on. aw“. '6... THESIS LIBRARY Michigan State Univcmty ROOM USE ONLY ABSTRACT THE PARTIAL REINFORCEMENT EFFECT IN A WITHIN-SUBJECTS DESIGN by Gale A. Gordon This study constituted the first step toward investigating Weinstock's hypothesis of habituation to frustration for explaining the partial reinforce- ment effect. According to Weinstock, §s under partial reinforcement conditions compared with those under continuous reinforcement conditions show greater resistance to extinction because frustration has under- gone habituation in such animals and is less effective for eliciting competing responses during extinction. The subjects were 30 laboratory rats: 10 albino and 20 Tryon. The apparatus consisted of a pair of enclosed straight alleys, each made up of a runway and a goal box. The interior of one runway and goal box was painted black; the other, flat white. Each.§ received 32 acquisition trials in one of the alleys with a continuous reinforcement schedule and received 32 acquisition trials in the other alley with a 50% partial reinforcement schedule. At the end of the acquisition trials, the §s were matched for total Gale A. Gordon running time, and one member of each pair was arbitrar- ily assigned to one of two extinction procedures. Half of the §s were extinguished in the alley where before they had experienced only continuous reinforcement. The other half of the §s were extinguished in the alley where before they had eXperienced only partial rein- forcement. The hypothesis that those subjects extinguished in the partially reinforced alley would be more re- sistant to extinction than those §s extinguished in the continuously reinforced alley was not confirmed, although there was a definite trend in the predicted direction. The discussion centered around the rela- tionship of frustration and habituation to frustration to the observation and a possible explanation for the failure to confirm the hypothesis. Suggestions were given for fruther research. Approved: ”fl/ %M Date: zf/ ’/E7‘5F' / THE PARTIAL REINFORCEMENT EFFECT IN A WITHIN-SUBJECTS DESIGN BY Gale A. Gordon A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Psychology 1965 ACKNOWLEDGEMENTS The author wishes to express her gratitude to Dr. M. Ray Denny for serving as chairman of the thesis committee and for his support and encouragement during the last two years. The author also wishes to extend her appreciation to Drs. Stanley Ratner and Richard Hart for serving on the thesis committee. Special thanks goes to Bob Boice who designed and built the apparatus. The author is indebted to Lee Taylor who gave unselfishly of his time and talents when it was most needed, especially in the final preparation of the thesis. To my parents, my heartfelt thanks for their en- couragement and moral support. It is to them that this work is dedicated. 11 TABLE OF CONTENTS Page ACKNOWLEDGEMENTS . . . . . . . . . . . 11 LIST OF TABLES . . . . . . . . . . . . iv LIST OF FIGURES . . . . . . . . . . . v LIST OF APPENDICES . . . . . . . . . . vi INTRODUCTION. . . . . . . . . . . . . 1 METHOD. . . . . . . . . . . . . . . 8 Subjects Apparatus Procedure RESULTS . . . . . . . . . . . . . . 13 DISCUSSION AND CONCLUSIONS . . . . . . . . 20 BIBLIOGRAPHY . . . . . . . . . . . . 24 APPENDICES O O O O O O O O O O O O O 26 iii Table 2. 3. LIST OF TABLES Mean total running time in seconds for pairs of Se matched for running time on last day of acquisition training. . . . . Number of trials to the extinction criterion of the second trial in which the total running time equals or exceeds five minutes with a maximum of 64 trials. Number of trials to reach the criterion of the second trial in which the running time is at least 2% minutes. . . . . . . Wilcoxon test for matched pairs using the second extinction trial in which the total running time was at least 150 sec. iv Page 12 17 18 19 LIST OF FIGURES Figure 1 Mean reciprocal total running time in seconds for extinction trials in blocks of 8 of Se extinguished in the PR alley and Se extnguish- ed in the CR alley for remaining gs not ex- tinsuished000.0.0C.0...’..................0.0.0.016 LIST OF APPENDICES Appendix Page A. Diagram of the apparatus. . . . . . . . . 27 B. Start box latencies in seconds summed over blocks of 8 extinction trials for each animal. . . . . . . . . . . . . . . 