THE INFLUENCE OF STIMULUS SATIABILITY 0N LEARNING IN TWO DIFFERENT TASKS Thesls Ior II"; Degree-of DII. D. MICHIGAN STATE UNIVERSITY James Norman Allen 1963 THESIS This is to certify that the thesis entitled .' r i ‘1 ~' ‘ :"‘ ‘ " .1? . 1.1.; Is-Jllabd..vd L'J.‘ Q.‘.I.‘-U.L.\/-J Usnl~wdgLiil b... puuxlux} Int. ’BJU ULL‘WJFIMLJAVIr 1'an presented by J . Julio" K mt. ml lLLLLdIN has been accepted towards fulfillment of the requirements for —£’—-l—U—— degree “1%)! Majtyiprofessor Dateyzfiz/w‘l ‘: /7é 3 0-169 LIBRARY Michigan State University m 1171:0330! 03‘ 521mm BHWILITI m MING IX Two DIME!“ TASKS 17 Ian Sam n1. HEW hunt“ to film “at. University 5.: ”1‘1 mums o! the mum“ !» the up» of mammom Wormlcc' 1953 “SM I‘he reeeereh prone-e ef the dteeertettoh driee out of tepneetteee e! the 'etteelhe eetiettee' peetelatee of eederh lees-hie; theertete. ihe theory eeet eloeely eeretmsed 1e that e! Deny end Adele». the research wee eh etteept to detereue 11’ 'eettehntty' it e relatively etehle ehareetertette of en ermdee. eed e1ee it the different erunteee in e veep could he dtflerentteted ehe free nether wording to the degree to which they exhibited etinlue eettehtltty. The eeeoed point e! the reeeereh wee. it mum uttemeee in eettehutty exteted. did they contribute te the differenttel aunt: of era-nine to leere different um e! use prone-e. the epeettte prone-e uder tenettcattee were e eteple '1' eeee 1e whieh the te-he-leereed uteri-duties wee e. ten in e pertteeler dtreettee. end e 5 ehetee pent eedlfled '1' eeee in which the te-he-leerhed reeponee wee errerleee peeeece ire-theeterthex. throuhthefieheteepetetetethe‘eel her. In 'teth leer-1e; problem. the eottve wee ne‘er end the reverd wee teed pellete. he at for the experteeet were IIO eele elhthe rete tree m eeleer attuned ty the Michigan Stete hiveretty Psych- elea depot-em. All It were given the eeee meter e: trtele (681nm '3' me. that!» '2' Ieee) htheleareteg proneee which were let in e auterbheleeeed «up with he" the Be nentu the 'B' eeee ttret and the other half the '1' eeee tiret. After the Sc coupleted their triele on the first problee they were ehifted to the eeccnd prchlee. In edditicn to the learning problem cech 8 wee expceed to e equere ncee exploration tield for 6 ninetee u 3 eeparate ccceeieee. it wee the porter-choc of the to in the exploration use which yielded the 'eetieticn cecree' which were converted into rele- tive rem. end coupled with the relative teenage cf the Sc for their perfornencc in the 2 leaning eitnetione. In oddities eech 8 wet weighed on cech ecceeicn it explored the equre eeee. Iheee dete provided tb heeie for oil the ceepericonc in the experinertt. Ihe epeoific hypotheeee tested were ec fellcwet (1) rho exploreticn nemrec will pereit relie'tle renh- iec of the do tr. feet to clcw eetietere. (2) Blew eetietere will he aperior in porter-choc to feet cetieterc in the eieple '1' Inc. (3) net eetietere will he enperier in perforce-cc to claw eetietere in the ultiple choice point '1' me. n. eeelyeee er the reeelte hy cerreleticnel techniqeee indiceted that the tiret hypethecie wee confined. the second huetheeie wee pertielly confined. end the third hpctheeie eee not confined. Very few of the other coupes-item ette-pted chewed cry ciaificent decree of relationchip. end or e result the dieoeeeiea chapter cenieter Mill: in en enelyeie cf the deficieeeiec end indeqneciee of the experient. the general ehereeter c! the reenlte of the experieent were concietent with. though they provided no etrong mppcrt for. the petition that eetiehility ie e fundenental cherecter- ietie cl cranieu. end that the decreee of their eetiehility reletee cucwhet to their ehility te learn. pertinlerly in “male um, repetitive ei'teetionec. m mm 01 STIMULUS SELLER!!! 03 W IN TWO DIFFERENT TASKS By Janet Noreen Allen A MSIB Submitted to Michigan State Univereity in partial fulfillment of the require-ante ’ for the degree of 900203. 01' PHILOSOPHY Departnent of Pcycholcg 1963 MIMIC}! Since the doctoral dissertation serves as the keystone of anyoncs £9555} education. I respectfully dedicate this dissertation to the individuals who have had the greatest degree of influence upon no in the course of my formal edncation at all levels. To then met go a large ncasure of the credit for any of the good qualities of this work. hr the inadequacies and/or deficiencies I reserve full n- sponsi‘cility to myself. In that light I dedicate this work to the following alphabetically listed individuals: 8. H. Bartley It. 3. Denny O. H. Harley E. 3. Bolt A. 1.. Moore 0. A. Smith 11 bl immmrs The author would like to acknowledge the invaluable assistance given him at all stages of his fomal graduate education by the non- bers of his guidance committee. especially that of his najor professor I. Ray Denny who has been a steadfast friend as well as educator. I would also like especially to thank ferrancc Allen and l‘rank Beetle. who were very helpful in the early development of this problem but who were unable to continue on the committee until the conpletion of the problen. Stanley Rather. whose cautionary remarks were not always headed (and perhaps should have been) was of great assistance both on the dissertation and as a significant force in shaping w formal train- ing. Lastly. the two nenbcrs of the committee who stepped into the breach created by the loss of the two aforementioned committee Ienbers Allen and Beetle. Doctors Donald )1. Johnson and Charles H. Henley were gracious and. kind in their willingness to aid in the successful completion of e. problen in which they had no part of the early structe- uring. and no central interest in. 111 MELBOIGOEEEMS CHAPTER Page, WWIIONANDSTAMMOITBEPROBEM ........ l SUBJECTS......................... 10 APPARATUS........................ 11 mm........................ 15 RESULTS......................... 22 1330038103........................ I-I-l WYANDGONOLUSIONS................. 51 APPENDICES Appendix .A' (Historical Bammund)c e e e c e o e e 55 Appendix '3' (experimental Data). . . . . . . . . . . 66 BIBLIOGRAPHY O O O O O O O O O O O O O O O O O O O O O O C 7" iv I. II. III. 7. VII. VIII. LIS! 01‘ TABLES Page Correlations for Satiation. Activity. and Weight Loss 32 Correlations Between Satiation Rankings and Activity Weeeeeecceeeeeeeeeeoeee 33 Correlation: Between Weight Loss and Activity and Weight Loss and. Satiation (for the second and third Irplorationl).................... 3n Chi-squares for Errors '1' (first and Second Groups) and Brrors 'I' (run and Second Groups) . . . . . . 35 Correlations Between fetal Brrors '3' and i'otal lrrol‘l'I'....................o 36 Chi-square for Brrors '1' (Median or Above) and Satiation Banking (Above and Below Median) . . . . . 3'! Correlations Between Rankings for Satiation and IrrOflinLOmineralk'ceeecceeeeeeec 38 Correlation Between Rankings for Activity and ImuinLcaming!uks.............. 39 Correlations Between Bankings for Weight Loss and lrrorlinI-oamingllasks.............. l-I-O INTRODUCTION AND STATEMENT OF THE PROBLEM The present problem stems from the relatively recent int- erest in 'stimulus satiation' postulates in learning theory. The theoretical positions of Denny and Adelman (17). and of Glanzcr (22) are particularly germane in this regard. These postulates emphasise that, in responding to a part- icular stimulus. there arises not only an incremental strength- ening of the tendency to respond in the same way to subsequent exposures to the stimulus but also the dynamic properties of the evoking stimulus for this particular organism are themselves al- tered in such a way as to bring about the opposite response in the organism. Thus. each response to a stimulus on the one hand brings an increase in the probability that the response to sub— sequent exposures to that stimulus will be of the same kind. and on the other hand. the stimulus loses some of its power to elicit this particular response in this particular organism with each exposure of the organism to the stimulus. This can ultimately lead to the probability of the response being decreased. A quote of one of the postulates will suffice to indicate the general manner of formulation. The postulate immediately following is from the learning theory of Denny and Adelman. "Post. 2 Sensog Adaptation p; Sgtiatig With continued or repeated presentation e11 stimuli lose or partially lose the property to elicit a response as a decay function of the duration or frequency of presentation. IIfhe slaps of this decay function varies with the nature and intensity of the stim— ulus. i.e.. some stimli (food for a hun- gry animal. shock. etc.) are more resistant to adaptation than are other classes of stimuli. With the passage of time stimuli recover their capacity to elicit a responsc." This postulate then allows the Denny-Adelman theory to handle decrements in performance which may occur. particularly under massed trial conditions. Note however the difference between the above postulate and the inhibitory potential' or 'work dec- rement' type of postulate. Clark L. Bull in Essentials 3; _B_c_- _h_ag_ip_i_' (30) relates the inhibitory process directly to response and labels it 'QLE'- '... we find underlying it a variable of con- siderably wider application which is called inhibitog potential (Ir). This is believed to be a residual or after-effect left. apparent- ly. by all responses (B). which is in the nat- ure of a negative drive akin to tissue injury fatigue. or “pain“. It tends to inhibit the reaction potential. i.e. to prevent the occur- rence of the respgnse in question and possibly other responses." The exact formulation of the first two parts of Bull's postulate is as follows "I. d. Whenever a reaction (I) is evoked from an organism there is left an increment of primary negative drive (Ir) which inhib- its to a degree according to its magnitude the reaction potential (air) to that response 1. Benny. M.B. and Adelman. EM. "Elicitation Theory II: The formal Theory“. Unpublished Thegrotiggl 2322;. 1953. Mich- igan State University. p. 2 2. Bull. 0.1.. Basegtials 2; Behavior. Yale University Press. 1951- P- 73 '1. B. lith the passage of time since its formation. (11.) spontaneously dissipates approximately as a simple decay function of the time (t) elapsed. i.e.. 1,." I, x 10"“.0 the emphasis of these two theoretical systems is quite <11:- ferent. Bull focuses primarily on the response produced changes in the organism which inhibit further responding. and Dennybddelman focus on changes in the dynamic qualities of the stimulus which . elicit the response. It was not by accident that Dennyh-ldelman call their theory an 'clicitation' theory. In many ways the Denny-Adelman position is similar to that of Hichotte in his en- pcriments .on the perception of causality (32). Michotte held. as a result of certain experiments he carried out that certain stimulus conditions elicited a direct perception of causal re- lations on the part of his subjects. Iron the point of view of e behavioristic psychology the Bennyh-Adelman position seems preferable since it allows theoret- ical accounting for behavior without recourse to vague unspecifiable 'physiologising' about internal states of the organism. l'or then. the fact that a stimulus may lose its elicitation potential is a given (that is to say it is a primitive assumption of the theory). is such their theory represents a fundamentally dif- ferent sort of attempt to bring order into the relationships between the organism and its surrounds. 3. Bull. C. Lil- m 91 m. Iale University Press. 195]» P0 7 ._—.. Denny and Adelman's theory however is not to be construed as wholly stimulus oriented; the organism brings something to the relationship also. This. the current experimental problem follows directly upon the assumption that all white rats do not outlets to the stimuli of their environment at the same rate. but rather vary along a continuum from rapid to slow catiaters. The distribution of their satiability is assumed to bo-normalli Furthermore. Denny and Adclman's definition of stimulus includes the organism. It is an assumption of the present research that the “sat- iability" of individual organisms to stimuli is a general char- acteristic of the organisms. one that they bring with them to any situation they may find themselves in. It is further assumed that while moment-to-momcnt variations in the satiability of an individual organism do exist. the variations in satiability 333- 3293 organisms are larger than the moment-to moment variations in satiability BEEP. any one organism. Also. it is assumed that by placing an organism in a standard physical situation an estimate of the satiability of that organism can be obtained. To accom- plish this an Exploration task and an Exploration apparatus were devised by the present investigator. A simple preliminary experiment with an N of 39 white rats indicated considerable individual variability in the rate at which. 