3 "Mi r-. 31., fit; *N‘V!”F;:*% «z 3 Ms 1 :. ’“MJN 11M "w a Q . I n..t&,-4l..__.;:u M; ., This is to certify that the thesis entitled THE INTERSTIMULUS INTERVAL IN CLASSICAL AUTONOMIC CONDITIONING OF YOUNG INFANTS presented by Eben 1V1. Ingram has been accepted towards fulfillment of the requirements for _Eh.Il.___. degree in W //m if)” N Q/ ”4.4 Date 8/28/73 0-7639 ABSTRACT THE INTERSTIMULUS INTERVAL IN CLASSICAL AUTONOMIC CONDITIONING IN YOUNG INFANTS by Eben Maceo Ingram Conditioned discrimination and extinction of the skin potential response was attempted in four-month-old infants. In addition to condition- ability. a major focus of the study was to investigate the effects of four different interstimulus intervals on conditionability. Also of major im- portance was an attempt to investigate the extent to which individual dif- ferences in orienting response magnitude predict conditionability, and interact with the interstimulus interval. A delayed conditioning procedure was used in which CS onset preceded UCS onset by 1500, 3500, 5500, and 7500 msec.; UCS duration was 1000 msec.; CS and UCS terminated simul- taneously. The 088 were 75 ch 800 and 400 Hz square wave tones. The US was a 3.6 g/mm air puff to the infant's cheek. A control group was included. The results support the contention that conditioning of autonomic responses is possible during early infancy. Conditioning was very much a function of 181, with 7500 msec. 181 group showing superior conditioning. Moreover, the results indicated a strong relationship between OR magnitude and conditionability. such that individuals manifesting large magnitude ORs also showed the greatest amount of conditioning regardless of ISI. THE INTERSTIMULUS INTERVAL IN CLASSICAL AUTONOMIC CONDITIONING IN YOUNG INF ANTS by Eben Maceo Ingram A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1973 ACKNOWLEDGEMENTS The author wishes to express his gratitude to Dr. Hiram Fitzgerald, Committee Chairman, whose guidance and suggestions were invaluable in the preperation of this manuscript. The author would also like to thank the other members of his committee, Dr. Gordon Wood, Dr. Mark Rilling, and Dr. Lester Hyman, for their assistance and response to such short notice. In addition Special thanks is offered Dr. M. R. Denny and Dr. S. Ratner for their assistance as members of the author's guidance committee. The author would also like to eXpress his gratitude to all of the members of the Michigan State University Psychology Department for their assistance over the years, which finally culminated in the completion of this didsertation . ii TABLE OF CONTENTS LIST OF TABLES ................. LIST 0 F FIGURES INTRODUCTION....... ..... TEMPORAL FACTORS IN INFANT CLASSICAL CONDITIONING THE OPTIMAL INTERSTIMULUS INTERVAL . . . . . RESEARCH WITH INFRAHUMAN ORGANISMS . . . . RESEARCH WITH HUMAN ADULTS: THE OPTIMAL ISI FOR SOMATIC AND AUTONOMIC CONDITIONING. THE OPTIMAL ISI AND THE DIFFERENTIAL CONDITIONING PARADIGM. . . . . . . . . . . . . . OPTIMAL ISI RESEARCH WITH INFANTS . . ..... METHOD..... ..... ...... SUBJECTS APPARATUS.... ..... DESIGN AND PROCEDURE ...... . . . ..... RESULTS......... HABITUATION............. ....... CONDITIONAL DISCRIMINATION . ...... Overall Effects Individual Differences iii 12 12 12 13 18 18 19 19 23 Page ANALYSIS OF OR MAGNITUDE AND CONDITIONABILITY . . . . . . . . . . . . . . . 24 EXTINCTION . . . . . . . . . . . . . . . . . . 27 SECOND INTERVAL RESPONSE . . . . . . . . . . . 30 DISCUSSION . . . . . . . . . . . . . . . . . . . 35 LISTOFREFERENCES................. 40 iv Table LIST OF TABLES EXPERIMENTAL DESIGN ...... . . . ..... ANALYSIS OF VARIANCE OF THE MEAN DIFFERENCE SCORE FOR THE FIRST INTERVAL SPR SUMMARY OF DUNCAN'S MULTIPLE-RANGE TEST ANALYSIS OF THE EFFECTS OF ISI ON CONDITIONABILITY . . . . . . . . . . . ...... MEAN RESPONSE MAGNITUDE FOR EACH _8_ TO THE CS+ AND CS— DURING DISCRIMINATION CONDITIONING. . . . ........ . ...... . . RESPONSE FREQUENCY TO CS+ AND CS- FOR HIGH AND LOW OR is AND FOR ALL§§ ...... THE MEAN MAGNITUDE OF THE MEAN DIFFERENCE ON THE LAST BLOCK OF CONDITIONING AND THE LAST BLOCK OF EXTINCTION TRIALS . . . . ..... MEAN RESPONSE MAGNITUDE TO TEST AND CONTROL STIMULI FOR THE SECOND INTERVAL RESPONSE DURING DISCRIMINATION CONDITIONING PEARSONr.............. ....... Page 15 21 22 25 31 33 LIST OF FIGURES Figure Page 1. THE MEAN RESPONSE MAGNITUDE FOR ALL § DURING HABITUATION, AND THE MEAN DIFFERENCE SCORE FOR ALL _§_s_ DURING ACQUISITION AND EXTINCTION . . . . . 