28 C. Running time in alley in seconds summed over blocks of 8 extinction trials for eachanlmal................29 D. Goal box latencies in seconds summed over blocks of 8 extinction trials for each animaloeeeeeeoeeeeeeeeoe 30 vi INTRODUCTION Various studies of partial reinforcement in a runway situation have shown the superior resistance to extinction of partially reinforced as compared with continuously reinforced animals. Hull (1952) and Sheffield (1949), posing the problem in stimulus-response terms, hypothesized that stimulation from nonreinforced trials remains as a stimulus trace for the partially reinforced animals, becoming part of the stimulus complex on subsequent trials when the animal receives reinforcement and then becomes conditioned to the running response. Thus when extinction trials are presented, the partial- ly reinforced animals will have had stimulus components from nonreinforced trials previously conditioned to running, while the continuously reinforced animals will not have had. Since the partially reinforced animals will have more of the components of the stimu- lus complex which occurs on an extinction trial conditioned to the response, they should prove the more resistant to extinction. This stimulus trace remaining from the stimulation of reinforced and non- reinforced trials should dissipate in time so that with highly distributed training trials, the 1 2 superiority of partial reinforcement should disappear. Weinstock (1954) and later Wilson, Weiss, and Amsel (1955). in testing this hypothesis, discovered that even with widely spaced trials (as much as 24 hours) the partial reinforcement effect could be pro- duced. Weinstock then proposed an explanation couched in a contiguity theory framework. Standard contiguity theory makes clear that the role of reinforcement is to remove the animal from stimulation so that the last response made prior to the termination of the stimulation will then be con- ditioned. In eXperimental extinction the reinforcement is withheld, and the animal makes competing responses which are conditioned to the stimulation so that the result is a decrement in response strength of the original response class. Weinstock elaborates upon this position by pro- posing further that nonreinforcement where previously there has been reinforcement produces frustration, the effect of which is to increase motivation. Early in training the effect of this motivational increase is mainly to strengthen irrelevant and interfering responses produced by frustration. With increased practice the competing responses tend to disappear because they are never reinforced. The competing re- sponses, elicited by the stimuli in the empty goal box, habituate over the course of nonreinforced trials. 3 The competing responses, having habituated, will occur with low frequency, and there will be little decrement in the strength of the original responses due to the presentation of a nonreinforced trial. During experimental extinction, animals previously trained with continuous reinforcement are frustrated for the first time, and interfering responses are elicited which have long since been habituated in the partially reinforced subjects. As a result, the run- ning response of the former group extinguishes quickly. Similarly, Amsel (1958) and Kendler (1957) suggest that when §s are switched to extinction, the partially reinforced §s have been trained to respond in the presence of antedating frustration stimuli, whereas the consistently reinforced gs have not. Adelman and Mastsch (1956) have demonstrated that frustration, acting as a powerful reinforcer, contri- butes to resistance to extinction. Spence (1960) has endeavored to clarify the role of frustration as an origin of the competing responses by suggesting that with each reinforced trial there develops in the organism a fractional anticipatory response (r6) which becomes conditioned in some degree to the stimulus cues in the instrumental chain. Non- reinforcement of a previously reinforced response results in an organismic state or response which Amsel (1958) has designated rf which contributes to the 4 general drive level (D) of the subject. There is a positive relation between the strength of r8 and the strength of rr, 1.6., the inhibitory effect of a nonreinforcement is related to the strength of the response tendency and would be zero or at very low levels in the earliest stages of learning. Not only would rf occur at the end of the response chain, 1.6., in the goal box, but as in the case of r would be ex- 8 pected to become conditioned to stimulus events earlier in the response chain. That is,‘§ should also come to make the anticipatory frustration response in the runway (Sc-rf). It is assumed that during experimental extinction, the frustration-aroused response, rf, through its own response-produced cues, Sf, tends to elicit previously learned or unlearned overt responses, some of which are incompatible with the learned instrumental response of running forward. Occurring as they do in the goal box, these competing responses should become conditioned to the stimulus cues present there and generalize to the highly similar cues of the alley. With repeti- tion of the nonreinforced trials during extinction these incompatible responses would become more strong- ly conditioned to the situation with the consequence that they would compete more with the learned instru- mental response. The purpose of this study is to investigate the 5 partial reinforcement effect, using a within-subjects design. This design has been used successfully by Peckham and Amsel (1964) in studying frustration effect, and by Black (1965) in analyzing the role of secondary reinforcement in extinction. The design used for this investigation attempts to minimize the stimulus-trace aftereffects of nonreinforcement which the Hull- Sheffield hypothesis holds to be the critical factor in the partial reinforcement effect. Likewise the design attempts to maximize the amount of nonreinforcement frustration, the habituation of which, Weinstock hypo- thesizes, is critical in producing the partial rein- forcement effect.1 A group of Se received two types of acquisition procedures: on one-half of the acquisition trials, gs experienced only continuous reinforcement in a white straight-alley, and on one-half of the acquisi- tion trials, §s experienced only 50% partial reinforce- ment in a black alley. For any one day's session, 1 When S is in the partial reinforcement situation, nonreinforcement will result in frustration-produced competing responses which habituate over the course of practice and drop to a low frequency of occurrence. When §,is in the continuous reinforcement situation, no such habituation will occur. During extinction the frustration-produced competing responses from nonrein- forcement have dropped to a lower level of occurrence through habituation (PR) and when rf is conditioned to the running response, than where the frustration-pro- duced competing responses from nonreinforcement appear at a high level of occurrence (CR) where no such con- ditioning of rf has occurred. 6 half of each S's trials were of each type of acquisi- tion training. Capaldi and Spivey (1963) suggest that the re- inforced trial (R) during partial reinforcement is in part conditioned or cued to the aftereffects of non- reinforcement (SN) from the previous nonreinforced trial (N). Thus a response under partial reinforcement is more likely to occur during extinction when the SN prevails than for a response under continuous rein- forcement. Capaldi and Stanley (1965) demonstrated that stimulus aftereffects of reinforcement and nonrein- forcement appear to persist for as long as 20 minutes. Thus the inter-trial interval of 5 minutes in the pre- sent study may minimize stimulus aftereffects but probably does not eliminate the effects of the stimulus-trace components. In the present study, where the acquisition trials were randomly altered between the partial reinforcement condition and the continuous reinforcement condition, a nonreinforced trial was as likely to be followed by a trial on the continuously reinforced white alley as it was by a trial (either reinforced or nonreinforced) on the partially reinforced black alley. If this is the case, §s extinguished in the partially reinforced alley, according to the stimulus-aftereffects position, should be no more resistant to extinction than those gs 7 extinguished in the continuously reinforced alley. By minimizing the effects of the stimulus-trace, the effects of habituation to frustration should stand out more clearly for the habituation to frustration hypothesis only requires that the partially reinforced alley and goal box be distinguished from the continuous- ly reinforced alley and goal box. If the Se, having experienced acquisition trials in both the white alley (CR) and the black alley (PR) demonstrate greater resistance to extinction when extinguished in the partial reinforcement black alley, then support is given Weinstock's hypothesis of habituation to frustration. METHOD SUBJECTS The subjects for this