the present study was to determine what effect shocking the twenty-fifth response within a fixed ratio 50 schedule of reinforcement had on time- out behavior in pigeons. Results should be considered exploratory and suggestive, rather than definitive. Except for S-1055 during sessions 8-16, few timeouts were taken from the fixed ratio 50 schedule without shock. The timeouts were usually located within the character- 'istic pause-after-reinforcement, which agrees with the previous findings of Azrin (1961) and Thompson (196a; 1965). It appears that under some conditions pigeons will take timeouts from shock. Variables of importance may include shock intensity and its degree of accompanying response suppression. with responding almost completely suppressed, 3-660 took few timeouts. For 5-832 responding was maintained and many timeouts taken. Unfortunately, the percent of session spent in timeout varied and even- tually decreased to zero. Attempts to achieve stability by changing shock intensity were unsuccessful. Although the present study gives no definitive explanation for this decrease in timeouts or its absence in some subjects, several possibilities are suggested. As responding has been shown to result in a characteristic recovery from mild punishment (e.g., Azrin, 1960; 29 30 Hake & Azrin, 1963), the decrease in timeouts may simply be a recovery phenomenon. In conjunction with this, the constant location of the shock may have allowed it to acquire discriminative properties (Holz & Azrin, 1962); support for this may be seen in figure 7 and in the Dardano & Sauerbrunn study (196“). Further, timeouts did not only decrease on further exposure to shock, but their location within the ratio also changed. While initially timeouts occurred after the shock, they eventually occurred only before the shock and most often early in the ratio (figures 3 and 4). Although this change is less clear for 8-1055, the timeouts did eventually occur most often early in the ratio (figures 5 and 6). Following Hearst & Sidman (1961), the delivery of both food and shock under either a variable ratio or variable interval schedule might have been more effective in maintaining a conflict situation and possibly timeout behavior. A third possible explanation for the decrease in timeouts and their relative fewness concerns the dura- tion of the timeout. Zimmerman & Forster (196“) found timeout duration to be an important factor in timeout from simple fixed ratio schedules. A fixed duration 10 second timeout resulted in fewer timeouts than a "free- switching” procedure. Possibly a ”free-switching” procedure, which allows the subject to both initiate and terminate timeouts, would have been more effective in the present study. If a thirty second timeout has been shown 31 effective in maintaining escape from a simple fixed ratio (e.g., Thompson, 1965), it does not necessarily follow that this duration will be sufficient to maintain escape re- sponding when shock is added. In this regard, it should be mentioned that Hearst & Sidman (1961) employed a fixed duration timeout of either 10 or 15 minutes. More recent research (Hearst & Koresko, 1964; Hearst, 1967) has employed the “free-switching" procedure. Gross obser- vations in the present experiment revealed that the subjects when not responding would often turn away from the intelli- gence panel and face the backwall. This may suggest that the timeout duration was not sufficient to maintain escape responding or that the locating of the timeout key on the backwall would have generated more escape. Another important variable