20 2. THE MEAN RESPONSE MAGNITUDE FOR HIGH AND LOW OR SS DURING HABITUATION, THE MEAN DIFFERENCE SCORE FOR HIGH AND LOW OR g. DURING ACQUISITION AND EXTINCTION . . . 26 vi INTRODUCTION Fitzgerald and Porges (1971) refer to the 60's as "a decade of surging research interest in all aspects of infant behavior." Increased attention was then, and continues to be, devoted to factors involved in infant classical conditioning and learning. The focus of this research has been in two directions. The first of these was due to the consideration by several investigators of the conditions under which specific stimuli serve as effective elicitors of infant behavior (e.g., Brackbill 8: Fitzgerald, 1969; Kay, 1967). According to Soviet Physiologists, there is an immut- able developmental order for conditional stimulus (CS) effectiveness in conditioning. The Soviets contend that the effectiveness of the sense modalities from the earliest to the last in effectiveness are in order: vestibular, auditory, tactile, olfactory, gustatory, and visual (Brackbill & Koltsova, 1967). Brackbill and Koltsova suggest that temporal stimuli be positioned at the low end with thermal stimuli positioned at the upper end of the continuum. A second issue for infant conditioning research concerns the role of the CS. The Soviets contend that the CS is the most important component of those involved in classical conditioning for predicting the course of conditioning (see Brackbill & Fitzgerald, 1969). 2 Recent reviews of the literature on infant learning clearly show that human infants are conditionable (see Fitzgerald & Porges, 1971; Brackbill & Koltsova, 1967). However, the data on autonomically condi- tioned responses is meager in comparison with data on somatically medi- ated responses. Using the classical conditioning paradigm Kasatkin et aL , (1953), Koch (1965), Kaye (1965), Lipsitt and Kaye (1964) and Lintz, Fitzgerald and Brackbill (1967) have obtained evidence demonstrating conditioning of somatically mediated responses such as the Babldn Reflex, sucking, eyeblinking, and head rotation. The major extent of the conditioning of autonomically mediated responses in infants has been done by Brackbill, Fitzgerald and their associates (Brackbill & Fitzgerald, 1969; Brackbill, Fitzgerald, & Lintz, 1967; Brackbill, Lintz, & Fitzgerald, 1968; Fitzgerald, Lintz, Brackbill, & Adams, 1967; Fitzgerald & Brackbill, 1971; Forbes & Porges, 1972; Ingram 8: Fitzgerald, in press). These studies involve conditioning of pupillary reflex dilation and constriction, heart rate, and skin potential to auditory, tactile and temporal stimuli. Sometime ago, Jones (1930) reported conditioning of the GSR in nine-month old infants using auditory, tactile, and visual stimuli with electroshock as the US. Although the combined results of these studies are by no means conclusive, they cer- tainly suggest that autonomic conditioning in infants is possible. 3 Temporal Factors in Infant Classical Conditioning Attention toward the temporal factors involved in infant behavior has increased in the last few years, however, the focus of this research has generally been on the ontogeny of biological rhythmicity. This includes studies of endogenous biological rhythms such as body temperature. electrodermal phenomena, heart rate changes, and sleep wake cycles by Hellbrugge (1960). Other researchers have studied rhythms involved in infant sucking and crying behavior (Wolff, 1967); and REM -NREM cycles (Stern, Parmelee, Akiyama, Shultz, 8: Wenner, 1969). A second aspect of temporal regularity is the organism's ability to perceive the quantitative difference between intervals of time and to respond on the basis of this perception. For example, Brackbill, Fitzgerald 81 Lintz (1967) and Brackbill 81 Fitzgerald (1969) have found temporal conditioning of autonomic responses in infants, and have used this successful demonstration of simple temporal conditioning as a means of studying the development of more complicated temporal behavior, i.e. , the organism's ability to make complicated temporal pattern discrimin- ations (Brackbill 8: Fitzgerald, 1972). The Optimal Interstimulus Interval One of the many conditions affecting the formation of conditioned reflexes is the interstimulus interval (ISI). The ISI refers to the amount of time elapsing between the onset of the conditioned stimulus (CS) and the 4 onset of the unconditioned stimulus (US). The problem of the optimal ISI was first discussed by