1+. Personal communication from 11. Bay Denny. Michigan State University and the extent to which. the 8s explored a small alley maze. i'he task and apparatus were then incorporated as a part of the present experiment. the purpose of this task being to estimate the satiability of the individual Ss. The next consideration was to relate satiability to success- ful performance of a given task. The successful completion of some tasks requires considerable persistence in repetitive act- ivity. while other tasks require considerable variability in activity for their successful completion. This. obviously. is also a consideration in the acquisition of these various tasks. l'hus. in learning a simple repetitive task such as a right or left turn in a '1‘" maze for food reward. a considerable amount of extremely similar activity involving very repetitious visual- motor stimulation of the organism is involved. Successfully carrying out this task involves turning the same way at the same choice point looking at the same alleys trial after trial after trial. And perhaps most important is the fact that all this takes place in an extremely limited space which by its very nature has a limited amount of possible differential stimulation in the first place. Contrast this situation with a much larger and more complex man. one with several choice points. a number of cul-de-sacs. and a relatively lengthy true path from start box to goal box. In the more complex situation the organism is required to engage in quite variable behavior to “solve” the maze. Indeed. it is in the entry into. and the withdrawal from the cul-de-sacs in the maze that the organism receives the greatest stimulus redun- dancy. Time. the successful acquisition of these two habits would seem to call for two different kinds of reaction patterns. In the simple '1" maze the most successful performer would be the 8 who persisted in repetitive activit'yin spite of a high degree of stimulus redundancy - an 5 who satiates slowly. On the other hand. in the complex maze would this help at all? It was the observation of the present investigator that the most successful performer would be the 5 who satiated most rapidly to redundant stimulation. and therefore eliminated entries into cul-de-sacs most rapidly. We need look no further than the welter of studies of explor- ation (1.2.3.h.5.6.7.s.11.13.1l+.21.3h.35.36.37.38.39.uo.h1.u2. 50.56.57. 58) to to. what a prepotent eliciter of approach reap- onses new and novel stimulus situations constitute for the organ- ism. Indeed. several experiments (2.10.19.58) indicate that Ss will learn to perform some response solely for the “reward." of approaching new and novel. or varied stimuli. In addition. the results of the exploration studies generally indicate a decline in approach responses as the situation in which the 5s are placed loses its newness or novelty. Thus. there is ample evidence to show that novel situations do elicit approach responses from various classes of organisms. and that behavior such as would be predicted on the basis of the Denny-Adelman Sensory Satiation postulate occurs with the prolonged or repeated exposure of the S to the same stimuli. The following experimental design and hypotheses were formulated in an attempt at an empirical valid- ation of the above argument. DESIGN AND HYPOTI-IESES The first element of the experimental design to be discussed is the exploratory task. As was stated earlier. this particular method for estimating the variation in satiability from individual to individual was devised in a preliminary experiment in which the writer posed the question of how to measure the difference if it did in fact exist. It was also desirable that the method of estimation take as short a time as was practical. In the course of letting the first few Ss of the preliminary experiment explore the maze. I noted that after about 5-6 minutes in the maze the 3s seemed to do little or no moving-about in the maze. so a cut-off time of 6 minutes was established and the remainder of the 8s were removed from the maze after that amount of time. The 8s were then ranked from fast to slow satiaters according to how long it took them to do 50% of the total exploration they carried out during the 6 minutes. This figure ranged from less than 1 minute to more than ’4 minutes. Within any one minute the number of units explored was assumed to be evenly spread through- out that minute. in S exploring 2 units in 1 minute was assumed to have explored l of the units in the first 30 seconds and 1 in the second 30: for 3 units in 1 minute. 1 in each succeeding 20 second period. and so on. l'or the current experiment the procedure was the same as that outlined for the preliminary experiment. with the exception that each 8 explored the maze a total of 3 times over a 2 month per- iod. The explorations were spaced about 1 month apart. The 3 rankings thus obtained were then compared using Kendall's Oo— efficient of Concordance (5h). The final composite ranking thus obtained. derived from the sums of the 3 ranks. reflects the best ranking of the Ss in the “least square“ sense.5 This de- rived ranking was then used in further cerrelational comparisons with the data obtained from the learning tasks in the experiment. The 2 learning tasks were run in counterbalanced order. half the 8s running first on the simple learning task. the other half on the complex. and the tasks reversed for the 2 groups after 13 days of trials. This method asstmes that the transfer effects will be in the same direction. and of close to the same magnitude from simple to complex task as they are from complex to simple task. and was adopted to economize on 3s by having each 8 serve as its own control. The specific hypotheses tested 5. Walker. H. 14.. and Lev. J. Statistical Inference. Henry Holt and company. 1953. p. 286 in this experiment were as follows: (1) The exploration measures will permit a reliable ranking of the 8s from fast to slow satiaters. (2) Slow satiaters'will be superior in performance to fast satiaters in the simple two choice discrimination task. (simple 'T" mass) (3) l‘ast satiaters will be superior in performance to slow satiaters in the complex. multiple choice point. two choice discrimination maze. (multiple "Y” maze) 10 SUBJECTS The Ss for this experiment were ’45 male Albino rats from the colony maintained by the Michigan State University Psychology department. During the course of the experiment ’4 of the 5s died. and l of them refused to run in the experimental mazes. Thus the experimental data are based on an N of Do. All of the Ss were approximately 10 months old at the start of the experiment. and Just beyond a year of age at its completion. 11 APPARATUS There were 3 major pieces of equipment used in this invest- igation: 1 for the exploration tasks. and 2 for the learning tasks. In addition to the major equipment certain incidental equipment and supplies were utilized. and these will be listed at the end of this section. Apparatus go}; _t_h_e_ egloration 59915;: The exploration field for the Ss in this experiment was a simple alley maze constructed of white pine. fir plywood. and one-half inch hardware cloth. The alleys of the maze were con- structed in the form of a square with fir plywood floors. white pine sides. and a hardware cloth top. The sides of the alleys were 5 inches in height. the inside width of the alleys was 3 inches. and the inside length of each side of the square. meas- ured midway between the 2 alley sides. was 1 feet. There were no culs in the maze. The hardware cloth top was a single piece of material hinged along one side of the maze to facilitate open- ing and closing it for the purpose of inserting Ss into. and re- moving them from. the maze. hiring the exploration tasks this maze was placed on a square table approximately 30:30:29 inches. Sufficient natural colored burlap. and also sufficient tar paper and black monk's cloth were used to drape the table completely around from a point about 9 feet off the floor to below the level of the table top. The apparatus was centered directly underneath 12 a 150 watt incandescent light bulb also suspended about 9 feet off the floor. The draperies. either the natural burlap. or the tar paper and monk's cloth. were tapered in such a manner that though they barely encased the lamp reflector at their upper attachment. they completely surrounded the table top just above their lower point of attachment thus providing a relatively homo- geneous visual extra-maze universe for the experimental Ss. Amatus £93; £1.12 gimp}; m 911933 discrimination w: The apparatus was a modification of a 'T' maze in which the goal boxes at either end of the cross-bar of the '1" were pivoted through 90 degrees. and paralleled the initial arm of the mass. In outline the maze looked like an upper case '1'. Six guillotine doors. 3 regular and 3 inverted (Operating upward) and powered by rubber bands were used to close off various sections of the mass at the following locations: One regular gravity powered guillotine door separated the start box from the rest of the maze. One inverted mbber band powered guillotine door separated the initial stem of the “I“ from the cross bar and the goal boxes. Continuous with the 2 sides of the initial stem of the '3' and in the cross bar were 2 inverted rubber band powered doors op- erating independently of each other which could separate either the left. or the right. or both halves of the cross bar from the rest of the maze. finally. at either end of the cross bar was a regular gravity powered door which closed off the goal box 13 from the rest of the maze. For this experiment the alleys were left in natural wood finish. The alleys. start and goal boxes were constructed out of 1 inch white pine. and the guillotine doors out of } inch fir plywood. The alleys and goal boxes were covered with g inch hardware cloth. the start box with a hinged white pine cover. All alley sides were 5 inches high. and all alleys were ’4 inches wide. The start box was 10 inches in length. the initial arm of the maze 12 inches. the total length of the cross bar 214 inches. and the goal boxes 10 inches in length. The food cups in the goal boxes were ordinary clear glass cups whose original purpose was to protect rugs. carpets or floors from permanent marring by the casters on the legs of heavy pieces of furniture. M $193 the gqmple; 15.119. 31M discriminatiog learnig 13515: The apparatus used for this part of the experiment was a modification of a maze originally constructed by Jensen in 1957 (31) as an apparatus for the study of latent learning. and fur- ther modified by Allen in 1958 (2) for the study of exploratory behavior. is modified for this experiment. the maze contained 5 different 2 choice. choice points. and a wrong choice at any 1 of the choice points leads the S into a cul-de-sac. Entry into a cul blocked further progress through the maze until the S retraoed his steps to the choice point and took the other arm. This maze is best described as a modified multiple ”Y“ maze. "Modified” because. rather than having the straight arms of the 114 ordinary 'Y“ maze. the arms leading from each choice point were bent in the middle. (Each arm was formed from alleyways which were originally the two adjacent arms of a hexagon.) The maze was constructed of white and yellow pine used in the alley sides. 3/’+ inch fir plywood for the floor. and i. inch hardware cloth used to cover the tops of the alleyways. The sides of the maze were 5 inches in height. the interior of the alleys were 3% inches wide and each segnent of the alloy was 1 foot long measured along a line midway between the 2 sides. For this experiment the entire maze was left in natural wood finish. Incidental equipment: Incidental equipment for the experiment consisted of a stOp watch for timing Ss. paper and pencil for recording their choices of path in the mazes. All the apparatus was housed. and the ex- periment was carried out in a single large brick walled room about 20:30 feet in size. illuminated by 6 large windows and 1t 150 watt incandescent reflectored bulbs. 15 PROCEDURE The general. features of the experiment were as follows. Each 8 took part in 3 exploration tasks and 2 learning tasks in the following order: First. 