SATlATIQN ELEANNG: ONE TREAL PER DAY Wash in: 1412:0991“ of M. A. MICHIGAN STATE UNIVERSITY Eruce Thomas tanker-f E963 LIBRARY 3 Michigan State ‘ ‘1 University [J ABSTRACT SATIATION LEARNING; ONE TRIAL PER DAY by Bruce Thomas Leckart The problem which this research was designed to investigate can be stated in two parts: (1) Can stimulus satiation mediate avoidance learning when highly spaced "forced" trials are administered to one side of a T-maze? (2) If so, what is the function which relates the number of forced trials to the learned satiation effect (83' preference for the opposite side). Sixty albino female rats were randomly assigned to five independent groups receiving either 0, 2, 6, 12 or 20 forced reward trials to one side of an enclosed T—maze, with an ITI of 24 hours. One day after the last forced trial a series of 10 free test trials was initiated. On these free trials the reward was on the same side as on the forced trials; the free trials were also administered with an ITI of 24 hours. A significant persistent preference for the opposite side was found in Groups 6, 12 and 20 on the 10 free trials, Bruce Thomas Leckart supporting the notion that stimulus satiation can act as a mediator for avoidance learning when highly spaced forced trials are given to one side of the maze. The most per- sistent preference occurred in the group that received 6 forced trials. In other words, a curvilinear relationship was found between the number of forced trials and the number of visits to the previously unvisited side during free trials. The learned effect was explained within an elici- tation framework which includes a postulate on stimulus satiation. According to this position, responses which are produced by stimulus satiation are contiguously conditioned to the cues at the choice point, mediating an avoidance of the previously experienced cues. The curvilinear effect was attributed to two competing response tendencies (approach and avoidance). On the one hand there is a tendency to avoid the previously visited side as mediated by stimulus satiation. On the other hand there is a tendency to approach the previously visited side resulting from repeated pairings of food with cues in that side of the maze. The tendencies increase at different rates, resulting in the curvilinear function obtained. fi' Date (41% I; /,5)(‘(1 Approved 2% am“? gm SATIATION LEARNING: ONE TRIAL PER DAY BY Bruce Thomas Leckart A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Psychology 1963 -1 \ ~ 5) '.<\\ I (\x u‘ \ a (it To Annette The Prime Mover ACKNOWLEDGMENT The author wishes to thank Dr. M. Ray Denny,without whom not one word of this thesis would have been possible. Thanks are also extended to Dr. Stanley C. Ratner for his aid in preparation of the manuscript and to Dr. William Stellwagen for his suggestions on the statis— tical analysis of the data. ii HISTORY . PROBLEM . APPARATUS SUBJECTS . PROCEDURE RESULTS . DISCUSSION SUMMARY . REFERENCES TABLE OF CONTENTS iii Page ll l3 14 18 28 31 33 LIST OF TABLES Table Page 1. Alley preference during spaced free trials by blocks for all groups . . . . . . . . . . . 22 iv LIST OF FIGURES Figure Page 1. Schematic drawing of the maze . . . . . . . . . 12 2. Mean running times during forced trials for all groups . . . . . . . . . . . . . . . . . l9 3. Mean number of responses to the opposite side during spaced test trials for all groups . . . . . . . . . . . . . . . . . . . 20 4. Percentage of responses to the opposite side during spaced free trials for Groups 6 and 12 combined and Group 0 . . . . . . . . . . . 23 5. Percentage of responses to the opposite side during spaced free trials for Groups 6 and 12 combined and Group 2 . . . . . . . . . . . 24 6. Percentage of responses to the opposite side during spaced free trials for Groups 6 and 12 combined and Group 20 . . . . . . . . . . 25 7. Mean number of massed free trials to ex- tinction criterion for all groups . . . . . . 27 HISTORY Tolman (1925) first observed what was later to be called spontaneous alternation. He found that rats run in a T-maze in which both alleys led to the same goal box tended to alternate between paths chosen on successive trials. In general, there have been two types of theoretical explanations of spontaneous alternation: (a) response— oriented theories and (b) stimulus-oriented theories. The former centers around Hull's (1943) concept of reactive inhi— bition (Ir). According to this interpretation the occurrence of a response temporarily reduces the probability that the same response will be evoked on a subsequent trial. Stimulus- oriented theories explain spontaneous alternation by hypothe- sizing that perception of a stimulus reduces the organism's potential reaction to the same stimulus on subsequent trials. Hull's (1943) Postulate 8 states that: "Whenever a reaction (R) is evoked in an organism there is created as a result a primary drive (D); (a) this has an innate capacity (Ir) to inhibit the reaction potentiality (sEr) to that response . . . Zeaman and House (1951) make three deductions from Postulate 8. They state that alternation tendency will be: "(1) a negatively accelerated decreasing function of the time between responses; (2) a positively accelerated increas- ing function of the work involved in the execution of the response; and (3) a simple linear increasing function of the number of evocations." Heathers (1940) presents evidence in support of the first deduction. Assuming that Ir is responsible for spon- taneous alternation and that it dissipates with time he hypothesized that the avoidance of repetition of a response would decrease with increases in the time between responses. Subjects were run six trials per day in an elevated T—maze with the ITI varying between 15 and 120 seconds. Evidence was found that the amount of alternation decreased with in- creases in the ITI thus supporting the conceptualization of Ir as a determinant of spontaneous alternation. Solomon (1948) tested the second deduction by in- clining the arms of a T-maze 16 degrees from the horizontal. Hypothetically this would increase the work involved in making the responses and hence the amount of Ir associated with each response resulting in an increased tendency to alternate. Supporting data in the form of an increase in the percentage of alternation in the tilted alleys as compared to horizontal alleys was found. However it was also found that increasing the weight that S carried through the maze did not increase the amount of alternation as the theory would predict. Zeaman and House (1951) tested the third deduction. They hypothesized that increases in the number of trials would increase Ir and the amount of spontaneous alternation. Using an elevated T—maze with one alley removed 85 were forced 10 trials to one side. This procedure was duplicated with forced trials to alternate sides of the maze on succes— sive days until each side was rewarded 100 times. Ten seconds following each block of 10 forced trials a free or alternation trial was given by replacing the missing alley. After the first 100 forced trials 100% alternation was ob- tained on free trials. "A series of a smaller number of forced trials was next tried." 1 2 4, or 7 forced trials were administered and a free trial given one hour later. In support of the third deduction it was found that the greater the number of forced trials the greater the likelihood of a visitation to the less frequently visited side. The relationship between the number of forced trials and the percentage of alternation was found to be positive and linear. Evidence was also found to support the first deduction. The percentage of visitations to the less frequently visited side was shown to decrease with increases in the time between forced and free trials. No significant alternation was found with delays of 18 and 24 hours between forced and free trials. Dennis (1939) was one of the first investigators to find good evidence for a stimulus interpretation of spon- taneous alternation. Using a two unit square maze he found a significant avoidance of repetition of exposure to the same stimuli but no significant avoidance of repetition of the same response. He concludes that there is an avoidance of a specific pathway rather than a specific response. This finding obviously seriously questions the validity of the explanatory concept of Ir in dealing with spontaneous alter- nation. Glanzer (1953) has formulated a stimulus satiation postulate to deal with spontaneous alternation. Accordingly, stimulus satiation is an incremental function of the time S perceives the stimulus and a decremental function of the time spent away from the object. Increases of stimulus satiation are hypothesized to bring about a reduction of the Ss tend- ency to respond to the same stimulus on a subsequent trial. In this framework spontaneous alternation is based upon non— repetition of exposure to a stimulus as opposed to non—repe- tition of a response. Glanzer tested his deduction in a cross—shaped maze by starting Ss on opposite ends of the maze on successive trials, thus pitting the stimulus satiation and reactive inhibition explanations against each other. Avoidance of the gige previously visited is evidence in support of stimulus satiation whereas avoidance of a previous response is evidence in support of reactive inhibition. The results obtained sup- ported the conception of a stimulus satiation postulate. walker, Dember, Earl and Karoly (1955) extended the work of Glanzer in determining what stimulus class the organism responds to when spontaneously alternating. By ro- tating a cross—shaped maze 1800 between trials they succeeded in pitting stimulus, response, and place against each other as determinants of alternation. First, they found that $5 alternate on the basis of intra—maze stimuli as opposed to responses. Second, evidence was found supporting the hypoth- esis that $3 also alternate with respect to place (extra-maze stimuli in the experimental setting). The results lend