Hull (1943), who concluded that the optimal ISI is 500 milliseconds. Subsequently many researchers studied this phenomenon confirming Hull's conclusion (Spence, 1960). Nevertheless, a number of recent studies point to areas of difficuky in establishing the optimal ISI in classical conditioning (Stewart, Stern Winokur 8; Fredman, 1961; Lockhart, 1966; Lockhart 82 Grings, 1963; Hartman & Grant, 1963; Kykman, 1967). These problems center around the various criteria used for the identification of the conditional responses (CR), the various methods for establishing a conditional response, the differences between species, and the differences between response systems. The following briefly outlines some of the findings with regard to the optimal ISI in infrahuman and human adult S_s with respect to various response systems. In addition, two other areas are examined—-autonomic conditioning and conditional discriminations. Research with Infrahuman Organisms Attempts to determine the optimal ISI, in addition to having involved both animal and human is, have included infra-mammalian organisms as S_s_. Recent studies of infra—mammalian species suggest that the optimal ISI is greater than one second: [Coppock 82 Bitterman (1955) using marine analids; Adams, Hardesty, 8: Noble (1960) using the earth- worm; and Smith 8: Baker (1960) with the horseshoe crab]. 5 Research with organisms higher on the phyletic scale also demon- strates an Optimal ISI that is longer than one second (Noble 8: Adam, 1963; Noble, Gruender, 8: Meyer, 1959; Noble 8: Adams, 1963, 1963a and Noble 8: Harding, 1963). Caldwell 8: Werboff (1962) found that among young mammalian rats the Optimal ISI was longer than for adult rats, thus suggesting an ontogenetic relationship between the ISI and conditioning. ‘ 9 Research with Human Adults: The Optimal ISI for Somatic and Autonomic Conditionig The results of studies using human S_s are different from those employing animal _S_s_.. The interval producing the greatest amoumt of con- ditioning in humans is 500 milliseconds for at least two response systems: the eyeblink and the finger withdrawal responses (Beecroft, 1966). How- ever, for the conditioning of autonomic response systems the optimal ISI is longer than the optimal ISI for the classical conditioning of motor responses. Reports of successful conditioning of the eyeblink response have demonstrated that superior eyeblink conditioning occurs at 1818 of 500 milliseconds (Reynolds, 1945; Kimble, 1947). Additional studies have found 1813 of 500 msec. to produce superior conditioning of the eyeblink response (Kimble, Mann, Dufort, 1955; Ebel 8: Prokasy, 1963). In addition Hansche 8: Grant (1960) studied the effects of 150 msec. to 1750 msec. 1813 for both CS onset and CS offset and found that a 500 msec. ISI provided superior performance. 6 Similar results have been reported for a second response system-- the finger withdrawal response. Using a measure of percent responding several investigators found that SUperior finger withdrawal performance occurred with the use of the 500 msec. ISI (Wolf, 1930, 1932; Spooner 8: Kellogg, 1947; Fitzwater 8: Reisman, 1952; Fitzwater 8: Thrush, 1956). The Optimal ISI for classical autonomic conditidning is longer than for the conditioning of motor responses (Pavlov, 1927; Gerall 8: Woodward, 1958; Fitzgerald 8: Brackbill, 1968; Hilgard, Dutton, 8: Helmick, 1949; Kakigi, 1964). However, with respect to the conditioning of the GSR, research findings seem to fall into two categories, differing somewhat from findings for other autonomic systems. The first category involves studies reporting superior GSR conditioning with short ISIS ; for example, from 450 to 500 msec. (White 8: Schlodberg, 1952; Moeller, 1954; Prokasy, Fawcett, 8: Hall, 1962). The second category includes those studies reporting superior conditioning with long ISIs--1000 to 3000 msec. (Bierbaum, 1959; Wickens 8: Cochran, 1960; Jones, 1962). III attempting to explain these differences, Prescott (1965) suggests that the GSR has orienting and unconditional response tendencies to the onset of the CS, which have not been sufficiently controlled in most re- search. Stewart, Stern, Winokur, 8: Fredman (1961) also refer to the lack of apprOpriate controls. Stewart, et a1., advocate the use of a long ISI which they feel would allow the "true CR" to develop. However, during long ISIS multiple responses are likely to occur (Lockhart, 1966) thereby confusing the situation as to which is a true