1 of the 3 exploration tasks: sec- ond. the second of 3 exploration tasks; third. I of the 2 learn- ing tasks; fourth. the second of 2 learning tasks; and last. the third of 3 exploration tasks. A more detailed account of the procedure. including a more elaborate description of the exper- imental tasks is given below. After the Ss were selected from the colony they were removed from the breeding and rearing area and placed in the experimental room in cages containing 5 animals each ( except for 2 cages. l of which contained 14 and the other 6 animals). This was done to permit the Ss some time to accustom themselves to their new sur- roundings. The number of animals in a cage was fortuitous in that it represents the number ordinarily reared in l cage. and all animals were left in the cages in which they were reared to eliminate the necessity of their undergoing a social reorganizat- ion along with their acclimatization to the experimental room. After the Ss had spent about 1 week in the experimental room the investigator commenced handling the animals. Tor 2 days the Ss were placed. 1 cagefull at a time on a bare table top about 30 inches square for about i hour. Handling of the 8s consisted of accustoming them to being picked up and replaced on the table. 16 This. incidentally. was their first experience outside their home cage for any length of time. and thus served to somewhat accustom them to being put in novel surroundings. During these 2 one-half hour handling sessions. the Ss were marked for positive identification. After the 2 days of handling the first experiment- al task was begun. The first experimental task was an exploratory task. The exploration maze described in the apparatus section was placed on a small. square table positioned directly underneath l of the 150 watt light fixtures which provided general illumination for the experimental room. Natural colored burlap was then draped from 2le wood supports at the level of the light fixture (about 9 feet off the floor) down around all 1|» sides of the table. This provided fairly homogeneous visual surrounds for the animal while it was in the maze. During all 3 of the exploration tasks the 150 watt light was left on at all times to provide a reasonably uniform level of illumination in all parts of the maze. Ss were then introduced into the maze singly. and allowed to explore for a 6 minute period. after which they were removed from the maze. The observation undertaken during the exploration period was a minute-byaminute account of the number of 'units' of the maze explored by the S. Thus. the data for this part of the experiment (and for all other exploration tasks) consisted of a set of 6 numbers signifying. for each of the 6 minutes spent in the maze. the number of ”units“ explored by the 8. On the 17 basis of these data the 8s could be differentiated 1 from another and placed along a continuum of satiability from fast satiaters to slow. After each 3 explored the maze it was weighed on a balance scale accurate within 1-2 grams. and the base weight of the S was recorded. The actual carrying out of the exploration measures took 2 days because the investigator wanted to restrict the span of time over which the exploratory behavior was measured to min- imize the individual differences which might occur due to changes in general activity level resulting from the well known diurnal activity cycle of the rat. After the 2 day period spent in meas- uring the explorations of the Ss. all Ss were placed on a reduced diet of 10 grams of Wayne Lab Blox per animal/per day for about 30 days. with water available 5g. lib throughout all phases of the experiment. It was during this 30 day period that 3 of the Ss died. After 30 days of a reduced diet during which time the Ss lost an average of around 15$ of their basal weight. the 8s were again introduced into the exploration maze. l'or their second (and also for their third) exposure to the maze the alleys of the maze were lined with black smooth textured construction paper. and the burlap drape was replaced by a drape of black tarpaper on 3 sides and. black: monk's cloth on the fourth. Other than these changes the procedure for the second exploration (and for the third) were identical to the procedure used in the first. 18 During the later stages of the deprivation period the regular diet of Wayne Lab Blox was partially replaced each day by the pel- lets later used as rewards during the learning tasks. After the second exploration task was completed all Ss were fed additional numbers of these pellets in the goal boxes of the two apparatus' used in the learning tasks. lash 8 was fed 10-15 minutes in each of the 3 goal boxes involved. Oompletion of this part of the experiment took an additional 2 days. and then the 3s were start- ed on the first learning task. Also at this time the daily ration of the Ss was raised to 12 grams of Wayne Lab Blox per animal] per day where it remained for the rest of the experimental period. The 142 surviving 8s were split into 2 groups by randomly assigning 9953; of animals to l task or the other. and on the completion of 13 days trials in l emeriment the 2 groups were switched and run for 13 days on the opposite task. On the third day of the learning part of the experiment the fourth death occur- ed among the Se. and a fifth 8 was discarded for failure to leave the starting box of the complex maze. At this time the final size of the experimental groups was reached. and data from the 1&0 remaining Ss are complete for all phases of the experiment. The 2 learning tasks. and the manner in which they were carried out was as follows. Task 1 was a simple 2 choice dis- crimination in a modified 'T" (or an '1') maze described fully in the apparatus section. The alleys were left in natural wood finish. and the discrimination to be learned was a simple left 19 (or right) turn at the choice point for a food reward. On the first trial for each 8 the food cups of path, goal boxes were loaded with the standard reward consisting of two .03 gram food pellets. On all subsequent trials the correct response was to whichever side the s visited on the first trial in the maze. 8s were given ’4 trials a day spaced as nearly as possible 15 minutes apart for the first 12 days of the task (a total of #8 trials). On day 13 each 8 was given 20 massed trials spaced as closely together as picking the 3 out of the goal box. reloading the food cup. replacing the S in the start box. and raising the door between the start box and alley permitted. The data collected for this part of the experiment was solely in terms of errors. that is. incorrect choices at the choice point. Time scores were not obtained. A non-correction technique was used on all trials on all 13 days. and Ss were left in the goal box after an error for 30 seconds. The basis for the choice of which arm of the maze would be correct was as follows. The investigator expected differences in the performance of 8s according to whether they were relative- 1y rapid. or relatively slow satiaters at the exploration task. It was hypothesized that these differences would favor more ent- ries into incorrect alloys by rapid satiaters than by slow sat- iaters during the course of the experiment. The procedure of selecting the side visited by the S on the first trial as the subsequently correct side anticipated a possible criticism that these ”errors“ occuring at a later time merely represent a visit by the S to the preferred side. Insofar as a single trial could be considered a measure of innate or strongly learned preferences on the part of the Ss. then visits to that side of the maze and/or turns in that direction were rewarded.as a matter of procedure. The errors expected to occur later would thus result from a visit by the S to the 'unpreferred' side. or a turn in the "unproferred' direction. At the completion of 13 days trials. a 2 day rest period was introduced during which time the 8s were maintained on a deprivation diet of 12 grams of‘wayne Lab Blox per animal] per day. After the two day rest period the second learning task using the modified multiple '1“ mass was begun. In this task each S was given 2 trials per day spaced as nearly as possible 20 minutes apart for the first 12 days. and on the 13th day a total of 20 trials massed as closely as possible together. .A modified correction technique was used for all trials in this maze as follows. Once the S left one ”I" of the maze. and pro- gressed to the next “Y”. he was barred from.regressing to a.pre— viously visited "Y” by stiff cardboard.blocks inserted downward into the alleyway at the choice point it had Just quit. However. if the 8 should choose the incorrect arm at any"Y' it was allowed to come back out of the cul-de~sac and enter the correct arm of the 'Y". indeed. so long as it did.not pass the next choice point it could.turn arOund and reenter the cul-de-sac once more. or for that matter he could repeat this several times so long as 21 the S remained within one 'Y“ of the maze. In this manner every 8 ultimately reached the goal box on every trial and.was allowed to consume a food reward of three .03 gram.pellets identical to those used as rewards in the simple learning task. The data collected in this part of the experiment consisted of error scores for each S on each trial. An error consisted of an entry into a cul-de-sac at least beyond the depth of the shoulders. and multiple errors at any choice point on any trial were possible. They were. however. infrequent. This completed the learning tasks of the experiment. and all Ss were then given 2 days rest during which time the depri— vation diet continued. Following the 2 day rest all Ss were given the third and last exploration period in the square maze. The last exploration task was carried.out with the identical apparatus and.procedure used in the second exploration task. again consuming a total of 2 days time. The section of the procedure dealing with the 2 learning tasks has described it in correct order for only half the se. The other half received the two tasks in the reverse of the order described above. 22 ESULTS In reporting the results of this experiment. the writer would like first of all to focus upon the exploratory task. In placing the same Se in the same task situation several times and measuring their performance again and again we are in effect asking a two-fold question; first. is the performance in quest- ion stable or changeable from instance to instance. and second. to what degree is the instrument chosen a reliable measure of performance! If the performance of each S is identical from occasion to occasion you can infer that not only is the perform- ance remarkably stable. but that the instrument used is perfect- ly reliable for measuring this performance. On the other hand. if your findings are more usual. and the performance of individ- ual Ss differs from occasion to occasion. the question of the reliability of the instrument and/or the stability of the per- formance is more complicated. Changes in measured performance may be the result of using an unreliable measuring device. or they may be the result of instance to instance differences with- in individuals in the measured performance. Perhaps more like- 1y what is involved is the measurement of an imperfectly stable performance with an imperfectly reliable instrument. The writer has assumed that “satiability“ is a characteristic of organisms which is relatively stable. and that one can properly describe certain organisms as “rapid" satiaters and other organisms as 'slow' satiaters. Fundamental to this assumption is the assertion 23 that. for a given number of organisms (n) it should be possible to order them from 1 to N as to their satiability if m possess Lb; promr mensurational device. No measurement device is now known which reliably measures this aspect of organismic perform- ance. This experiment then is at least as concerned with an attempt to assess the usefulness of one such proposed device for this task as it is in attempting to relate the satiability of organisms to other aspects of their performance. The net-- hod of assessing the satiability of the Ss was described in detail in the procedure chapter. Essentially. it involved the Ss' exploring a rather limited field on 3 separate occasions. The degree to which the Ss tended to explore the field in the same manner from occasion to occasion would indicate the stab- ility of this satiability. The degree of association of the 3 sets of satiation rankings was assessed statistically by using Kendall's W". or Coefficient of Concordancel. The results of this statistical treatment are sumnarized in Table l on page 32, The results indicate that there was a significant relationship between the 3 separate rankings of 8s for satiability. Il'rom the discussion by Siegela of Kendall's w". it is clear that the best estimate of the 'true' rank-order is obtained by rank- ing the sums of the 3 ranks for each 8 from 1 to It. thus creating l. Siegel. 