may be the manner in which the shock is introduced. For 3-1055 and 8-440, the inten- sity was gradually increased and although this proved successful in maintaining responses to the food key, it did not result in timeouts accounting for much of the sessions' length. Examination of the location of timeouts within the ratio for S-1055 showed timeouts occurring with- in the early portions of the ratio, after the initial exposure to shock. While Hearst gt_§l have been able to manipulate shock intensity in maintaining timeouts with rats, the present experimenter was unable to duplicate this with pigeons. Several procedural differences that might account for this were suggested above. 32 In summary, results suggest that the shocking of the twenty-fifth response within a fixed ratio 50 schedule of food reinforcement may result in pigeons taking time- outs from positive reinforcement, although the results are not clear as to the important variables underlying this behavior. Subjects which did take timeouts showed a tendency to take them early in the ratio on further exposure to shock. LIST OF REFERENCES LIST OF REFERENCES Appel, J.B. “Aversive aspects of a schedule of positive re- inforcement. Journal of the Experimental Analysis of Behavior, 1963, §, 5233528. Azrin, N.H. Punishment and recovery during the fixed ratio performance. Journal of the Experimental Analysis of Behavior, 1959a, Z, 3UI-355. Azrin, N.H. A technique for delivering shock to pigeons. Journal of the Experimental Analysis of Behavior, - g g "' e Azrin, N.H. Effects of punishment intensity during variable interval reinforcement. .Journal of the Experimental Analysis of Behavior, 1965, 3, 1233132. Azrin, N.H. Timeout from positive reinforcement. Science, 1961 9 1,229 382‘3830 Azrin, N.H., Hake, D.F., Holz, W.C., and Hutchinson, R.R. Motivational aspects of escape from punishment. Journal of the Experimental Analysis of Behavior, I§6;, a, 31-“. Azrin, N.H. and Hake, W.C. Punishment during fixed-interval reinforcement. Journal of the Experimental Analysis of Behavior, 1961. 5. 333:357} Azrin, N.H. and H012, W.C. Punishment. In W.K. Honig (Ed.). 0 erant behavior: areas of research and a lication. New York: Appleton- en ury- re s, , p. - 7. Azrin, N.H.. Hols, W.C., and Hake, D.F. Fixed-ratio punish- megt. 6Journal of the Experimental Analysis of Behavior, 19 39 _9 " s Dardano, J.F. and Sauerbrunn, D. Selective punishment of fixed-ratio performance. Journal of the Experimental Analysis of Behavior, 1964, 7, 253-265. Forster, 0.3. and Skinner, B.F. Schedules of reinforcement. New York: Appleton-Century-Croffs, I937. 33 34 Hearst, E. Oscillatory behavior during approach-avoidance conflict. Journal of the Experimental Analysis of Behavior, 1 , __, - . Hearst, E. and Koresko, Minnie B. Self-presentation and self-termination of a conflict-producing stimulus. Science, 1964, 346, 415-416. Hearst, E. and Sidman, M. Some behavioral effects of a concurrently positive and negative stimulus. Journal of the Experimental Analysis of Behavior, 9 9 " 0 Hols, W.C. and Azrin, N.H. Discriminative properties of punishment. Journal of the Experimental Analysis at BOTH-7101‘ , , ’ ‘l' 2 e Hols, W.C., Azrin, N.H., and Ulrich, R.E. Punishment of temporally spaced responding. Journal of the Ex er- imental Analysis of Behavior, 1963,‘§, ITS-122. Thompson, D.N. Escape from SD associated with fixed-ratio reinforcement. Journal of the Experimental Analysis of Behavior, 1964, z, 1-8. Thompson, D.N. Timeout from fixed-ratio reinforcement: a systematic replication. Psychonomic Science, 1965, 3, 109-110. Zimmerman, J. and