CR. An orienting 7 response or first interval response occurs shortly after CS onset (Leonard 8: Winokur, 1963; Lockhart 8: Grings, 1964). A response occurring just prior to the onset of the US or "pre-US" response is referred to by Stewart, et a1., as the "true CR" in GSR conditioning. In this study these responses will be referred to as second interval R's. Stewart also considers the response occurring in the position of the UR when the US is absent such as on test trials as a CR. However, Lockhart & Grings (1963) point out that the first interval or orienting response shows conditioning if a magnitude rather than a frequency criterion is used. Therefore conditioning can be demonstrated with either a frequency or a magnitude criterion when the long latency ISI is used. A hypothesis has been proposed to explain the fact that autonomic responses generally have longer optimal ISIS than somatic responses (Jones, 1961). Jones suggested that the optimal ISI for conditioning depends on certain characteristics of the reflex being conditioned. For example, some reflexes have long latencies while others have short latencies, therefore, long latency responses would condition at longer ISIS than would short latency responses. Eccles (1964) provides some neurophysiological evidence confirming that autonomic response systems generally have longer latencies than some somatic responses. 8 The Optimal ISI and the Differential Conditioning Paradigm Again regarding the eyeblink response, there is an additional factor exerting influence on the ISI; namely the differential conditioning paradigm. Recently investigators have begun to study the effect ISI on differential conditioning of the eyelid response. Hilgard, Campbell and Sears (1938) were the first to examine this relationship by conditioning the eyelid response using ISIS of 600 and 650 msec. The longer ISI was found to give better performance. This general finding is supported by Hartmm and Grant (1962) with a similar study which indicated that conditioned discrimination increased as the ISI increased in length. Differential conditioning of the GSR was found to increase with increasing ISIS (Kimmel 8: Pennypacker, 1963). However, this effect was attributed in part to reduced responding to the negative CS in the longer ISI conditions. The above authors accept an explanation similar to one offered by Hartman and Grant (1962) for similar differences found in differential eyelid conditioning. Hartman and Grant point to the possibility that inhibition of reSponding may require more time. Inhibition of the response is also seen to be the cause of the attenuation of the response magnitude peak such that the peak of GSR responding falls off after fewer numbers of reinforced trials. Therefore, the greater amount of time al- lowed for response inhibition to occur in the long ISI conditions would improve performance . Optimal ISI Research with Infants Research with animals and human adults on the question of an Optimal ISI seems to indicate that the longer ISIS afford better conditioning performance when the response is an autonomic response, and discrimina- tion conditioning rather than simple conditioning is attempted. However, most of the 61:1er research up to this point has yielded little evidence for assuming that any specific ISI would give superior performance with infants (Brackbill, Fitzgerald, 8: Lintz, 1967; Lintz, Fitzgerald, 8: Brackbill, 1967; Morgan 8: Morgan, 1944; Natio 8: Lipsitt, 1960; Rendle- short, 1961; Wenger, 1936). However, a closer examiination of some of the above studies seems to indicate that differential performance does occur to different ISIS. Natio 8: Lipsitt found that an ISI of 500 msec. produced a mean of 27 percent CRs in an eyeblink conditioning study. On the other hand Lintz, et a1. , in a similar study found that a 1000 msec. ISI produced responding at about the 90 percent level. Since these studies differed methodologically to some extent, the differences in the percent of responding may have been due to any of the methodological differences. However, recent research presenting a much clearer picture of the rela- tionship of the ISI to conditioning in young infants and children suggests that the longer ISIS afford better conditioning (Ohlrich 8: Ross, 1968; Little, 1971). Even among the animal research there is evidence to suggest that younger animals require longer ISIS for optimal per- formance (Caldwell 8: Werboff, 1962). 