8.. gonpgrametric Statistics. MoGraw—Hill. 1956. pp. 229-238 2 e 1b ide PP. 229-231 an a composite ranking for all Be based on their performance on 3 separate explorations of the field. Accordingly this was done. In addition to differing in satiability. the Ss differed on 2 other dimensions on which they could be ranked. general activity. and weight loss. It was thought that individual dif- ferences along these two dimensions might contribute some of the variance in the 2 learning situations. and accordingly. they were analysed in a manner similar to the analysis carried out on the satiation rankings. General activity is defined. in this experiment. as the total number of units of the maze entered in a 6 minute period. Since satiation ranks are based on the percentage of this same 6 minute period that it takes an s to go through A» of the total . units that the 8 enters into. it is quite possible that the 2 dimensions of satiation and general activity may be highly rel- ated. though it is not necessary that they be related at all. In a preliminary experiment with an I of 39 Allen (3) concluded that general activity and satiation were essentially independent (rho-.05). To determine whether in the present experiment the same independence of the 2 dimensions was observed the relation- ship between them was statistically analysed by a Spearman rank- order correlation}: In addition. the degree of relationship between the 3 separate rankings for general activity was assessed 3. Siegel. s. Nonmrametric Statistics. McGrawh-Hill. 1956. 25 in the same manner as was the relationship between the 3 sep- arate sets of satiation ranks. The results of analysing the relationship between the 3 general activity rankings are sum- marized in Table l on page 32 and the results of correlating the satiation rankings and the activity rannngs are summarised in Table 2 on page 33- l The significant relationship between the 3 separate general 1 activity rankings indicates. Just as it did in the case of the satiation rankings. that the best estimate of the 'true' ranking for general activity is obtained by ranking the sums of the 3 separate ranks for each 8. and thus creating a composite rank- ing of the Ss from 1 to I on the general activity dimension based on the results of all 3 explorations of the field. In Table 2 on page 33 are the r.'s obtained from correlating the rank-order for satiability with the rank-order for general activity. The 8s were ranked on these 2 dimensions according to the following plan. In the satiation rankings the most rapid satiater was ranked number 1. and in the general activity rankings the most active 8 was ranked number 1. The obtained correlations thus indicate that for every exploration of the me rapid satiation was associated with low activity levels. lhile the magnitude of the obtained correlations indicated a cer- tain degree of relatedness between the dimensions of satiation and general activity there is no support for the contention that they are unidimensional. IThe obtained correlations are far too modest to support any such contention. At this point then we were possessed with 2 related but separate dimensions of organismic performance. either or both of which might be related to the ability of the organism to learn. On each occasion that the Se explored the square maze they were also weighed, and we thus possessed measurements over 2 periods of time of the amount of weight lost by each S. This measurement has obvious possibilities as a determiner of the 8s performance in a learning situation via its relation to pos- sible differences in motivational level. That is. an 8 who had lost but 10% of its body weight might be said to be less highly motivated than an S which had lost 20% of its body weight. This might be a rather academic relationship if all Ss had lost about the same amount of weight after corresponding deprivation per- iods. but in point of fact they did not. There was a consider- able range from the least percentage of base body weight lost to the greatest percentage lest. For the first weight loss period. that is. the period Just prior to starting the discrim- ination learning problems. the range of base body weight lest was from 3.18% to 22.18% hr the second period, that is. until Just after the completion of the discrimination learning tasks the range was from 11.63% to 38.32%. A further finding of int- erest was the result of correlating the percentage of weight lost in period 1 with the percent lost in period 2. The obtained 27 rank-order correlation of .069 indicates no relationship between the relative amounts of weight lost during the 2 periods. Because of this lack of relationship between the percentage of weight lost during the 2 periods no attempt was made to devise a comp- osite ranking. Weight loss rankings for the first period which ended Just prior to commencing the first learning task were com- pared to performance on the first learning task. Weight loss rankings for the second period which ended Just after the com- pletion of the second learning task were cempared.with performance on the second.task. The only significant correlations indicated in the results of comparing weight loss rankings with.other’meas- ms of performance (Table 3 page 31L were between weight loss rankings and “I“ maze performance on the first learning task. plus a single significant correlation between weight loss rankings and general activity rankings. This also scoured during the first weight loss period. is for the remaining statistical tests. very few correlations attain statistical significance. The correlations between sat- iation rankings and "I" maze performance are in the predicted direction of more errors by rapid satiaters. but only 1 is high enough to be statistically significant. Regarding the relation- ship between satiation and “I" maze performance, none of the correlations attain statistical significance. and all of them excepting l are in the opposite direction from that predicted. The activity ranking had no significant relationship to any aspect of performance in the 2 learning situations. but was re» lated to weight loss rankings for the first period. the obtained rank-order correlation of .1108 was sinificant beyond the .01 level. In Table 1% on page 35 appear the results of the analysis of the transfer effects found in this experiment. .L.dhi-square analysis”. indicated,significant degrees of positive transfer from.'!' mass to 91' mass. and significant negative transfer from “3" mass to '2' mass. the Chi-squares of 12.1 and M.9 being significant beyond the .01 and.between the .05 and .02 levels respectively. An inspection of the mean error scores for the appropriate groups indicates clearly that these trans- fer phenomena are not merely the result of the particular direct-r ion adopted.en trial 1 in the 91' mass. but represent other aspects of their mass perfonance. It might be argued.that what the auther'has called "transfer- effect' are really'nothing of the sort. It is possible at least at first glance to propose as the main factor involved nothing more than a sampling error in assigning 3s to one or the other group and thus. since one group is superior on both tasks. and the other inferior one need not invoke the concept of transfer at all to explain the results. This seems unlikely. and.the results summarised.in.Table 5 on.page 36 indicate that an explanation as simple as sampling error is not enough. If one 29 group is composed of superior subjects. and the other inferior. it would seem predictable that the performance of the Se in a group on the two tasks would be significantly correlated. In fact neither of the two correlations are significant which in- dicates that the performance of Se in the 2 situations is un- related. Table 6 on page 37 indicates that the clearest difference between rapid and slow satiaters in the '1' maze occurs in the last half of the trials. This is precisely the place where they were enected to appear as indicated in the introductory chapter. a Chi-Square analysis with a median or above split in errors. and a median split in satiation ranking yielded a Chi-Square of 11-.9 with 1 degree of freedom which is significant at between the .05 and .02 levels. An analysis of the some portion of the trials for the 'I' maze indicates no significant difference between rapid and slow satiaters. Tables 7. 8. and 9. on pages 38. 39. and 1&0. summarise the results of the correlations between weight loss. activity. and satiation rankings with performance in both the '1' mass and the '1' mass. In all these tables there were 3 significant correlations. one reflecting the superiority of the slow satiaters of one of the two groups over the fast satiaters of the same group in '3' maze performance. This difference was indicated far better by the results of comparisons which were summarized in Table 6. as the comparisons in Table 7 included errors made on all trials of the 30 learning task. and thus included errors made during the very early trials when any 8 regardless of satiability would be necessarily be making errors. The remaining 2 significant correlations occur in Table 9 ans occur in correlations of weight loss with errors in the '1' mass for the group running the '1' mass first. The argument for the relevance of these two correlations in the overall picture is involved and derivative but. it seems to the author valid. The argument centers around the results of a number of the size-of- reward studies. Typically they show that varying size-of-reward affects performance in complex learning situations but not in simple. and the direction of the effect is for larger rewards to yield better performance. In this experiment it is true that we did not have different sizes of reward but there were differences in the amount of weight lost by individual 8s. The next link in the argumentative chain is that because of differing weight loss among Se we have differing motivational levels and for this reason the rewards. while not differing in physical size would be function- ally different in size because of the differing motivational levels. It is tempting to speculate on the relationship between this argu- ment made for subhumen organisms and the results of experiments with human children as Se in which the experimental task was set- imating the sins of coins. The poorer and presumably more deprived (in a monetary sense) children typically show greater positive size distortions (estimate the coin size as larger) than do the children 31 of economically more well off families. lhile a disappointingly small number of the comparisons attained statistical significance. the general character of the results of the experiment were consistent with the position that satiability as a fundamental characteristic of organisms is re- lated to performance in a learning task. at least in simpler situations. and that slow satiaters do perform significantly better in 9!“ mass learning tasks than do rapid satiaters. TABLE! OWIONS FOR W103. LOTIVIH. AND WEIGHT LOSS Kendall's '1' for 3 Satiation Rankings Iendall' s '1' for 3 Act ivity Rankings 3 for [eight Loss 1 and Weight Loss 2 I r or r. :3 Sig. 13-0 .501 60.12 .05 ) .025 ’40 .941; 77.28 .01 > .001 no .069 ms. , 33 mu GOWIONS mm SLIM!“ W m ACTIVIM MEGS I r first Imploration ho -.5uu Second Exploration MO -.503 Third.lxploration. no -.350 Composite Rankings 1+0 -.538 It! h.05 3.63 2.32 3.95 Sig. beyond .001 beyond .001 .05 ).02 beyond .001 31} um III OWIOIS mm EIGHT 1.058 m AMI“!!! m) axon LOSS m) warms (to: the Second and. mm Between Weight Loss and. Activity Second hploration Third hploration Between Veight Less and Satiation Second hploratien Third lxploration kplorations) r Sig. .hoe .01 > .001 CW '08. .230 1.8. 0%! '08. 35 TAN-'17 mequms 1'03 mons '1' (ruse 1m: steam) 030m) 1m) mass 'r' (1138! mm 53001111 camps) s :2 a: Big. Total lrrors '1'.- no 12.1 1 beyond .01 Total nrrors '1' to 11.9 1 .05) .02 36 m1 OWIOHS 31mm row. ERRORS '1' AND EMAIL 133038 ”I” 1! r Sig. GROUP '3' Iirst - '1' Second 20 .1611 3.8. '1' Iirst - '3' Second 20 .321 1.8. 37 2131871 CHI-SW rm mono '3' (mm on non) m 3121mm nmm (non m: mow mm) s 12 0: Significance no 11.9 1 .05) .02 39 1mm VIII omnzmmmnmnmsmmmu mmansnnmmusxs ammun'r'mrmr lrrorlree Trialein'Y'llase SpaeedTriallrrorsin'T'llase naseedTriallrrorsin'I'Mase Totallri'ersin'f'lese 0301? me “I“ m noon lrror Tree Trials in "I" was. Spaced 221.1 Errors in Ir- a... leased Trial Errors in '1' Ease Total lrrors in "I" Hare 0301!? m0 '1' am HRS! Total lrrore in '3' lace 0300? mm} '3' was SECOND Total lrrors in '3' Haas 8888 8388 4339 -.158 -.203 -.186 .181 . 019 Big. 1.8. 3.3. 1.5. '08. 3.8. 1L1 DISOUSSIOI As was reported in the Results chapter. the number of stat- istically significant findings was rather small. ll'or this reason any discussion of the experiment must include consideration of the faults. flaws. or inadequacies of the research. That is. of course. true of almost any research undertaking but it is of particular importance in an experiment which yields so few positive results. Therefore. in the main. the experiment will be discussed as to its inadequacies for demonstrating the valid.