them- selves to a stimulus satiation interpretation, relevant stimuli being both intra and extra-maze in nature. They con- clude that intra—maze cues (stimuli), extra—maze cues (place) and responses are, respectively, the most important determi- nants of alternation. Rothkopf and Zeaman (1952) varying the number of forced trials in an elevated T—maze found evidence for an in— crease in alternation with increases in the number of forced trials to one side of the maze. They also found evidence for a learning effect, alternation increasing with practice in alternating. They postulate a two-factor theory, in- cluding concepts of both stimulus satiation and reactive inhi- bition, in dealing with the data. First, there is an adap- tation to external cues (stimuli) during traversing of the maze which results in a preference to respond to the less adapted stimuli. And second, making a response leads to a fatigue like state (Ir) which inhibits the repetition of that response and results in a preference of the organism for the less fatigued response. Denny (1957) has offered evidence that stimulus sati- ation may act as a mediator of avoidance learning. Accord- ingly, satiation with stimuli in one alley of a T—maze leads to avoidance of those stimuli on succeeding trials. Two trials were given per day with an ITI of thirty minutes. Forced and free trials were interspersed and controlled by the experimenter so that one side was visited twice as often as the other. A significant tendency for $5 to visit the less frequently visited side on free trials was found. Evi- dence suggesting that this preference was learned is supported by three empirical observations. (1) The preference develops along the lines of a typical learning curve. (2) After initial training 53 continued to visit the less frequently visited side on successive free trials. (3) There was re- tention of the preference for at least a week. Denny's results were interpreted with the Elicitation Theory (1955) framework which includes a postulate of stimulus satiation. Accordingly, satiation with cues in one arm of the maze elicits an avoidance response with respect to those cues. This avoidance reaction is assumed to be contiguously conditioned to these cues. Finally, Hill, Cotton and Clayton (1963) using a forced trial technique have found no preference for one side of a T-maze that was rewarded twice as often as the other. They also found evidence that Ss visiting one arm but not the other showed more of a tendency to visit the non-experienced arm than those Ss that had previously experienced it. They interpret their results as evidence against a frequency of reward theory of learning and as support for Denny's findings. THE PROBLEM The present research is related to the work done on spontaneous alternation, Heathers (1940), Solomon (1948), Dennis (1939), Glanzer (1953), and Walker §p_§l. (1955), but unlike the typical studies done in this area the Ss are not always permitted to respond freely. The present method, first used by Zeaman and House (1951), consists of giving a series of forced rewarded trials to one side of a T-maze followed by a free trial or a series of free trials. Zeaman and House and Rothkopf and Zeaman (1952) applied this method in an elevated maze and found clear evidence for an increased alternation effect with forced trials to one side though. after 24 hours the effect was gone. The latter investigators, however, found a suspicion of a learned alternation effect. Since this time many investigators, Glanzer (1953), Dember and Earl (1957), Montgomery (1952), and Walker gp_§1. (1955), have obtained evidence that alternation, regardless of pro- cedure used, seems to be a stimulus satiation effect rather than a response produced effect (reactive inhibition) as assumed by Zeaman and associates. In the past five years one published study (Denny, 1957) and a number of other studies have been conducted by Denny and associates which indicate that when the Zeaman and House technique is used in an enclosed maze, rather than an open elevated maze, the stimulus satiation effect as revealed by a preference for the non-visited side is long lasting or seemingly learned. This particular series of researches culminates with an as yet unpublished study by Denny and Leckart (1962) and the present master's research. The problem which this research was designed to in— vestigate can be stated in two parts: (1) Can stimulus sati- ation mediate avoidance learning when highly spaced "forced" trials are administered to one side of a T-maze? (2) If so, what is the function which relates the number of forced trials to the learned satiation effect (55' preference for the oppo— site side). The possibility of a learned effect has been evaluated by Denny and Leckart by using 10 free trials spaced 24 hours apart. This method virtually eliminates from the test trials any immediate non-learned stimulus satiation effects which could result from a single trial or any accumulation of such effects. In the present