10 Eyelid conditioning studies by Little (1971) involving ISIS of 500, 1000, 1500, and 2000 msec. suggest that the only appreciable increase in responding over trials occurred in the group conditioned at ISIS of 1500 and 2000 msec. Subjects conditioned at the ISI of 2000 msec. did not show as large an increase in responding as did S_s_ conditioned at an ISI of 1500 msec. In a second experiment performance was compared between ISIS of 500 and 1500 msec. Conditioning was found to be significantly superior for those _S_§ conditioned with the 1500 msec. ISI. These data tend to support the contention that infants tend to condition best at ISIS longer than 500 msec. , at least with somatic responses. Finally, Fitzgerald 8: Brackbill (1973) have concluded from their research with stereotype conditioning that Simple associative connections are formed more rapidly and more uniformly than are complex associations such as those required in a discrimination task (also, Brackbill 8: Fitz- gerald, 1972). On the basis of the above, one can contend that the relationship between conditionability and the ISI is quite complicated and the formation of a developmental theory of conditionability specifying only a small portion of the components of the classical conditioning paradigm is questionable. To summarize: in the area of infant research the laborious task of examining in detail the temporal relationships involved in infant learning is required in order to understand the nature of conditional: reflex formation. For example, conditionability has been found to be related to the magnitude of the orienting reflex (Ingram & Fitzgerald, in press; Aslin 8: ll Fitzgerald, 1973) and to other factors such as the type of CS (Brackbill 8: Fitzgerald, 1969; Fitzgerald 8: Porges, 1971), and type of CR (somatic or autonomic) (Fitzgerald 8: Brackbill, 1971). The present study was designed to extend the investigation of variables affecting conditionability in young infants. Specifically, the present study investigates the effect Of ISI on conditionability of the skin potential response (SPR). METHOD Subjects The S_s_ were 30 full -term, clinically normal human infants from the local area surrounding Michigan State University. Subjects were 85-120 days old (mean = 112 days); 18 were male and 12 were female. Permis- sion for participation in the experiment was obtained in writing from the parents once a complete description and explanation of the experiment and procedure were given. Initially 33 S_s were contacted, 3 were dropped from the experiment because of excessive State changes (excessive crying). Appaeatus The apparatus consisted of a Grass model 7 polygraph for recording the SPR. The two Stimuli used as CSS were 800 and 400 Hz tones. The tones were delivered to S through a speaker located in front of and about 60 cm. away from §_ with an intensity of 7Sdb recorded at S_‘s ears. The tones were produced by an Eico tone generator Operating through a Sony amplifier. A 3.6 gm/mm puff of air served as the US. The US was delivered to _S_'§_ right cheek from a distance of 2.5 cm. The air pressure was regulated by a Hoke Regulator Valve, and delivery was controlled by a 12 13 Hunter Silent Solenoid. Stimulus duration was regulated by Hunter Electronic Timers. UCS duration was 1000 msec. for each group. The CS durations were 2500, 4500, 6500, and 8500 msec. for groups 1 through 4 respectively. The CSs and US terminated simultaneously. Inter-trial intervals varied ran- domly among 10, 15, and 20 seconds. The SPR was recorded as a change in millivolts with Beckman Biopotential electrodes placed as follows: the active electrode was placed on the arch of the foot halfway between the ankle and the first phalange; the referent electrode was placed over the tibia, 1/8 of the way up on the shin between ankle and knee; the ground electrode was placed on the outside of S'_s upper thigh. The skin surface was cleaned with 70 percent ethanol, dried, and covered with an electrolyte paste for contact between the electrode and the skin (Ingram 8: Fitzgerald, in press). Design and Procedure Subjects were assigned to each group on the basis of the magnitude of the orienting reflex (OR). The OR was measured according to the magnitude of the first response during habituation (Ingram 8: Fitzgerald, in press). Assignment to groups on the basis of OR magnitude was made such that each group was composed of 3 individuals whose OR magnitude was greater than 3.0 Mv. and 3 individuals whose OR magnitude was less than 3.0 