— ity of the propositions which were derived from the theoretical framework of the Denny-.Ldelman elicitation theory. W W a 1m .21 5.1a 22m mama mania: Borlyn- (6). Glanser (22). and Denny and Ldelman (17). .11 refer in their theories in some form to the positive relationship between novel stimuli and approach responses. llsewhere. this writer has pointed out reasons that seem to males the Denny-Melman position preferable. and they will not be recapitulated here. but g of these theorists have as a consequence of their theory. taken the position that exploration of the environment is an en- ceedingly prominent behavior in the eccnow of the organism. In addition. all of the theorists mentioned above talus theoretical account of the self limiting character of exploratory behavior. Dew and Adel-an do so with their postulate of sensory satiation. The rapidity with which this process of satiation occurs in an )44?. individual organism is not. however. to be taken as a constant decay function for all members of a particular species. It. libs any number of other population variables is distributed ever a range of different values. Denny assumes that the frequency distribution of these values follows a Gaussian none. tio 1. me being the case. a standardised situation in which the plwsical characteristics of the environment are rcpt as nearly as possible constant should enable the investigator to determine the degree of this satiability in each member of any sample of organisms he chooses to expose to the standardised environment. Indeed. this is precisely the rationale given at an earlier point in the paper for adopting the procedure that was adopted. However. there exists virtually no empirical evidence for this preposition. It is true that Allen (3) in an earlier unpublished experiment discovered a rather wide range of what he called 'satiability scores' were obtained from a group of 39 albino rats. These rats were allowed to explore a small mass for a short time with the mass. exploration time. and scoring procedure identical with those used. in the present experiment for the eqloratory task. On the basis of this earlier experiment. and. considering the lack of information about the stability of this satiability factor over time. the decision was nade to collect data on the satiability of the same Se to him similar environments over a somewhat extended period of time. At 1. Personal communication from 11.11. Denny. Michigan State University h3 the time it was realised by the writer that this might p_9_s_s_i_b_l,y be a weakness in the experimental procedure. but it was adopted as a simplified procedure to the alternative of attempting to standardise a large number of small mases and provide a new and different mass for each occasion on which exploration was meas- ured. When one takes into account the principle of stimulus gen- eralisation. it is perhaps inappropriate to consider a sit- nation that an organism with any appreciable sensory past en- counters to be completely novel in the fullest sense of the word. It is. however. perfectly proper to speak of the greater or lesser degree of novelty that a stimulus complex is likely to possess for a given organism. If novel stimli decline in their approach-eliciting value as stimulus novelty is lost. them any program which includes multiple exposures to essentially the same field neggssarily involves a progressive decline in the novelty. and in the approach-eliciting value of the stimuli from occasion to occasion. Denny and idelman hold forth the possibil- ity that the full amount of the approach-eliciting value of the stimulus is potentially recoverable if the interval of non-exp posure is long enough. but their own research typically shows persistence of some loss in eliciting power over rather extend.- ed periods of time. Specifically then. any further experiments along the lines of the present investigation could benefit from the use of a number of distinctly different mass situations to assess the satiability of the Ss. There is still another problem in this area which deserves some comment. The mass chosen for use in this experiment was of square configuration with 1|» serially connected. equal length ams Joined at right angles. The more used had no choice points in it. and no cul-de-sacs. This layout was different from the layout of the mases used in the learning part of the experiment in that they had both choice points and culs. Informal obser- vation of fairly large numbers of rats by the author has con- vinced him that the behavior of individual animals confronted with the blank wall at the end of a cul-de-sac is as variable. relatively speaking. as it is likely to be in any one of numerous other occasions. Since this variability in behavior was not systematically investigated. there remains the strong possibility that it might have had a discernible effect on the rated satiabil- ity of an individual S. The inclusion of mases shaped like a cross. or like a “I“. both of which would have choice points and oul-de-sacs. as well as the inclusion of a circular runway. which would lack not only choice points and cul-de-sacs. but even the corners possessed by the square mass. would be advise- able in future attempts to arrive at the satiability quotient of individual organisms. Lastly. with regard to the general problem of relating em- 1.5 ploratory behavior to behavior in a learning situation. Allen in 1958 (2) commented upon the confounding of real exploration of an environment with locomotion through that same environment. Some exploration situations require by their very nature a great deal of point-to—point locomotion Just to traverse. while others. such as the exploratory field used by Velhsr in 1957 (57) require but little point-to-point locomotion. and concentrate on a high- ly varied visual-tactual environment. In actual fact the mass used by the author as an exploratory field was smaller than that used by Walker in the experiment cited above. but there is a great deal of difference in the emphasis of the two fields. The Allen mare requires point-to-point locomotion even though in a limited field. there being little emphasis on variegated environ- ment. In fact the attempt was to make the physical surrounds as uniform as possible. Velhsr. on the other hand. in a field only slightly larger. obviously emphasised variegated textures. shapes. and brightnesses in his exploration field. In this experiment the decision to use the type of exploration field that was used was predicated on the assumption that it more nearly duplicated the conditions of the learning problem whose outcome we were interested in predicting. Obviously. there is no logical necessity that this be the case. It is Just as logical to predict the satiability of individual animals on the basis of the rate at which they cease to explore a highly diversified field as it is to predict on the basis of how quickly they cease to explore a highly uniform field. It is 1L6 perhaps likely that the satiation rate will be slower because of the increased richness of the visual/tactual environment. and. that the assessment of the satiability of an individual S might be somewhat more elaborate. but any significant improvement in the rating of Se satiability would be well worth a more couplicated procedure. gimme—.2240“ t gamma amazement In this experiment the Ss were on a severely reduced diet for two months from start to finish. It is quite possible that the lengthy period of reduced diet so increased the saliency of the hunger motive that it diminished the importance of the sat-e iation variable. thus obscuring the effect of the variable os- tensibly under investigation. the I (comment). The author feels that a much milder deprivation diet would obviate this situation but is unable at this time to do more than speculate that this procedure will indeed increase the effect of stimulus satiation on the mass learning performance of the Ss. ‘ mmmmm t mmmmnmmme The number of Se with which this experiment began was a rather modest ”-5. the II with which it finished an even more modest l#0. Since the main variable under investigation is assumed to follow a Gaussian frequency distribution in the population. it is possi- ble that attempting to select 35 to ’40 So from the extremes of a much larger population which had been rated for satiability would enable the investigator to secure a sample more extreme in their n7 satiability than was actually obtained in this experiment. If. as is possible. such factors as hunger tend to mask some of the satiation effects in a learning situation. selection of extreme indivihals from the population might allow a larger residual of the effect to show up. the drawback to this procedure is the uneconomical use of experimental animals and of time. Also there is the possibility that no matter how slowly. or how rapidly. an 8 satiates under conditions in which it is being fed ad lib. m deprivation diet might wash out the effect nearly completely. levertheless. it is a procedure which might be extremely useful. and ought at least to be tried. Lia mu: m 21 22W alien as; am; mu is as Win am: !he implicit assuption of the writer about the discrimin- atory power of rapid as against slow satiaters in the planning of this experiment was that the rapid satiater was more sensi- tive to smaller differences in its environment. In fact. this was thought to be the essential difference between them. Ior this reason the decision was made to leave all the alleys in the mass in a natural wood finish. i'he thinking behind this decision was that the fast satiaters would be affected more by the minute differences in stimuli throughout the maze. and they would thus be more affected by the relative differences in the novelty of stimuli. Because of this their behavior would be more variable than the behavior of the slow satiaters who were not so sensitive to these small differences. It is possible however that the discriminatory power of the animal is not re- lated. or is only slightly related. to its satiability. If this is the case then names in which the stimulus values of the different arms are made more discriminently different would per- haps allow the satiation variable to show to greater effect. If in addition to turn direction the Ss also had available the cue of brightness difference (as they would have 1: 1 arm at each choice point were painted white and the other black) the learn- ing task would have been easier because of the multiple cues all pointing the same direction. Also. with greater plwsical difference in the stimuli. the novelty of an arm not recently entered would have greater saliency. and thus greater approach- eliciting value. this procedure must be viewed with caution however since it might also. in an experiment in which each 5 served as its own control as they did in this experiment. tend to increase the already mnsiderable and unwelcome interaction effects. This could well have the effect of further obscuring an already difficult interpretation even further. One last change in the physical environment of the learning mazes would seem desirable. fhe surrounds of these mazes ought to be made more hohogeneous in the same manner that the surrounds of the exploration mazes were. That is. a system of drapes skirting around the periphery of the maze ought to be erected. 1‘9 Some of the lack of positive results might well be attributable to the fact that the rapid satiaters. who in general did not perform as expected. failed to do so because. satiating rapidly to the redundant stimulation of the interior of the maze. they turned for stimulus variety to the relatively rich and uncontrolled extra maze environment. transfer W: One of the findings of this experiment was that there was positive transfer from the '1' mass to the '3' maze. and negative transfer from the 'E' maze to the 'I' maze. rho positive trans- fer from '1' to '2' was so great that 3 of the Ss run through ‘ the 2 learning tasks in that order did not make a single mistake in a total of 68 trials. and 2 more Ss made but a single error. 'i'his seems to result as a function of 2 interacting circumstances. first. the adoption of the procedure of making the correct arm of the '1' maze uniformly correspond to the side of the maze visited by the S on its first trial in the maze. and second. the speed of running through the alleys that the 8 had built up over nu trials in the multiple '1' maze. the running time of man of the Be who had the '3' maze problem second did not exceed 2 or 3 seconds on their very first trial in the maze and. further every 5 was well accustomed to consuming the food reward immediately upon entering the goal box. This combined with the fact that which ever way they went on trial 1 was the correct way. the extremely short time interval between the 50 critical turn response and. the receipt of the reinforcement could. well explain the rapidity with which learning took place. 