experiment were 10 ex- perimentally naive male albino rats, 10 male Tryon Brights, and 10 female Tryons. The age of the animals was not held constant, this being a within-subjects design. The range of ages was roughly 70 - 150 days at the beginning of the experiment. Animals were housed together in groups of 6 and were maintained on Efl.ll2 water at all times. APPARATUS The apparatus consisted of a pair of enclosed straight alleys (see appendix). The various components of the apparatus were a first runway, a first goal box; a second runway, and a second goal box. The interior of one runway and one goal box was painted black, the interior of the other runway and goal box was painted flat white. Each runway was 4' x 4%" x 6” and each goal box was 14" x 4%" x 6". The only door incorporated into this simple apparatus is a guillotine-type door located at the end of the alleys to prevent retracing from the goal box. A start door 10" long, painted the same color as the alley covered the first 10" of the 8 9 the runway. The rest of the apparatus as a unit had a hardware cloth top. The food cups were two Mason jar lids placed near the back and center of the goal box. PROCEDURE Habituation The preliminary training period was 10 days. Throughout this period §s were maintained on a 25- hour food deprivation schedule. Water was available at all times throughout the experiment. On each of these 10 days all §s were handled for 3 minutes prior to being fed. At the end of the feeding hour, all remaining food was removed from the cage. On days 8-10 each S was placed in the apparatus for 4 minutes and was allowed to eXplore the apparatus, 2 minutes in each alley. Acquisition Training For a period of 8 days, each S was given 8 trials per day under 25-hour food deprivation. Sixteen of the §s received continuous reinforcement in the black alley on 4 of his 8 daily trials and received 50% partial reinforcement in the white alley on 4 of his 8 daily trials. For 14 of the S3 the procedure was reversed. The presentation of the trials was random- ized as much as possible. On the trials where rein- 10 forcement was present, the S was removed from the goal box immediately after consumming the two, .045 gram Noyes pellets used as reinforcement. 0n the nonrein- forced trials, §_was confined to the goal box 45 sec. Between trials §_was placed in a small dark running box. The intertrial interval was 5 minutes. After each day's session, the gs were returned to their home cage and fed for one hour after which any remaining food was removed. For the duration of the experiment, including the habituation period, all of the S3 were kept in the experimentation room so that the stimuli for the §s were constant at all times. Latency, running and goal-entry measures were recorded by‘g using a step watch. Extinction or Test Trials Following the 64 acquisition trials, 32 exposures to continuous reinforcement and 32 exposures to partial reinforcement schedules, §s in each of the two groups were paired on the basis of total running time on the last day of acquisition training. One member of each pair was arbitrarily assigned to extinction on the previously continuous reinforcement alley while the other member of the pair was assigned to extinction on the previously partial reinforcement alley. The extinction trials were run exactly as were the 11 nonreinforced acquisition trials with s being confined to the goal box 45 seconds each trial. The inter-trial interval was kept as close to 5 minutes as possible, never exceeding 8 minutes. A S was considered to have been extinguished when it had two trials in which the total running time was at least 5 minutes. 12 TABLE 1 MEAN TOTAL RUNNING TIME IN SECONDS FOR PAIRS MATCHED FOR RUNNING TIME ON LAST DAY OF ACQUISITION TRAINING Mean running time of animal Mean running time of animal Assigned to extinction assigned to extinction under the previously under the previously Partial Reinforcement Continuous Reinforcement condition condition 1 2.63 3.25 2 2.65 3.25 3 4.00 3.63 4 4.13 3.75 5 6.00 5.13 6 6.25 5.38 7 6.38 5.88 8 11.75 6.75 9 13.38 8.75 10 21.00 21.00 11 1.13 1.28 12 1.25 2.00 13 1.63 2.50 14 2.38 2.75 U! U 1‘0 U'l \JJ 0 O O RESULTS The §s extinguished in the partial reinforcement alley and the §s extinguished in the continuous rein- forcement alley were compared as to differences in resistance to extinction. While there is a definite trend in the predicted direction, that the §s extin- guished in the PR alley would