research not only are the free trials separated by 24 hours but the original forced trials are also separated by 24 hours. Thus it is possible to see whether lO discrete, noncumulative stimulus satiation effects can mediate avoidance learning (according to the elicitation framework this is entirely conceivable). Given this procedure, the present study involves varying the number of forced trials which independent groups receive and evaluating the effects with 10 spaced free trials. APPARATUS The apparatus was an enclosed T—maze constructed of unpainted wood shown diagrammatically in Figure l. The over- all dimensions of the arms and stem were 28" and 16" respec— tively. The goal boxes and the starting box were 12" long. The alleys and goal boxes were 4" wide and 6" high and covered with hardware cloth. The start box was covered with unpainted wood. To provide differential intra-maze cues, on each wall of the right arm, 2" from the bottom, and down the center of the floor of the right arm there were full length strips of 3/4" black electrical tape. The laboratory situation was sudh as to provide a wealth of differential extra-maze cues. Opaque guillotine doors were located at the entrance to the stem, and at the entrance to each arm and goal box. These doors were closed behind S as he traversed the maze to preclude retracing. ll 12 B ’T Left Arm : Goal Stem Box r--d Start- ing Box Figure l. ---0 Right Arm Goal Box A, B, C, D, E — Guillotine Doors Diagrammatic drawing of T-maze SUBJECTS The subjects used in the present study were 61 experimentally naive female albino rats selected from the colony maintained by the department of psychology at Michigan State University. One animal was discarded for failing to reach an habituation criterion. All animals ranged in age from 90 to 120 days at the beginning of the experiment. 13 PROCEDURE Pretraining All Ss were given 12 days of habituation. On each of these days S was first handled for two minutes and then allowed to eat Lab Blox for one hour. All Ss were under a 23 hour food deprivation schedule throughout the experiment. Water was ad lib at all times except in the maze. On Day 9, immediately after handling, each S was placed first in one goal box and then in the other for food- association trials. On each of these "visits" six 45 mg. P. J. Noyes reward pellets were available in the goal cup. The S was allowed to remain in each goal box until it had eaten all of the pellets or until three minutes had elapsed. The usual one hour feeding followed the last visit. Because many 83 did not eat during this first period, 12 hours after the morning handling the 35 were given additional food familiarization trials. At this time they were twice placed in either one of two clear plastic boxes with two reward pellets in the food cup and were allowed to remain there un- til they had consumed both pellets or for a maximum of three minutes. These same procedures were duplicated on Days 10- 14 15 12 with visits to each goal box and each plastic box counter- balanced. One S failed to meet the habituation criterion by not eating any of the pellets and was discarded. Acquisition The subjects were randomly assigned to five inde- pendent groups (0, 2, 6, 12, and 20). The numbers corre— spond to the number of forced trials that each group received during acquisition. A forced trial consisted of denying access to one side of the maze by closing the door leading to that alley. Prior to all trials each S was given thirty seconds of handling. S was then placed in the start box for 10 seconds at the end of which time the door leading to the stem was Opened. On all trials, as S traversed the maze the guillotine doors were closed behind him, as all but his tail traversed the door jamb, to preclude retracing. Each of these forced trials was rewarded by one 45 mg. P. J. Noyes reward pellet placed in the goal box where S was allowed to remain until the pellet was consumed. All forced trials for a particular S were to the same side of the maze, and the ITI was always 24 hours. One half of the Ss within each group were forced to the right and the other half were forced to the left. The time between leaving the start box and entering the goal box was recorded for all trials (running 16 time). This was accomplished by starting a stopwatch as S entered the stem and stopping it as the door to the goal box began to close. The order of running within each home cage was randomly determined each day to preclude the possibility of daily feedings reinforcing turning behavior in a non- random fashion. Test Trials One day after the last forced trial ten free trials were administered with an ITI of 24 hours. The free trial procedure remained the same as the procedure during acqui- sition with the exception that access was no longer denied to the previously unvisited side; both choice point doors were open. All visits to the previously non-experienced side were non-rewarded whereas visits to the previously experienced side continued to be reinforced. Group 