51 SUMMARY AND CONCLUSst The research problem examined in this dissertation stems from the recent interest in what might be termed 'stimulus sat- iation' postulates by modern learning theorists. The particular theoretical position from which this problem is examined is that of Dow and Melman (17). The essential point of the theory under investigation deals with the theoretical postulate on sensory satiation which states that the approach eliciting value of a stimulus complex is lessened for an organism with continued exposure of the organism to a particular stimulus cellplex and that the speed of this satiation varies in the pepulation accord- ing to a Gaussian function. / The argument is put forth that in performing different kinds of tasks it may at times be advantageous if the organism satiates rather slowly. and that at other times a rapidly sat- iating organism may have the advantage. The specific test of this argument was accomplished by rating the Ss for satiability on an exploratory task in a simple small maze. then having the Sc perform in 2 different learning situations in counterbalanced order. One of the learning tasks was a simple turn response in an: '3' maze. the other was a complicated maze problem using a 5 unit modified multiple '1' mass. The Ss for this experiment were its male Albino rats from the Hichigan State University Psychology department colony. with 52 the final I for the experiment shrunk to no because of 1+ deaths and l refusal to run. Three pieces of apparatus were used in the experiment. an exploration.maze. and 2 different mazes for the 2 learning prob- lems. The eqloration mass was shaped like a square. with 1t alleys each.a foot long connected.by 90 degree corners. One of the learning mazes was an '1' mass (a 'T' maze with the goal boxes reflected.90 degrees from the cross bar on the 'T' and parallel to the initial stem of the maze). and.the second.maze was a modi- fied 'I' maze with 5 choice points in a B.I..h.h.1. pattern. The particular modification.of this 'I' mess is attributable to its origins as a.multiple hexagonal maze. Each arm of the '1' con- sisted of 2 adJacent arms of the original hexagons and was bent in the middle. whereas in a conventional '1‘ mass each.arm is straight. All 3s were permitted to explore the square mass 3 times. 6 minutes each time. at approximately 1 month intervals. On the basis of their composite performance they were rated individually along a continuum from fast satiaters to slow. Their ranking on this continuum was then compared. using Spearman rank order cor- relations. with their performance in each of the 2 learning tasks. The 2 learning tasks were run in counterbalanced order. half the 8s starting with the '1' mass problem. and half starting with the multiple '1' maze problem. After 13 days of trials all Be were switched to the opposite problem. Of the 13 days of trials 53 the first 12 days were all relatively spaced trials (1; per day in the '3' mass. and 2 per day in the '1' mass). The last day in either mass consisted of 20 trials massed as closely together as possible. All Be were then ranked according to the number of errors made on the problem. and these rankings constituted the second variable in the correlations referred to in the previous paragraph. One additional measure was taken. the amount of weight lost after 2 different length intervals of a deprivation diet. All Be were weighed each time they explored the square mass. thus data was obtained as to the Ss weight on 3 different occasions (I) The base weight. or weight prior to any time on a deprivation diet: and (3) The weight of each 3 at the completion of tin experiment approximately 1 month after the second weighing. All Be were ranked for the percentage of their base body weight lost for each interval. This then constituted the third major variable considered in the experiment. It was considered im- portant because of its possible relationship to 'mctivational level differences which might lead to increased variance in per- fonance of the Se in the learning tasks. from the results of the analysis it is apparent that the presumptive motivational difference contributed more perhaps than any other variable to the perfonance variance of the 8s. 5’4 1. disappointingly low number of the correlations and other analyses attained statistical significance. weight loss and sat- iation being the only significant variables. and these only par- tially so. is a result. the discussion of the results consisted mainly in an analysis of the deficiencies and inadequacies of the experiment. Basic to the design of the experiment was the assumption of consistent transfer effects. and the experiment did not yield consistent transfer effects. The transfer from multiple '1' to '3' was positive. while the transfer from '1' to multiple '1' was negative. Also the positive transfer from '1" to '1' was more substantial than the negative transfer from '3' to 'I'. The general character of the results of this experiment however are consistent with. though providing no strong support for. the position that satiability is a fundamental characteristic or organisms. and that satiability is related to their perform- ance in a learning task. at least in the simple '1' mass. in the predicted direction. Except for autocorrslations the significant correlation between errors in the '3' mass and. satiability rank- ing was the highest attained in the experiment. 55 APPENDIX 'A' In discussing the historical background of this problem. three lines of research will be of primary importance; (1) Spontaneous activity studies as they relate to learning. (2) Genetic studies investigating the inheritance of 'intelligence' in rats. and (3) Exploration studies. Other studies which seem relevant to the problem will also be reviewed. but not as extensively. One of the latter sort of studies was a paper in l9ll2 by 3. tuppuswaq (33). who studied a single 'bachard' rat quite intensively. hoping to discover if the backwardness of this rat was all-per- vasive. or if perhaps he might be able to find some area in which this animal equalled or exceeded the average performance of his fellows. He discovered no such area of performance. this one S being in every respect inferior to all other Ss tested. Insofar as cerrelational studies of the performance of Se have been carried out at all. they appear to be correlations be- tween performances in quite similar situations. and. all the core relations are positive and quite high. Ounmins. Molnar and Stone in 1932 (12) report a correlation of .6 between abilities to learn several kinds of complex mazes (multiple 'T'. multiple elevated 'T'. etc.). Tryon in 1931 (51) reports a correlation of .8 for errors made by Ss in two different 'T' mazes. There was no attempt here to assess the learning ability of the Se in two fundamentally different situations. This is also true of the 56 genetic studies on inheritance of intelligence (21%. 25. 26. 1$6. 52). In all of these experiments the test situation was the pen. formanee of the Se in relatively complicated mazes. Heron and various others have. however. reported several measures of functioning in rats other than the acquisition of a response in their long term program studying the inheritance of intelligence. Heron and rugend in 1936 (29) reported that the Heron 'bright' strain of rats had a higher mm than the Huon 'dnll' strain. Heron in 19110 (27). and Heron and Skinmr. also in who (28) found that Heron's maze bright animals extinguish a learned habit faster than the maze dnll animals. Heron. in a 1935 article (26) reported the following pro- cednre used in the creation of the two strains of rats he created. He first tested a randomly selected group-of animals in the Heron automatic maze. then. selecting the best and worst per- formers in the maze. he inbred them. best to best. and worst to worst. The offspring were also tested in the same maze. and the best performers in the 'best' group were bred to each other. with the worst performers in the 'worst' group also being treat- ed analougously. This procedure was followed for several genera- tions (best bred to best. and worst bred to worst) until at last there was virtually no overlap in the performance of the 'bright' and 'anl' groups. The important thing to note from this study so far as the present problem goes is that the orig- inal group. selected randomly. contained individuals nearly as 57 extreme in their performance as the Se in the selected groups after several generations of inbreeding. ‘ Heron was also involved in the genetic investigation of other traits in the rat. such as activity. mindquist and Heron in 1935 (#5) reported the results of a comparison of the mass learning abilities of two groups selectively bred. using pro- cednres analogous to those outlined in the proceeding paragraph. to produce an ‘active' and an 'inactive' strain. Individuals from these two strains as well as animals from the 'bright' and 'dull' strains were tested in the Heron automatic mass. with the following experimental outcome. The 'active' and the 'bright' strains were comparable in. their ability to learn the maze. though the p'bright' animals were somewhat better than the 'active' animals. Both of these groups were better than either the ‘inactive' or the 'dull' 'Ss. In this particular experiment the bright—dull strains were in the fourth generation of the breeding program. and. the active-inactive strains were in the seventeenth generation. . Activity measures have not always shown such a consistent relationship to learning ability however. 1:in1 in 1925 (35) using lambs as Se and a simple mass as an apparatus for a learn- ing experiment found that spontaneous act ivity. measured by means of a pedometer attached to the loreleg of the lambs showed no relationship to maze learning ability. The activity measures 58 were taken over periods of various lengths between four and thirty-seven days. Shirley. in 1928 (M7) found maze learning to be only slightly related to activity. but Tuttle and Dykshorn. also in 1928 (53) round activity level and learning ability to be closely allied. Tuttle and Dykshorn were actually studying the effects of certain physiological changes. brought about in the rats by operative techniques. on spontaneous activity and learning ability. One must be cautious in comparing their results to other experiments of the same general type because of the extensive changes wrought in the physiology and behavior of the 8s due to the castration operation. Lee and YaniBuskirh also studied this problem in 1928 (31*). and concluded that the spontaneous activity level of their Se was unaffected by the changes in m brought about by the thy- roidectomy which they performed on their Ss. There is possibly an interesting relationship between the research of Lee and Van Huskirk. the research of Heron and Ingend and that of Hnmdquist and Heron. Heron and Tugend found the Heron 'bright' strain had a higher M than the 'dull' strain. hindquist and Heron found that the 'active' and the 'bright' strains were fairly comparable in mass learning ability. and Lee and YanHushirk found that m and activity were not related. There is thus a possibility that the findings of Heron and Yugend regarding the higher m of the 'bright' strain are entirely 59 fortuitous. Allen in 1959 (3) in some unpublished research also determined that. for animals fed a normal diet. stimulus satiation and gen- eral activity are not related. 