Mv. The basis for the apriori cutoff was that this was the median OR magnitude found by Ingram and Fitzgerald in previous research 14 with infants of the same age level as those in the present study. The 4 different ISIS (1500, 3500, 5500, 7500 msec.) determined the 4 experimental groups. A fifth group served as a pseudoconditioning control. Each group contained 6 S_s, with half of the S_s in each group receiving the 800 Hz tone as CS+, while the other half received the 400 Hz tone as CS+. The experiment was divided into 5 sessions: one habituation, three conditioning sessions, and one extinction session. Table 1 shows the experimental design. To control for any confounding due to endogenous rhythmicity, sessions were given at the same time on consecutive days for each S for a total of 5 days. During session 1 each _S_ received 15 presentations of the 800 Hz tone, and 15 presentations of the 400 Hz tone in mixed order for a total of 30 presentations. The habituation ISI varied randomly among 15, and 10 seconds, with each stimulus duration being the same length as the CS duration for each respective experimental group, after the first habituation 3111, On the first trial all _S__s received the 800 Hz stimulus for a dura- tion of 5000 msec., then all subsequent trials were as described above. During sessions 2, 3, and 4 all experimental S_s received 15 CS+ and 15 CS- trials for a total of 30 trials. Three Of the 15 CS+ trials given per session were test trials (non—reinforced trials), for a ratio of 4:1 reinforced to non-reinforced trials. The test trials occurred on pre- sentations 4, 8, and 12 of the CS+ stimulus in each conditioning session. The CS- trial occurring prior to, but not necessarily adjacent to each 15 TABLE 1 EXPERIMENTAL DE SIGN Experimental Sessions 1 2 3 4 5 Habituation ' Conditioning Extinction 800Hz 8: CS+ -_— 800Hz 800Hz 8: N = 3 400Hz CS- ___. 400Hz 400Hz (mixed) All ISI 800Hz 8: CS+ = 400Hz 800Hz 8: Groups 400Hz 08- = 800Hz 400Hz (mixed) N = 3 30 trials total 30 trials daily 30 trials consisting of: 15 800Hz 15 CS+ 15 CS—. 15 400Hz (Of the 15 CS+ trials, 12 are reinforced and 3 (test trials) are not. Control Group 800Hz 8: Random presentations 400Hz of the 800 and 400Hz (mixed) tones N = 6 30 trials total 30 trials daily 30 trials 15 800Hz 15 800Hz 15 400Hz 15 400Hz 16 test trials served as the control trial. Thus, there were 3 CS- control trials per conditioning session. The only difference between CS- control trials and the additional 12 CS- trials occurring in each conditioning session was that the control trials served as comparison trials for the CS+ test trials . Extinction took place during session 5. All S_s_ (both experimental and control) were presented 15 trials of the CS+ unpaired with the US, and 15 trials of the CS-. The presentation schedule was the same for extinction as during conditioning. Therefore, extinction differed from conditioning in regard to stimulus presentations only with respect to the absence of the US. The pseudoconditioning control S3 differed from the experimental _S_§_ only during sessions 2, 3, and 4. The control is received the same Stimuli and schedule of presentation in session 1 (habituation) and S (extinction) as did the experimental §§° During sessions 2, 3, and 4, the control _S_s_ received a total of 30 presentations per session of the 800 Hz and 400 Hz Stimuli presented in random fashion. The basic difference between experimental and control S_s was that the US was omitted for the controls. The interval between stimuli varied randomly between 10, 15, and 20 seconds. Each _S_ appeared in the laboratory accompanied by at least one parent, to whom a complete description and explanation of the experiment was given, and from whom final permission for the infant's participation was obtained in writing. Care was taken to assure that permission was l7 granted only after the parents clearly understOod the exact nature of the experiment. The infant was then placed into an infant seat. The seat was located inside a sound attenuated booth (ambient noise level = SOdb). The electrodes were then placed as described above. RESULTS Habituation In order to demonstrate a conditional discrimination, it is necessary to demonstrate that the response magnitude to the CS+ either remains constant, or increases, while the response magnitude to the CS— shows a decrement, or no increase over trials. Analysis Of the mean response magnitudes to both stimuli during habituation indicated that