be more resistant to extinction, the results, on the whole, are not statis- tically significant. Pooling the data from all groups (strain and sex) a sign test for matched pairs was performed to compare the group extinguished in the partial reinforcement alley and the group extinguished in the continuous reinforcement alley. The criterion for this compari- son was the second extinction trial in which the total running time for the S was 150 seconds or more. This comparison revealed that the S3 extinguished in the PR alley appear to be more resistant to extinction in terms of maintaining faster running speeds on extinc- tion trials. The associated probability of .046 (N213, x=3) is significant at the .05 level. A subsequent Wilcoxon-T test for matched pairs was performed on these same data to consider the rel- ative magnitude of the differences as well as their 13 14 direction. The associated probability (T=34.5, N=14) of .13 is not statistically significant. The reason for this discrepancy between the results of the two tests appears to be that those pairs of Se whose data are Opposite to the predicted results are so extreme as to render the more numerous results in the pre- dicted direction without significance. Figure 1 reveals this trend in graphic form. During the first several blocks of trials, the trend of the extinction data is fairly typical of the partial reinforcement extinction gradient, the partial reinforcement group having the shorter latencies, and both groups having longer latencies with successive trials. However, on later trials, the §s in the CR group that are remaining are running rapidly enough that on the eighth day of extinction trials, it appears that the OR group is more resistant to ex- tinction than the PR group. A further comparison was made between the same two groups, this time using the second extinction trial in which the running time exceeded 30 seconds as the criterion. The results of the sign test (N=12, x=3) have an associated probability of .973 while the Wilcoxon test (T=28, N212) yielded a probability of .199, both of which fall below statistical significance at the .05 level. At the 30 second criterion, the PR 15 group and the CR group are even closer together than the comparison made at 150 seconds. Other tests were performed, all of which reveal the same trend as that reported above. There were strain differences between the Tryon brights and the albinos, with the Tryons having general- ly slower running times on all trials both during acquisition and extinction. 16 FIGURE 1 §s extinguished in the PR alley ————— §s extinguished in the 27:0 I § CR alley 25.0 13.0 11.0 run- 9-0 ning time 7.0 I/BOC. 3.0 UuO 012 3 45 6'78 Blocks of 8 trials MEAN RECIPROCAL TOTAL RUNNING TIME IN SECONDS FOR EXTINCTION TRIALS IN BLOCKS OF 8 OF SS EXTINGUISHED IN THE PR ALLEY AND Se EXTINGUISHED IN THE CR ALLEY FOR REMAINING SS NOT EXTINGUISHED 17 TABLE 2 NUMBER OF TRIALS TO THE EXTINCTION CRITERION OF THE SECOND TRIAL IN WHICH THE TOTAL RUNNING TIME EQUALS OR EXCEEDS FIVE MINUTES WITH A MAXIMUM OF 64 TRIALS Animal extinguished under Animal extinguished under the previously Partial the previously Continuous Reinforcement condition Reinforcement condition 1 64 64 2 64 40 3 64 45 4 64 38 5 64 54 6 64 64 7 24 54 8 64 64 9 55 64 10 64 39 11 64 12 12 64 15 13 44 24 14 64 64 24 64 - UI 18 TABLE 3 THE NUMBER OF TRIALS TO REACH THE CRITERION OF THE SECOND TRIAL IN WHICH THE RUNNING TIME IS AT LEAST 2% MINUTES Animal extinguished under Animal extinguished under the previously Partial the previously Continuous Reinforcement condition Reinforcement condition 1 64 37 2 37 26 3 64 43 4 64 38 5 39 64 6 10 64 7 24 18 8 64 64 9 16 54 10 64 39 11 64 10 12 64 15 13 43 24 14 64 37 15 24 34 Trial reaching criterion 19 TABLE 4 WILCOXON TEST FOR MATCHED PAIRS USING THE SECOND EXTINCTION TRIAL IN WHICH THE TOTAL RUNNING TIME WAS AT LEAST 150 SECONDS rank of CR group PR group d d 37 64 +27 10.5 26 37 +11 3 43 64 +21 5 38 64 +26 9 54 39 -25 -7 64 10 -54 ~13.5 18 24 + 6 1 64 64 0 0 64 16 -48 -12 39 64 +25 7 10 64 +54 13.5 15 64 +25 7 24 43 +19 4 37 64 +27 10.5 34 24 ~10 -2 z = 1.13 has an associated probability value of .13 DISCUSSION AND CONCLUSIONS The hypothesis pr0posed that §s experiencing continuous reinforcement in a white alley and partial reinforcement in a black alley should extinguish more rapidly in the white or continuously reinforced alley than in the black, partially reinforced alley cannot be definitely