Forster, C.B. Some notes on timeout from reinforcement. Journal of the Experimental Analysis of BehavIor, 1964, 1, 13-19. APPENDIX A APPENDIX A 3-832 shock session percent in session intensity responses length timeout 16 0.0 ms. 2500 16.92 min. 00.0 17 0.0 2500 18.61 00.0 18 0.0 2500 18.75 00.0 19 0.0 2500 19.67 00.0 20 0.0 2500 58.77 00.0 21 0.0 2500 18.37 00.0 22 0.0 2500 17.81 00.0 2 0.0 2500 20.56 00.0 2 0.0 2500 17.92 00.0 25 0.0 2500 20.20 00.0 26 0.0 2500 18.41 00.0 27 0.0 2500 18.77 00.0 28 0.0 2500 19.19 00.0 29 0.0 2500 19.14 00.0 30 0.0 2500 18.56 00.0 31 0.0 2500 19.19 00.0 32 0.0 2500 33.20 16.6 33 0.0 2500 31.89 20.4 34 0.0 2500 18.62 08.1 35 0.0 2500 25.77 11.6 36 0.0 2500 17.57 02.8 37 0.0 2500 19.44 05.1 38 0.0 2500 16.37 00.0 39 0.0 2500 15.36 00.0 40 0.0 2500 15.21 00.0 41 0.0 2500 15.84 00.0 42 0.0 2500 16.53 00.0 43 0.0 2500 16.27 00.0 44 0.0 2500 18.90 00.0 45 0.0 2500 17.30 00.0 46 0.0 2500 16.94 00.0 47 0.0 2500 17.98 00.0 48 0.0 2500 18.79 02.7 49 0.0 2500 16.10 00.0 50 0.0 2500 16.08 00.0 51 0.0 2500 16.84 00.0 52 0.0 2500 17.48 02.9 53 0.0 2500 17.29 02.9 54 0.0 2500 17.36 00.0 55 0.0 2599______§%;,%g 00.0 36 1275' 175 . 00.2 57 12.0 175 240.87 07.1 35 36 S-832 shock session percent in session intensity responses length timeout 58 12.0 83 234.79 12.1 59 12.0 175 240.29 11.9 60 12.0 114 240.25 18.5 61 12.0 103 240.22 22.1 62 12.0 193 240.19 23.3 63 12.0 25 240.54 20.6 64 12.0 354 420.28 40.2 65 12.0 0 360.81 08.0 66 12.0 177 359.62 08.9 6? 12.0 0 360.00 07.5 68 8.0 232 . . 69 8.0 500 359.84 28.1 70 8.0 157 360.43 21.4 71 8.0 1075 359.04 12.7 72 8.0 707 359.87 09.7 7 8.0 844 360.36 19.3 7 8.0 1394 259.12 19. 75 8.0 1319 359.93 30. 76 8.0 2266 359.44 05.3 77 8.0 213 362.89 00.6 78 8.0 656 362.13 00.8 79 8.0 2500 219.56 03.0 80 8.0 2500 317.66 00.5 81 8.0 2500 311.23 00.3 82 8.0 1300 259.55 00.7 832 _58,9 2 00 40.11 00.4 84 0.0 .46 . 85 0.0 2500 19.68 00.0 86 0.0 2500 28.48 00.0 87 0.0 2500 21.90 00.0 88 0.0 2500 19.33 00.0 89 0.0 2500 21.20 00.0 90 0.0 2500 19.16 00.0 _21 0.0 2 00 18.34 00.0 92 712.0 II 361. 2 03.9 93 12.0 1403 361.06 12.3 94 12.0 2500 236.37 12.5 95 12.0 2408 359.50 10.6 96 12.0 1950 360.23 00.0 97 12.0 2500 293.26 00.0 98 12.0 2500 350.60 00.3 99 12.0 2500 299.30 00.0 100 12.0 2500 182.02 00.0 101 12.0 1400 110.34 00.0 37 S-660 shock session percent in session intensity responses lengph timeout 16 0.0 ma. 2500 34.86 min. 00.0_ 17 0.0 2500 34.68 00.0 18 0.0 2500 30.02 00.0 19 0.0. 2500 26.36 00.0 20 0.0 2500 21.96 00.0 21 0.0 2500 20.00 00.0 22 0.0 2500 22.84 00.0 23 0.0 2500 32.26 00.0 24 0.0 2500 20.94 04.7 25 0.0 2500 28.20 05.3 26 0.0 2500 36.79 00.0 27 0.0 2500 28.18 00.0 28 0.0 2500 22.39 00.0 29 0.0 2500 20.51 00.0 30 0.0 2500 23.36 00.0 31 0.0 2500 22.27 00.0 32 0.0 2500 19.54 00.0 33 0.0 2500 16.79 00.0 34 0.0 2500 16.15 00.0 35 0.0 2500 16.77 00.0 36 0.0 2500 18.66 00.0 37 0.0 2500 19.06 00.0 38 0.0 2500 25.53 00.0 39 0.0 2500 20.69 00.0 40 0.0 2500 19.05 00.0 41 0.0 2500 15.89 00.0 42 0.0 2500 13.82 00.0 43 0.0 2500 1 .24 00.0 44 0.0 2500 1 .05 00.0 45 0.0 2500 17.64 00.0 46 0.0 2500 14.13 00.0 47 0.0 2500 15.56 00-0 48 0.0 2500 13.52 00.0 “9 000 2500 13075 00' 50 0.0 2500 13.94 00.0 51 0.0 2500 13.92 00.0 52 0.0 2500 15.23 00.0 53 0.0 2500 14.57 00.0 54 0.0 2500 14.89 00.0 55 0.0 2500 14.16 00.0 56 12.U* 375 361.97 v1.2 57 12.0 28 363.21 00.6 58 12.0 2 363. 1 00.3 59 12.0 27 363.43 02.2 60 12.0 0 363.51 00. 