1 rank-order correlation between general activity in a small maze (defined in terms of the number of arms visited in a standard time interval). and stimulus sat- iation (defined in terms of the rate at which the animals slowed up and/or ceased their movement through the maze). was --.05 with an I of 39. This last experiment by Allen was importanttto the overall planning on the design of the experiment carried out in the present problem for it strongly indicated a possible method of differentiating Ss along lines which are theoretically related to individual differences in ability to learn. and further. this method is apparently not confounded with differing general act- ivity levels. A fuller discussion of the possible effects of confounding these two variables is contained in Allen's 1958 unpublished in thesis (2). essentially. the possible effect is as follows: In small mazes such as the one used by Welter in 1957 (57). not much locomotion is required in the exploration of the maze. while in large mazes. such as the one used in Allen's Ii thesis. or in Dashiell's classroom demonstration maze develop- ed in the twenties and reported in 1925 (in). a great deal of peimt-to-point locomotion is required for the exploration of the maze. Thus. when large mazes are used to evaluate stimulus 60 satiation. the satiation measure may be confounded with 0.111315 ences in general activity level among the 3s which will lead to more or less movement through the maze. and these differences in movement through the maze need not be at all related to stimulus satiation as such. There are also differences in the maze behavior of animals related directly to their motivational state. the extremely com- mon experimental procedure of depriving animals of food for some period of time to motivate them to learn the maze for a reward of food being an obvious example of this feet. When. however. we remove reward from the situation and do an experiment in exp ploratory behavior instead of one in learning. evidence does not support the contention that the resultant behavior of the Se is so directly related to their motivational state. primarily because of the confounding of exploratory behavior and locomotive behavior mentioned in the proceeding paragraph. Montgomery. in 1953 (39). and ildsrstsin and l'ehrer in 1955 (1). also rehrer in 1956 (21) came to distinctly different conclusions about the interrelations of these variables. Alderstein and J'ehrer. and Iehrer. concluded that food deprived animals explore much more than animals fed ad lib. while Montgomery concluded exactly the opposite of this. To quote Montgomery. his study gives evidence that exploratory 'drive' is a '. . primary drive which undergoes a decrement in the presence of other primary drives."1 To cite l. Hontgomeryj I. 0. “The Effect of Hunger and Thirst Drives upon Exploratory Bohavior'u 1m 2; We 59.4 W 2mm 1953: '01- #6. 10- 319 6l rehrer. hunger W exploratory drive. ' . . . in the sense that hungry animals are more likely than sated ones to leave familiar foodless territory."2 The most important variable operating to produce these dif- ferences seems to be the maze itself. Montgomery typically used a simple 'I' maze of limited areal extent. and l'ehrer used a much larger maze with considerably more locomotion required to traverse it. Hontgomery himself was apparently aware of this difference since he says in one of the many papers in which he concluded that satiated animals explore more than deprived animals. that in a. larger maze the deprived animals would explore more. though he does not attempt to explain why this would be so. It is inter- esting to note that one experiment by Thompson in 1953 (50) in- dicated no differences in exploration rates attributable to the motivational state of the as. and in this experiment the maze used as exploration ground for the 8s was intermediate in size to Montgomery's small 'I' maze. and Tehrer's very large maze. Hontgomery in particular went to great lengths to try and separate 'erploratory drive' from other motivational states of the organism. In one study in 1953 (38) he concluded that ex- ploratory drive is separate from activity drive. In another stuw by Montgomery and Honkman in 1955 (#0) a distinction be— tween exploratory drive and fear motivated responding is the 2. rshrer. :7 fans lffects of Hunger and l'amiliarity of Locale on prloration'. M 9; W a; Pksiolggical Pszcholg 1955s '01. 1*9s Po 551 Mk.——~ “, a... —’ -- .« e__a a“ ..~ 62 point at issue. lurther studies by Montgomery in 1952 (36). and in 1953 (37). examined.exp1oratory'drive in relation.to spontanp eous alternation. and to stimulus generalization. respectively. Derlyne (6). Glanzer (22). and Denny and Adelman (17). have taken theoretical account of stimulus satiation in somewhat similar fashions. The relevant aspects of Denny and Ldelman's theoretical position were quoted in the first chapter of this dissertation on account of their more direct relationship to the formulation of this problem. but a quote here of Derlyne's two postulate system might serve to illustrate its similarities to. and differences from. the theoretical position of Denny and Adelman. lotice that the sec- tion of Denny and idelman's theory 'Postulate 1. When a novel stimulus affects an organism's receptors. there will occur a drive-stimulus producing response which we shall call curiosity. l'l’ostulate 2. is a curiosity-arousing stimulus continues to affect an organism's receptors. curiosity will diminish" 3 which deals with the recovery of eliciting value by stimuli as a result of the passage of time during which the organism is not exposed to the particular stimulus. is only hinted at rather weakly in Derlyme's second postulate. and that solely by implication and not direct statement. Trom other articles generated by Derlyne's theory. it is apparent that some recovery of curiosity does occur with periods of non-exposure in spite of his theoretical silence 3. Derlyne. D: I. 'Hovelty and Curiosity as Determinants of In- Ploratory Behavior“. Dritigh M if Lem 1950. vol. 51 P. 70. 53 on that point. Derlyne in 1955 (7) states that complex environ- ments arouse curiosity more readily than do more simply structured environments. Also in this experiment. Derlyne's results indicate that the satiation of curiosity is related to previous emosure to the stimuli. the massing of the exposure trials. and exposure time within a trial. whether we conceptualize the relevant theoretical positions in terms of novel stimuli bringing about a drive-stimulus produc- ing response called curiosity. or in terms of the elicitation of approach responses being a prepotent property of novel stimuli which diminishes in potency with prolonged or repeated exposure to the stimuli. and recovers some (perhaps all) of its potency with the passage of time during periods of non-exposure. exposure to novel stimuli has a powerful effect on the behavior of organ- isms so exposed. Derlyne and Slater in 1957 (8) carried out an experiment which showed that rats have a definite preference for entry into the arm of a 'T' maze leading into a more complex maze which they are allowed to explore as opposed to entering a plain 'T' maze arm where they receive a small amount of food re- ward. Denny in 1957 (19) performed an experiment in which He were rewarded for visiting either arm of a 'T' maze with dis- similar arms (blacb-natural). Two trials per day spaced thirty minutes apart were given to all 8s. and the trials were so arranged that by forcing the 8 to respond in one direction. every 3 visited one side of the mass twice as often as the other (one 6h half of the 8s visited the black arm twice as often. the other half visited the natural arm twice as often). The result of this treatment was that the 5s came to take slightly more than 90% of their 'free' choices toward the less often visited side. This remember. in spite of being rewarded with food regardless of their choice of arm. A quote from the discussion of his experimental results by Denny is useful in illustrating how his theory deals with such events. "According to this (Denny's) position the responses prepotently elicited in a sit- nation are the ones that become conditioned to this situation. In the present instance an avoidance reaction is assumed to be con- sistently elicited by the more tequently em- periencfid stimuli and thus conditioned to them.“ The important result of the Denny experiment which Justified this treatment was that Denny's 8s persisted in choosing the less often visited and of the maze in a series of two free trials given after an intervening one week interval. Thus. to Denny. novel/ familiar stimuli elicit approach/avoidance responses which are as effective mediators of learned patterns of responding as are food for an hungry S or water for a thirsty 8. Butler in 1953 (10). and again in 195M (11) showed that rhesus monkeys will learn to correctly choose one of two alter- natives for the reward of being allowed to peek out of the en- closed box in which the discrimination is made through a small u. Dom I. 3. "Learning Through Stimulus Satiation'.1ourna,1 EWW- 1957. vol. 5n. p. 63 65 window which allows the S to look out of his small box into a larger room. Lastly. the typical picture of the laboratory rat busily exploring each new stimulus to which he is exposed.may at least partly derive from the fact that while not exactly domesticated. he has had countless generations to accustom.himse1f to man with his often strange ways. Barnett. whose personal bravery may well surpass that of the arerage experimental psychologist. in 1958 (’4) used trapped mature wild rate in an exploratory task and com- pared.their'performauce to that of typical laboratory animals. He found.that in contrast to the lab animals who.gig'husily explore the novel enrironment in which they were placed. the wild animals exhibited.'neophobia'. or extreme fear of their novel surrounds. This finding would certainly embarrass Berlyne. but not Denny.'who could.simply say that the response for the class of organisms including wild rats which is most prepotently elicited by exposure to novel stimuli of the sort encountered.in the lab- oratory is of the class of escape. withdrawal. or fear responses. 66 APPENDIX '3' 2111s appendix contains the raw data of. the experiment. including the weight 01’ Be at each exploration. the exploration scores. and the error scores in each learning situation. A particular 8 has the same nunber throughout the appendix. Subject We ight rirst Us ight Second Weight rhird lumber Exploration Exploration Exploration 1. 535 #60 3110 2. 5’6 1‘80 395 a. #30 his 380 . £633 “5 377 5. £110 395 6- 35 lH5 330 7. 75 has 360 s. 1135 380 326 9- #35 10 360 10. 1490 33;) 73 11. 515 13 12. I55 323 302 1a. 505 1405 1 . 1190 Mo 372 15. no hos 360 16. £150 1110 305 17. n55 130 too 18. 530 MED 330 19. M0 1120 331 20. 1190 360 370 21. 1195 1170 363 22. 1475 l+50 370 23- “20 390 335 . .+55 390 295 25. 25 l$75 is: 26. 90 M5 27. #70 1130 378 28. 190 mo 31.0 29. t$90 W5 367 332. £22; #33 375 . ' 393 32. 1+9!) inc 3033 3 . 1190 1:50 3 . #70 1110 - 350 35 . '#90 m5 370 36. 1185 l155 377 37. 11-00 80 3M5 38. M50 15 370 39. lL30 Inc 365 1‘0. 525 hso n13 1456 Smmswhmue 6 123 Scores by Minute 1 23 #5 67 6 mnnnnmumswmmamcwmanaunm Scores by Minute Coupoeite l‘irst lxploration Second Exploration third Exploration 1 23 M5 mmmm Sub e M. 3301301333231.03002135.49.305.4033010011102 360331325213120503123140453013°u20301u2233 5501120250551051230032hfi572131520122214‘438 sonuuaoulou ou010330u6h 53913311410332 33223 7002310563502063131.“.5655821335032203‘0‘4 9&63nh71n89655n9852595mn9h673885637h6687 #1013000222250302202232632130112h0311001 lOOthOOlZZl3030210h121h3222206221021202 h0035h05033¥00110523503631030276h2500h02 3001h1063l320063O3u3hh27332h215551212623 57036526112112h036hM22h8512623767h1125lh 1.15103016313325235533387n662556n666232‘“. 31030066nuu 55661h3253u65395366761u11.4.45 66203006h8h5h07hh55355557588h6685586h239 3500 50h. 92.“. 36 22 76 356.“.7558 035.“. 58 77.87531“. 528 66 o o 50 3751‘. 775.“. 7‘“. 5s“. 56 7‘“. 5.8 nu. 990‘“. 966666 5667 80h3933232O977838685h769M009667390960819 1111 111 l l 5 5 55 5 5 O 0 O O O O nalwmsnasummmaBsewwnmmaaaufiamnenauanvasw OnaOOOOOOOOL’M.0.“.Jfb..&°0’00O...’°OO&OOO&OOPH”/OOOOOO '63 'Y' W ERRORS AID mon-rm IRIALS Sub. manning rot. lrrore rot. Errors Tot. Error: by Eot. Error- Massed. Trial! lo. Order * Spaced trials 1. + 36 2. «7- 60 a. + ’46 . + 21 5. + 61 6. + 22 7. + 33 8. 4- 23 9. + 53 10. + 117 11. + ha 12. + 5h 13. I 19 1 . I a; 15. I 16. I 12 17. I 72 18. + 178 19. + Mo 20. + 37 21. I 57 22. I 31 23. I 3h 29. : g5 25. 2 26. I Eh 27. I 60 28. I 52 29. I 57 30. I 56 31. + 17 320 + an 33- + 3 3 . + 17 35. + . 7 32 37. I 13 38. I 37 go. I 39 . I 52 ‘ + '- Ben 'I' use flat 38 Run '2' use eecond 1h 9 11 2 16 Choice Point 1 2 3 1 5 9 6 2 29 h a 1n 8 19 20 9 16 7 12 13 5 8 h 3 3 1h 15 1h 20 1 9 6 u h 7 5 7 7 16 6 1o 7 h 3 6 1om12m11 12 9 7 1h 13 s 16 5 11 s 7 17 .6 11 22 2226 81019 10 22 7 9 21 13 21 5 5 17 9 16 6 e 7 17 1s 12 33 1s 11 7 8 16 12 19 9 5 27 8 1o 7 7 1o 6 1h 21 12 20 15 7 11 s 9 8 9 11 7 8 5 19 11 12 26 9 17 15 1h 23 16 16 21 1o 1- 15 15 16 1o 20 17 8 15 9 23 19 10 1s 8 16 15 13 11 6 19 13 h 5 2 3 5 6 5 6 15 17 11 6 5 23 9 s h 2 5 3 9 8 a 9 1o 5 16 5 12 1h 9 1h 5 s 16 9 12 6 6 11 1210 611716 5 22 s 13 20 l‘ree Trials 9 15 15 31 1h 27 19 27 1s 1s 20 15 u 11 13 20 2 20 s 26 6 21 21 15 11 1o 12 7 1n 13 32 1h 11 30 25 19 17 19 11 15 M ‘10 .l' 'I‘ ll. 1‘ “ V..."I".Ill‘llni Ill OI. 4". _ w III. I, '1' I III @ mm 01‘ 'l' m RESPONSES uuuruau 13an 13 Response Pattern + Icorroct. 0 carrot nu 123n56789012 I00 MthMMMQ Order +¢ ++ ++ 0+ ++ ++ ++ 0+ 0+ ++ ++++ ++++ f0++ 0+++ ++++ ++++ +00¢ ++0+ +0++ ++++ 0+++ +000 1. ++ ++ ++ ++ ++ +0 0+ ++ ++ ++ ++?