supported by this study. Results having probability values in the range of .05 to .20 suggest that the percentage of partial reinforcement may have been too high to obtain the maximum partial reinforcement effect. In any one test- ing session, each § received a total of 75% rein- forcement with only 2 nonreinforced trials, and a total of only 16 nonreinforced trials in the course of the experimental procedure. Amsel (1958) hypothesized that the partial reinforcement effect will be evident only after a critical number of training trials, i.e., only after sf becomes conditioned to the running reaponse as opposed to other competing responses; the number of trials will depend on the training situation. Com- paring four groups: a 24-trial CR group, a 24-trial 50% PR group, an 84-trial CR group, and an 84-trial 50% PR group, Amsel discovered that the two 24-trial groups did not differ significantly in amount of resistance 20 21 to extinction, but that the CR group was slightly more resistant at this stage. The differences between the two 84-trial groups was highly significant, the PR group showing much greater resistance to extinction. Thel§s in the present study had a total of 64 trials: 32 trials continuously reinforced and 32 trials 50% partially reinforced. Comparing this breakdown with Amsel's 24-trial acquisition groups, both similarities and differences can be noted. In Amsel's study the CR group was still demonstrating a tendency to be more resistant to extinction. Presumably the PR group is still making frustration-produced competing responses in the goal box which have not yet habituated. In the present study, the SS having received only 32 PR trials are al- ready demonstrating a definite trend for the partially reinforced §s to be more resistant to extinction than the continuously reinforced gs. This strongly suggests that in the PR group, the competing responses made in the empty goal box are beginning to disappear as the result of habituation, but that the total number of non- reinforced trials is not large enough to allow the maximum partial reinforcement effect. Presumably, sixteen nonreinforced trials is not sufficient to allow habituation of competing reaponses in the present design. The results of this study suggest a further re- finement in the design: that the total number of trials 22 be increased in order to have a larger number of non- reinforced trials and/or that the percent of partial reinforcement be decreased below 50% in order to have a larger number of nonreinforced trials. Either alter- ation would serve the function of increasing rf in the goal box more rapidly so that the frustration-pro- duced competing responses would habituate and would drop out with rf becoming conditioned to the running response. By using one of the above suggested methods to increase rf sufficiently to allow the frustration-pro- duced competing responses to habituate with rf becoming conditioned to the approach response, it may then be possible to demonstrate that the group extinguished in the partial reinforcement alley will be more resistant to extinction than the group extinguished in the con- tinuous reinforcement alley in this within-subjects design. An independent check on the same hypothesis in- volved in the present study would be to employ an identical procedure with the one exception that a common goal box (grey) be used for both the black and the white alley. Frustration-produced competing responses made in the common grey goal box should not be chained to the stimuli of one alley any more than to the stimuli of the other alley; in this case one would predict that 23 there would be no difference in resistance to extinction between the §s extinguished in the partially reinforced alley and those gs extinguished in the continuously reinforced alley. One further condition which may account for Ss' failure to completely discriminate the two alley conditions is the two identical Mason jar lids used in the goal box in each alley. It would be better if these two food cups were painted the color of the alley or if the pellets were placed directly on the floor of the goal box. BIBLIOGRAPHY ADELMAN, H.M., and MAATSCH, J.L. (1956) Learning and extinction based upon frustration, food reward, and eXploratory tendency, J, exp, Psychol., 5g, 311-3150 AMSEL, A., (1958) The role of frustrative nonreward in noncontinuous reward situations, Psychol, Bull., {55, 102-119. BLACK, R.W., (1965) Differential conditioning, ex- tinction, and