38 S-66O shock session percent in session intensity responses length timeout 12.0 0 363.48 00.1 2% 12.0 0 363.47 00.0 63 12.0 0 363.43 01.9 64 12.0 1 363.42 00.1 65 12.0 0 363.46 01.9 67 12.0 0 363.58 01.5 68 12.0 0 363.60 01.7 69 12,0 0 363.60 01.4 70 12.0 0 363.60 00.0 71 12.0 0 363.65 00.0 72 12.0 0 363.65 00.4 73 12.0 0 366.01 00.3 74 12.0 1 365.79 00.3 75 12.0 0 365.96 00.0 76 12.0 0 366.05 00.5 77 12.0 0 366.19 00.8 78 0.0 79 0.0 2500 21.49 00.0 80 0.0 2500 23.27 00.0 81 0.0 2500 22.00 00.0 82 0.0 2500 32.27 00.0 83 0.0 2500 21.48 00.0 84 0.0 2500 18.66 00.0 85 0.0 2500 18.95 00.0 86 0.0 2500 32.13 00.0 87 0.0 2500 16.09 00.0 33 0.0 2500 16.92 00.0 89 0.0 2500 16.75 00.0 90 0.0 2500 14.52 00.0 91 0.0 2500 14.52 00.0 92 0.0 2500 13.08 00.0 93 0.0 2500 13.01 00.0 9a *8.0‘ 104 365.90 *19.7 95 8.0 0 366.63 00.1 96 8.0 25 366.60 03.1 97 8.0 0 366.35 00.4 98 8.0 0 366.27 00.4 99 8.0 0 366.63 00.1 100 8.0 3 366.30 01.6 101 8.0 1 358.20 00.4 102 8.0 0 366.28 01.2 39 S-440 , shock session percent in session intensity responses length timeout 8 0.0 ma. 2500 123.29 min. 00.0 9 0.0 2500 43.48 00.0 10 0.0 2500 41.61 00.0 11 0.0 2500 67.50 00.0 12 0.0 2500 40.01 00.0 1 0.0 2500 40.68 00.0 1 0.0 2500 24.00 00.0 15 0.0 2500 5.40 00.0 16* 0.0 2500 130.15 '00.5 17* 0.0 2500 1 9.65 00.0 18 0.0 2500 33.43 00.0 19 0.0 2500 27.41 00.0 20 0.0 2500 26.89 00.0 21 0.0 2500 32.86 00.0 22 0.0 2500 28.37 00.0 23 0.0 2500 26.91 00.0 24 0.0 2500 26.53 00.0 25 0.0 2500 27.44 00.0 26 0.0 2500 27.52 00.0 27 0.0 2500 37.00 01.4 28 0.0 2500 30.68 00.0 29 0.0 2500 25.75 00.0 30 0.0 2500 27.47 00.0 31 0.0 2500 30.00 00.0 2 0.0 2500 31.17 00.0 :§2"'”"270351270 1000 360.09 ‘00.0 12.0 140 . 00.0 35 12.0 407 361.07 00.0 36 12.0 2500 280.68 00.0 37 12.0 1875 358.81 00.0 38 12.0 931 360.34 00.0 39 12.0 2050 358.52 00.0 40 12.0 351 361.18 00.0 41 12.0 75 361.20 00.0 42 12.0 '52 361.00 00.0 43 12.0 1247 359.60 00.0 44 12.0 1500 359.25 -00.0 45 12.0 325 361.28 00.0 46 12.0 2500 343.00 00.0 47 12.0 1615 359.12 00.0 48 12.0 2500 302.98 00.0 * apparatus failure session shock intensity 12.0 12.0 12.0 12.0 12.0 12.0 12.0 12.0 12.0 12.0 40 S-440 responses 2500 2500 2500 1301 1556 2025 2100 70 1935 2013 length 278.98 292.61 249.61 360.04 359.66 358.90 358.74 360.00 357.45 359.14 session percent in timeout 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 41 S-1055 shock session percent in session intensity responses length timeout 7 0.0 m. 2500 115.50 min. 0009 8 0.0 2500 126.75 16.6 9 0.0 2500 48.84 11.3 10 0.0 2500 41.51 01.2 11 0.0 2500 60.30 03.3 12 0.0 2500 109.07 13.8 1 7 0.0 2500 118. 71 16.0 15* 0.0 1805 360. 00 37 2 16* 0.0 2500 105. 47 21 8 17 0.0 2500 34.15 A 07 3 18 0.0 2500 29.70 05 1 19 0.0 2500 28.66 01.7 20 0.0 2500 29.68 01.7 21 0.0 2500 29.70 05.1 22 0.0 2500 31.07 03.2 23 0.0 2500 31.49 03 2 24 0.0 2500 26.84 00 0 25 0.0 2500 27.58 00 0 26 0.0 2500 27.21 00 0 27 0.0 2500 31. 44 00 0 28 0.0 2500 30.57 00.0 29 0.0 2500 28. 48 00.0 30 0.0 2500 28.45 00.0 31 2 0.0 2500 36.41 02.? "“32 2 .412.0 2590 “325.02 02(3"’ 1 .0 23° . '00:? :::;;g 17.0 2210 . A . 35 17.0 2500 51.90 00.0 36 17.0 2500 284. 0 00.5 37 17.0 2500 295. 9 08 5 38 17.0 1809 354. 44 01. 4 39 17.0 0 361. 72 07.0 40 17.0 2500 209 76 19 3 41 17,0 150 360. 00 06.4 42 17.0 2500 224.84 02. 9 43 17.0 2500 67.91 04. 4 44 17.0 2250 357.69 04.5 45 17,0 1250 359.26 0 .1 46 17.0 2500 313.58 0 .9 47 17.0 2050 359.54 01.8 * apparatus failure 42 S-1055 shock session percent in session intensity responses length timeout 48 17.0 2353 325.56 03.1 49 17.0 2500 355.32 ‘ 00.7 50 17.0 2500 222.92 00.0 51 17.0 2500 177.61 00.3 52 17.0 2450 155.41 00.0 5 17.0 2500 203.74 05.6 5 17.0 .2500 285.63 02.8 55 17.0 2802 354.44 02.7 56 17.0 2500 263.88 01.7 57 17.0 17 361.77 03.2 58 17.0 2500 126.87 00.8 Au'.