0, the control group was given free trials 24 hours after the last day of the habituation procedure. The side visited as well as the run- ning time was recorded for each trial. Forty-eight hours after the last free trial a series of massed free trials was administered with an ITI of 10 seconds. These trials continued until S had made two consec- utive visits to the reinforced side (extinction criterion for the satiation habit). The number of trials to criterion was 17 recorded for each animal as well as running time and side visited. RESULTS Acquisition The mean running times on forced trials are presented for all groups in Figure 2. With an increase in the number of trials the running times decrease as expected, though prior research (Leckart and Denny, 1962) has shown an in- crease after 5 or 6 forced trials when the forced trials are massed (when cumulative satiation effects are possible). Test Trials The mean number of visits, during spaced free trials, to the side opposite the side to Which the Ss were forced are presented for each group in Figure 3. Here we see a curvi- linear relationship between the number of forced trials and the number of visits to the Opposite side. An analysis of variance was performed on these data, comparing the five groups on the number of responses made to the opposite side during the first ten free trials. An F value of 6.538 was obtained, with 4 and 55 df's, which is significant at the .01 level of confidence. Tukeys test (Edwards, 1961) for a significant gap between group means was 18 l9 .mmoowm Ham How mamflup poouom mcflusp mmEHu mcflccow smog mica—mp. omUmOn. ON m. o. .v. N. O. m o .N wusmflm .v N I I ‘o\o or. ‘5 co m . II I a x .n. . w \\w ”WI .%W p s. o . (1 II S s..fi aapa. .1 4... .1 .11 .. . . om ...... a. ........ .IN_ a .8 .......... o .. .. ..................... N mn50¢0 O. N. .v. o. m. cm NN BWIJ. ONINNOH NVHW 20 .mmsowm Ham Mom mamas» umwu poommm mcflwsp open ouflmommo mnu ou noncommmu mo HmQESG cmoz .m ousmam 33m... Sumo... .3 mumzaz om m. a m o lullrlllllrllll—lrlllu BOIS 31|SOddO OJ. SESNOdSBB :JO HEBWON NVBW 21 performed. The differences between Groups 2 and 6, and Groups 12 and 20 are significant (P < .01 and P < .05, re— spectively). No significant differences eXist between Groups 0 and 2 and Groups 6 and 12. The extension of the median test, a nonparametric specified by Siegel (1956) was performed on the same data. A X2 value of 21.953 was ob- tained, with 4 df's, which is significant at the .01 level of confidence (P < .001). Selected median tests were per- formed to determine the locus of the significant differences between groups. The following differences are significant: (1) Group 2 vs Group 6 (P (.01), (2) Group 6 vs Group 20 (P < .05), (3) Groups 0 and 2 combined vs Groups 6 and 12 com- bined (P < .001), (4) Groups 6 and 12 vs Group 20 (P = .039). The differences between Groups 0 and 2, 6 and 12, and 12 and 20 are not statistically significant. The data for Groups 6 and 12 were therefore combined and are graphically compared, individually, with Groups 0, 2, and 20 on the number of re- sponses to the opposite side over the course of the ten free trials (Figures 4, 5, and 6). In order to test the persistence of the learned ef- fect the 10 free trials were divided into three blocks (Trials 1—3, 4-6 and 7-10) and the Z approximation of the binomial was utilized to test for significant tendencies to visit the sides of the maze. The results are summarized in Table l. 2 2 :0. V 5 mpfim popuwzfiu 93 woconmmwum pom mucmuomwum ucmoflmasmfim uGMUAMHcmHm oz Amo.uv no open wufimommo osu How mosmuwmoum mucoummowm Damnamacmfim oz pcmuwmwcmam :0. V 3 open ouflmommo on“ now moawummmnm :0. V 5 mp“; mufimommo 9.3 How oocmummmnm pennamwcmwm on :0. V3 open muwmommo map How oocwnwmmwm ucmoamasmflm NH :0. V my open mnemommo map mocwnomowm pcmoHMHcmHm oz DGMUHMHcmHm Mom mocwummoum Damnamacmflm o Aao.uv_mv open popHMBou map How mocwnwmoum Amo..v.mv open muamommo may oocmummmnm unmoflmacmflm usmoflmacmflm 02 Mom wouoummmum ucmowmwcmwm N Amo.uv my mpflm popnm3mu on» now wocwummoum mocwummwum HGMUHMHcmMm ucmoflmwcmwm oz wucwuwmwum pancamwcmflm oz 0 ACHInV Aolgv Amlav mmsonw mxoon an mHmHua .mmsowm Ham How mxooH£ >3 mamwuu moum pwommm mcwusp wonmnwmwwm m0HH¢II.H manna 23 .o msowo paw meHnEoo NH paw o mmsowo How mamfluu comm pwommm msflwsp opflm muflmommo map ou mmmcomwmw mo mmwucmowwm m.._<_m._. mum“. O. m m OIIOII!\O '.|.|O N. 02.4 0 mQDOmo h o m, V m N .w wwsmflm OOOOO vom— O in BOIS ELLISOddO 3H1 OJ. BSNOdSHH 1N3383d Om 24 .m msouw paw pmsflnfioo NH pom o mmsoww Mom mamawu omuw pmommm mcflnsp open wuflmommo map 0» noncommww mo wmmucwoumm m|_<_m_._. mum“. O. m m N. o m V m N COOOCCOOOOOOC N N_ 02.4 0 WQDOKO .m musmflm 0. ON 0m 0.? on 00 ON 00 om 00. 30:5 BJJSOddO BHI 01 BSNOdSBH 1N3383d ' 25 .ON msowo paw pmcflnfiou NH can 0 wmsoww How mHmHHu wmuw pmommm mcflwsp mpflm oufimommo may 0» noncommmw mo wmwusmouwm mixers”; mum... O. m m u. o m. V m N III III-IIVAUmw N_ O 2.4 O maDOmo .0 wusmflm o o o o o o m