secondary reinforcement, J, e p. PSYChOJ-e’ _6_2, 67-740 CAPALDI, E.J. and SPIVEY, J.E., (1963) Effect of goal box similarity on the aftereffect of nonreinforce- ment and resistance to extinction, J, exp, Psychol., fl, 461”465. CAPALDI, E.J. and STANLEY, L.R. (1963) Temporal properties 02 reinforcement aftereffects, J, exp, Psychol., £5, 1 9-175. HULL, C.L., (1952) A Behavior S stem, New Haven: Yale University Press. KENDLER, H.H., PLISKOFF, 3.3., D'AMATO, M.R., and KATZ, 8., (1957) Nonreinforcement vs. reinforcements as vari- ables in the partial reinforcement effect, J, exp. P8201101. , 22’ 209-2150 KIMBLE, G., (1961) ar d M r u s' c n iti nin and leapning, New York, Appleton-Century-Crofts, Inc. PECKHAM, R.H., and AMSEL, A., (1964) Magnitude of re- ward and the frustration effect in a within-sub- jects design, Psychon, Sci., 1, 285-286. SHEFFIELD, V.F. (1949) Extinction as a function of partial reinforcement and distribution of prac- tice, J, exp, Psychol., 32, 511-526. SPENCE, K.W., (1960) Behavigr thepry and learning: Se- lected a ers, Englewood Cliffs, J.J.: Prentice-Hall. 24 25 WILSON, W., WEISS, E.J., and AMSEL, A., (1955) Two tests of the Sheffield hypothesis concerning resistance to extinction, partial reinforcement, and disgribution of practice, J. exp. Psychol., 0, 51- 0. WEINSTOCK, S., (1954) Resistance to extinction of a running response following partial reinforcement under widely spaced trials, J, comp. physiol. Psychol.,‘£1. 318-323. APPEN DI CES 26 27 APPENDIX A INTERCHANGEABLE GB‘s DOOR HARDWARE (1,0111 4/ K o START soon 36 START noon 1" WHH’E DIAGRAM OF THE APPARATUS APPENDIX B START BOX LATENCIES IN SECONDS FOR EACH ANIMAL SUMMED OVER BLOCKS OF 8 EXTINCTION TRIALS 5 Day 3 4 5 1 18.5 7 34 68.5 45 33 12.5 30 2 26 21 35 49 162 119 37 27 PR 3 12 43.5 169 - - - - - alley 4 19 9 20 46 46 29 28 83 5 17.5 18 29 51 75 35.5 24 40 6 13.5 20.5 101 85 78 359‘ - - 7 14.5 47 88 102 162 196- 108 - 8 20 22 69 22.5 6 23.5 5.5 9.5 9 39 51 46 35 27 41 13.5 59 1C 124 151 830 - - - - - 11 4 4 11.5 5.5 10 26.5 11 9.5 12 4 4 35 10.5 21.5 9 10.5 24 13 4 8.5 11 23 11.5 12.5 5 5.5 14 4 17 31 26 18.5 42.5 47 38 15 4.5 22.5 41.5 15.5 23.5 15 34.5 18 16 48 22 124 99 58 81 38 33 17 13 23 83 37 40 44 37 87 18 19.5 11.5 23 74 48 31 11.5 21 19 33 38 260 59 127 - - - CR 20 21 42 67 - - - - - alley21 23 39 - - - - - - 22 19 47 232 132 68.5 71 75 38.5 23 16 40 61 39 36 29 58 20 24 50 56 60 159 120 125 47 77 25 54 109 - - - - - - 26 23 19.5 26.5 30 “40“ +5 - - 27 42 228 167 42.5 - - - - 28 6.5 24.5 505 10.5 17 "’ " '- 29 11 11.5 16 18.5 18.5 19 17.5 10 3O 19 33.5 26 60 84 21 14 11 28 APPENDIX C RUNNING TIME IN ALLEY IN SECONDS SUMMED OVER BLOCKS 0F 8 EXTINCTION TRIALS FOR EACH ANIMAL .2 Day 1 2 3 4 5 6 7 8 13 42 103 94 48 58 89 53 21 29 85 416 183 81 45 17 14 48 241 - - - - - PR 27 4O 46 68 27 48 29 46 alley 36 37 52 55 108 34 75 44 17 7O 42 369 257 46 - - 23 260 91 166 68 180 72 - 21 43 89 21 32 32 34 29 283 89 104 33 102 46 125 8 2O 82 59 41 15 42 8 15 104 156 25 46 200 24 16 35 122 59 13 13 33 65 61 62 46 150 13 19 68 11 170 88 13 21 ddfld—l-fi U'lk'UN-‘Oomflmm-FUN‘ ooooooooxo 16 44 36 84 49 75 4o 28 87 17 77 35 37 76 80 111 337 37 18 17 26 67 77 83 28 39 48 19 57 72 97 103 331 - 2o 34 42 334 - - 21 32 82 - - - 22 25 57 37 57 4o 27 117 25 23 45 63 123 88 48 35 66 25 CR 24 106 90 60 135 33 57 59 112 alley25 258 - - - - - 26 1o 20 94 92 112 28 - - 27 22 143 185 228 486 - - - 28 8 24 13 21 57 - - - 29 9 14 36 53 62 46 63 «9 30 1o 30 27 50 57 19 16 9 29 APPENDIX D GOAL BOX LATENCIES IN SECONDS SUMMED OVER BLOCKS OF 8 EXTINCTION TRIALS FOR EACH ANIMAL 5 Day 1 2 3 4 5 6 7 8 1 4 119 220 149 125 149 106 84 2 4 18 60 41 97 90 10 13 3 6 48 506 - - - - - 4 15.5 133 418 381 217 130 93 156 PR 5 5.5 45 70 279 230 196 189 302 alley 6 4 57 166 288 121 397 - - 7 54 286 185 279 315 597 785 523 8 11 30 156 23 13 23 48 13 9 145 412 285 47 26 38 115 32 10 18 99 84 - - - - - '11 4 4 60 209 160 188 208 89 12 4 4 158 261 503 218 '71 50 13 4 4 6 72 413 267 187 67 14 4 12 8 88 317 235 196 77 15 4 41 68 8 170 88 13 179 (D O\ 16 4 66 195 151 219 80 107 17 29 183 624 506 420 111 337 109 18 4 34 173 32 515 136 50 69 19 49 63 298 168 533 - - - CR 20 248 31 344 - - - - - alley21 5 562 - - - - - - CR 22 58 129 118 62 61 133 236 19 alley23 32 92 208 113 111 140 95 33 24 72 296 154 96 67 229 158 286 25 205 4 - - - - - - 26 4 8 169 157 162 878 - - 27 4 4 271 875 298 - - - 28 4 8 12 8 579 - - - 29 4 405 312 158 317 262 82 92 30 157 6 12 35 49 119 179 130