+ ++++ ++++ ++++ ++++ ++¢+ +++¢ ++++ ++++ 0+++ +000 ++00 2. ++ ++ ++ ¢+ ++ 0+ ++ 0+ ++ ++ ++++ ++++ ++++ 0+++ ++++ ++0+ ++~0 0+0+ ++++ +0++ 0++0 +440 3. ++ 0? ++ ++ +f ++ 0+ 0% ++ ++ ++¢+ ++++ ++++ ++++ ¢+++ ++++ ++++ ++¢+ ++++ ++++ 00++ ++0+ ?+ &+ ++ 0+ 0+ ++ ++ ++ ++ 0+ ++++ 0+++ ++++ +++0 0+++ 0++0 00++ ++¢0 ++00 +0+0 +++O +000 5. ++ ++ ++ ++ 0+ ++ ++ ++ ++ +0 ++++ 0+++ ++++ ++++ ++++ +0++ ¢++0 +000 ++++ 0+0+ 0000 ++++ ++ ++ 0+ f+ ++ ++ ?+ 0? ++ en... ++++ +++f ++++ ++++ ++++ ++++ ++++ 0+¢+ V+¢++ +¢++ +0 .90 +00+ 7o +o+++++++h ++++++++++ ++++ ++r+ ++++ ++++ ++++ ++++ +¢++ ++++ +++¢ ++++ ++++ +0++ 8. ++ ++ ++ +9 ++ ++ 0+ ++ ++ ++ ++¢+ 0.??? +0++ ++++ .o.++o 00+... +00+ 00¢.O 0+0» 0+0... +00 .u. +000 9. r r . ~ _ . _ _ _ T _ . _ _ _ ~ _ ._ _ * _ _ w _ _ _ _ _ _ N _ .. k _ _ u _ . _ _ ~ . . _ _ + _ _ _ _ _ H __ _ . _ _ _ a...— .4L '1‘ 'l.‘ 70 e P ttcrn 5456789012on (N A) G S‘JOI’I n .LL 1 Day ‘1 m 1 {LOT Or in? AiOO Subject Runn' + + + + + O + + + + + + + u —— l + + + + + O + + 10. + + O + + + + + O O + + + + + + + + + + + + + O + + O + + + + + + + + + + O + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + 11. + + + + + -L a + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + O O O l + I 1- O + + + + + + + + + + + + O O + O O + + + O + + + + O O + + + + + + + + + + + + + + 13. O + O O + O O + + + + + + + + + + + + + + + O O O + O + + O + O + + + + + + + + O + + + + O + + O + + O + O O + O O O O + O O + + O O + O + + O O + + + J. o + + O O + O O + + + O + + + O + + + + + + + 0 14. I .1- + O O O O + O O O + O + O + + O + + + O + + + O + + + + + O + + O + + + + + + + + O + + + + + O O + O O O O O + O + + + O + + + + + 15. .L. c + O + O O O O + + + + .i- + O O O O + + + + + + + + O + O + + O + + + 16. O O O + + + + + + + + O + + + O + O + + + + + + O O O O + + + O O O O + + O + + + + + + + + O + O + O + + O + + O + + + + + + + + O + + + + + + + + O + + O + + + + + + O O + + O + + O O + + + O + + O 17. O + + + O O + O + + + + + O + + + + + + + + + + + O + + + + + + + + O + + + O + + + + 18. + + + + + + + + + + + + + + O + O + + + + + + + + + + + + + O + + + + + + 71 Massed Trials on Day 15 Response Pattern er Dayl 2 5 4 5 6 7 8 9 O l 2 Running Subject , n \J- Or lIo. ske on any trial mist H 19. + + + + + O + + + + + + + + + + + + + + + + + + + + + + + O + + + + + + + + + + 20. + + + + .1. I + + + + + + + + + + + + + + + + O + + O + + + O + + + + + + + + + + O + + + + + O + + O O + + O + + O + + + + O + O + O + O + O O + + O + + O + + + O O + + O O O + + + + + O O O + + + + + + + + + + + + + + + + + + + + O + + + + + v .4— D + + + + + + + + + O + O O O + + O + + O + + + O + + 0 I _- u— 0 I + + O + O + + + + + + + + O + + + + + + + + 0,. + + + + + + + + + + + + + + + O + + + O + O + + + + + + + + + O + + + + + + O + + + + + + + + + + + + + + + + + O + + + + + + + + + + O + O + + + + + O + + 24. + + + + + + + + + + + O + + + + O + + + + + + + + O O O O O + + + + + + + + + O O + + + + + + + + + O O + + + + O + + + + + + + + l 't' O + + O O O O + + + O + + + O + O + + + + + + + O + + + + + + + + + + + + + + + O + + + + O + + + + + + O + -l. .1. I 0 J. a + + + + O + + O + + + + O O O O + O + O O + + + + + + + + O + + + + + O O + O + + + + + + + + O + + + O + + + + + + + + + + + + + + O O 0/. + + + + + + + + 0 O + O O O + + + + O O + O + + + + + + + + + + + + -L. I O + O + + + + 72 Patte n y 1 2 5 4 5 e 7 8 9 O 1 2 Response unning Order D» ‘3 .LL Subject City 1 5 on D n K... o. 28. + + J. l + O + + O + + + + + + + + O O O + + O O O + + O O + O + + + + + + + + + + + + .1. t + + O + O + O + + + + + O O + + + + + + + + + + + + + + + O + + + + + O + + + + O + + + .1. I + O + + + + + 29. + + + + + O O + O + + + + + + + O + + + + O + + O + + + I 1— + + + + + + + + + O + + O + + + + + + + + + O + O + + + O O + + + w) + + + + + + + + + + + O O O O + + + + O + + + + + O + O + + O + + + + + O + + + + + + + + + No errors except on Trial 2, Day 1 51. + O O + + + + + + + + + + + + + + O + + + + O + + + O O O + O + + + 52. + + + + + + + + + + O + + + O + + + + + + + + + O + + O + + O + O + l 5, Day 1 ('1 11 Tri. No errors except on 7‘) 7) :1. No errors on any trie No errors on any trial. 55- + + O + + + + + + + + O + + + O + O + + + + + + + + + I 1- + O O O + + + + + + + + + + + + + O + O O + O + + + + + + + + + + O + + + + + + + + O + + + + + + O + + O O + O + O + + + O + + + + O + + + + 37- + + + + .1. a + + + + + O + + O + + + + O + O + -1. a O + O + + + + + O + + + + + + + + O O + + + + + + + + O + + + O O + + .L. I 2/ + + + + + + + + + + + + + + + O O O + + + + + + + + + + + + O + + + + + + + 73 ‘5 C" rn v t Order Day 1 2 5 4 5 6 7 8 9 o 1 2 esponse P' .5 I. N C) sunnin‘ Subject ay 15 on D + O + + + + O + + O + O O + + + + + O + O + + + O O + O O + + -l- + + + + + + + + O O + O O + + + l _- l + .1. I + + + O + + + + + + + + + O + O + + + + O + + + + + O + O + O + + + + + O + + O O O O O + + + + 1+o. + + + + + + + + + + + + + + + O + + + O + + + + + + + + maze first 1. 2. 3. 7. 8. 9. 10. 11. 12. 71. BIBLIOGRAPHY Alderstein. A. and Iehrer. B. "The Effect of l'ood Deprivation on prloratory Behavior in a Complex Maze". {oumal 2; 9213- mm .5251 2112191261231 W. 1955. vol. 1310 250-53 Allen. J’. I. “Exploratory Behavior in a Complex Kass Under Conditions of Satiation and Deprivation” Unpublished 3.5. m. Michigan State University. 1958 Allen. J. I. “A rechnique and an Apparatus for Assessing Stimulus Satiability in Rats“ Mblished Besgggh. Michigan State University. 1959 ' Barnett. S. A. Experiments on 'leophohia' in Wild and Lab- ' oratory Bats". British M g; zszcholog. 1958. vol. 1+9. pp. 195-201 Barnett. S. A. "Exploratory Behaviour“. British 1m; 2; W. 1958. vol. 149. pp- 289-310 Berlyne. D. B. “Novelty and curiosity as Deter-inants of Exploratory Barnum“- W 2.1. W. 1950. vol. 51. pp. 68-80 Berlyne. D. B. ”The Arousal and Satiation of Perceptual Cur- 10-1ty". L%o 2.1.9.223?“ tiv MWW. 1955. vol. . pp. 233.2 Berlyne. D. 3.. and Slater. J. l'Pereeptual Curiosity. Explora- tory Behavior. and Mass Learning“. 1m .93, W QWW’ 010 e 1957. vol. 50. We 223-332 Brody. I. 6. “Genetic Bases of Spontaneous Activity in the Albino Rat“. WW cho 1 5222529228.. 19"?» vol. 17. #5. 2 pp. Butler. 3. A. “Discrimination Learning by Rhesus Monkeys to Visual Exploration Hotivation'. lamina 3; W and WM. 1953. vol. . pp- 95-98 Butler. 3. A. 'Incentive Conditions Which Influence Visual hplorauon'. 12m 21 W W. 1951+. vol. ’48. pp. 19-23 Cummins. V. D.. Melenar. 0.. and Stone. C. "Intercorrelations of Measures of Ability in the Bat“. m 3; mm mm. 1932. vol. 11L. pp. 225-235 13. in. 15. 16. 17. 18. 19. 22. 23. 2h. 75 Densiger. I. and lainland. It. “The Bahituetion of Exploratory Behavior“. W We: 291221.951. 1951*. vol. 6. pp- 39-51 Bashiell. J. l‘. 'A mantitative Denonstration of Animal Dri're'l gm; 3; Cmgrgtive Psycholggz. 1925. vol. 5. pp. 205-208 Dashiell. J. 1'. and Begroff. A. G. 'A l‘orward Going tendency in Haze Running". Jami 21 92mm film. 1931- 701. 12. pp. 77-914 Beater. U. B. and Peeler. E. 'Spontaneous Alternation Behavior“ WM 1958. vol. 55. pp. l+12—h28 Benny. II. B. and Adelnan. E. It. “Elicitation theory II: !he formal theory“. Unpublished Theoretical Paper. Hichigan State University. 1953 Denny. II. B. and Adelnan. B. It. IElicitation Theory 1: An Analysis of he Typical Learning Situations“. Psychologicg M21. 1955. vol. 52. pp. 290-295 Denny. M. B. “Learning through Stisulus Satiation'. M .9; W Psycholog. 1957. vol. 5h. pp. 62-6!» layrs. J. T. “Spontaneous Activity in the gags. B221 ‘33! £22232; gamm. 195’4. vol. 2. pp. 25-30 l'ehrer. l. "The Effects of Hunger and Ianiliarity of Locale on'lxploration'o at t mm W. 1956. vol. 9. pp- 9-552 Clanser. ll. "Stimulus Satiation: An Explanation of Spontaneous Alternation and Related Phenomena“. W Min. 1953. vol. 60. pp. 257-268 Glanser. ll. ”Curiosity. Exploratory Drive and Stimulus Sat- iation'. W Bevigg. 1958. vol. 55. pp. 302-315 Hall. C. S. "The Genetics of Behavior'. Mp; Ps olo (Stevens. 8. S. ed.). 1951. John Wiley and Sons Inc. pp- 3 329 Harris. 3. B. 'An Analysis of the lass Learning Scores of Bright and Bull Bats with Reference to Motivational Pastors”. WM 19%. "lo “- PP- 130-135 76 26. Heron. U. !. "l'he Inheritance of llaze Learning Ability in Bate“. Wigwam. 1935. vol. 19. pp- 77-89 27. Haron. V. I. ”he Behavior of Active and Inactive Hate in Experimental Extinction and Discrimination Problems. m- £251.29. Record. 191m. vol. 1.. pp. 23-31 28. Heron. I. I. and Skinner. B. E. ”The Hate of Extinction in Maze Bright and Ease m1 Bats“. Psycholggiggl m. 19%. vol. II». pp. 11-18 29. Heron. I. I. and legend. 3. “Basal Metabolis- and Ease Learn- ing in into“. bagels: mon 1222:2125:- volo 1‘9 30. Hull. C. L. Essential; 3; Boggy. 1951. Yale University Press 31. Jensen. P. I. 'Ehe Effects of Blinds and Erequenoy in Latent Learning“. W 2gp. Diesertati . Michigan State University. 1957 32. Krech. n. and Crutchfield. R. 5. meme. L: 2119332192. 1960. A. A. Knopf Inc. 33- mppusvany. B. 'A Hots on the Learning of a 'Bachard' Hat“ mmnm 19%. vol. 17. pp. 91-93 3n. 1.... u. o. and mama. s. I. '!he Effect of museum on Spontaneous Activity in the Hat'. m M 5;; W 1928. vol. 8’4. pp. 321-329 35. Liddell. H. 3. Who Relationship of Spontaneous Activity to Classical Conditioning in the She '. M g; ngrgtivg Psychology. 1925. m1. 5. pp. #7 3 36. liontgonery. I. C. 'Exploratory Behavior and Its Relation to Spontaneous Alternation in a Series of Haze Exposures". @flmmwmu 1952. vol. . pp. 50-57 37. Hontgonery. I. C. I'Exploratory Behavior as a Motion of 'Sinilarity' of Stimulus Situation“. 333% p; C r t MW 2312112123. 1953. vol. . Pp- 129-133 38. Hontgonery. I. C. 'l'he Effect of Activity Deprivation upon Exploratory Behavior“. Journal 31 C tiv Lag. Miolggi- gm 1953. '01- ‘5. pp- E35531 39. Ill. ll2. l$3. 1+7. 118. 119. 50. 51. 77 Montgonery. I. C. “The Effect of Hunger and thirst Drives upon Exploratory Behavior“. 129%: mm: and £91- siologicel Psychology. 1953. vol. . pp. 315-319 Montgomery. I. C. and Monknan. J. A. “The Relation Between Pear and Exploratory Behavior“. 19% p; Cognagatéve 29.!» WW. 19550 '01- s PP- 132-13 Montgomery. I. C. and Zinbardo. P. G. “Effect of Sensory and Behavioral Deprivation upon Exploratory Behavior in the Rat“ 1222.291 2!. 5.222193% £32 £23.95 mi: 1957. vol. 7. 1313-223-229 lissen. H. w. “Study of Exploratory Behavior in the White Hat by Means of the Obstruction Method“. W 2; Gengtig P.y3h°10ag 1930s '01e 37. me 361’}? Reid. L. S. and finger. I'. U. “Ihe Rat's Adjustment to 23-hour load Deprivation cycles". 124% 2! 29222133119. 2.9.4 Ma- leainal 121222125:- 1955. vol. . pp. 110-113 Reid. L. S. and linger. 1'. V. “the Effect of Activity Restrict- ion upon AdJustnent to Cyclical Iood Deprivation“. 9.! .ng t '0 MW nmum. 1957. vol. 50. pp- 91 9 Hundquiet. E. E. and Heron. I. !. “Spontaneous Activity and a... burning“. Jamel 2!. 9.9mm W. 1935. '01s 19. PP. 297.311 Searle. L. T. “!he Organisation of Hereditary Maze-Brightness and Kano-Whoa". 9.25mi; W Me. 1919. '01. 39. pp. 279-325 Shirley. M. “the Relation of Activity to Maze Learning and to Bra-1n mew. loans; as W ram 1928. vol. 8. pp. 137-195 81ml. 8- WM. 1956. McCray-Hill Sprague. S. D. S. “Anticipation as a factor in Maze Errors“ mnww 1933. vol. 15. pp. 313-329 theatpson. V. R. “Exploratory Behavior as a Junction of Hunger in 'Bright' and. “11111“ Rate“. Comgtive as mm 29229122. 1953. vol. . pp. 323-3 i'ryon. R. C. “Studies in Individual Differences in Mass Ability III: Ihe Community of function Between Two Mass Abilities“ malaiWW. 1931. '01. 12. pp- 95-115 .. — .— k“-_——-b-—* -—._...- o ’ I -fl 9 _ . ._.. ___._.... __ -7— .li. ... -kfl _ . 52. 53. 51+- 55. 56. 57. 58. 59. 78 i'ryon. R. C. “Genetic Differences in Maze Loarning Ability in Rats“. e rbook g; Matignal Society £93 1133 51:11ng W 19 . vol. 39. pp. 111-119 rattle. W. I. and Dykshorn. 8. “She Effect of Castration and Ovariectcw on Spontaneous Activity and Ability to Learn“. Proceegpgs 2i jigs; Society 2; hormonal Biology egg. Me - cig. 1923. vol. 25. pp. 569-570 Walker. He II. and. 1.07. Jo W M. 1953. Henry Holt and Company Inc. Warden. C. J. and Hamilton. E. L. “She Effects of Variations in Length of Mass Pattern upon Rate of Fixation in the White Rat“. Jgurnal p; Goggtic Psycholog. 1929. vol. 36. pp. 229-239 Mashburn. M. 1‘. and Broer. M. “‘Dhe Relation of Hunger and Act- ivity Drives in Mass Dinning by Mice to the rector of Habit- nation 11". Lana a: W W. 1932. vol. 13. pp- LF0340? Welter. W. I. " 'l‘ree‘ Versus 'I'crced‘ Exploration of a Novel Situation by Bate“. W nopog'gg. 1957. vol. 3. pp. 95-108 Millions. C. D. and Huchta. J. C. “Exploratory Behavior in Ewe Mazes with Dissimilar Alternatives“. M 9; W 2.911 W 201929.123. 1957. vol. 50. pp- 509-513 Mingfield. R. C. and Dennis. M. “i'he Dependence of the Rat's Choice of Pathvays upon the Length of the Daily Trial Series“ mew W. 1931‘. vol. 18. pp. 135-1147 I ROOM USE ONLY i . 0 ~ \l «I11111111111