there was no significant difference between the two stimuli in response magnitude (related measures t = .097, df = 29). Since the two stimuli elicited response magnitudes of basically the same level, any difference occurring during the subsequent course of conditioning should be due to the experi- mental manipulations. Each S'_s response during the habituation session was analyzed by dividing the habituation trials into three blocks of 10 trials each. Each block of trials consisted of 5 presentations of one stimulus and 5 presenta- tions of the other. The mean response magnitude for the first block of trials was compared with the mean response magnitude of the last block of trials. A related measures L—test (across all S_s on the first block, and across all _S_s on the last block of trials) indicated a significant difference between means (t_== 3.210, df -= 29, p<:.01). Therefore, across all E 18 19 the mean response magnitude to the first block of trials was significantly greater than the last block of trials, indicating that response decrement had occurred (see Figure 1). Conditional Discrimination Overall effects. Analysis of variance showed that the main effects of ISI, OR magnitude and Trials were Significant, as was the Trials X OR magnitude interaction. The analysis is summarized in Table 2. Since there were 5 different ISI groups it is not obvious from the analysis of variance which groups were different. The differences among the ISI groups were analyzed by Duncan's Multiple—Range Test. The test indicates that the 1500 msec. ISI was least effective in producing conditioning, followed by the control group, the 3500 msec. ISI group, and the 5500 msec. and 7500 msec. ISI groups as the most effective groups. The analysis is shown in Table 3. In all of the above analyses the mean difference score (the mean of the difference between the response magni- tude to CS+ and the response magnitude to CS-) was used as a measure of responding. Therefore, according to the Duncan Multiple Range statistic, differential responding to the CS+ increased as the ISI increased in length. In addition, the significance of the overall analysis on OR magnitude, Trials and the Trials X OR magnitude interaction indicates that differential responding to CS+ was also influenced by CR magnitude. The significant Trials effect indicates that S_s_ increased responding as a 20 1500 msec. 3500 msec. 5500 msec. Mean Mean 7500 msec. Mean Response Difference Control Difference 2. 2:- Magnitude (mv/ mm) (mv/ mm) (mv/mm) 2.0- 1.8%- 1.6- ml 1. 2r- 1.0~ J l l 1 2 3 HABITUATION ACQUISITION EXTINCTION TRIAL BLOCKS 123 Figure 1. The Mean Response Magnitude for all _S§ during habitua- tion and the Mean Difference Score for all ISI Groups during acquisition and extinction. ANALYSIS OF VARIANCE OF THE MEAN DIFFERENCE SCORE FOR THE FIRST INTERVAL SPR 21 TABLE 2 Source Ss df MS F Total 289 . 18 269 -- -— Between is 141. 65 29 -- -- ISI 65.28 4 16.32 5.49* OR 31. 10 1 31. 10 10. 47* ISI x OR 14. 15 4 1 -— Error (Between) 59.43 20 2.97 -- Within _S_s 147. 54 240 -- -- Trials 60.44 8 7.56 9. 34** Trials x ISI 33. 713 32 1 -— Trials x OR 103. 55 8 12. 94 15. 99** Trials x ISI x OR 112. 151 32 1 -- Error (Within) 129. 41 160 -- -- *p<.005 **p<.001 22 TABLE 3 SUMMARY OF DUNCAN'S MULTIPLE-RANGE TEST ANALYSIS OF THE EFFECTS OF ISI ON CONDITIONABILITY 35 1500 E Control 35 3500 R 5500 3: 7500 271500 = 1.25 Mv. -- .02 1.22 4.63* 5.75* ‘X‘Control = 1.27 Mv. -- 1.20 4. 61* 5.53* $3500 = 2.47 Mv. -- 2. 50* 3.62* 355500 = 5.88 Mv. -- 1.12 R7500 = 7.00 Mv. -- *p<.05 The overall mean difference score for each group. The figures on the right in the table refer to the difference between comparison means (Mv/mm). 23 function of practice. AS shown by the significant Trials X OR magnitude interaction the r_at_e_ of increased responding to the CS+ is related to OR magnitude. Figure 1 Shows the acquisition curves for all groups (1500, 3500, 5500, 7500 msec. and control groups). Generally all groups Show an initial rise in the difference score between trial blocks 2 and 5, however, after this initial increase all groups Show a decrease in the difference score, followed by a gradual rise in most cases. Specifically each group shows an initial sudden increase, with the 7500 msec. ISI group Showing the increase first, followed by the 3500 and 5500 msec. ISI group, with the 1500 msec. ISI demonstrating the phenomenon last. Following the sud- den rise all groups show a sudden decline with the exception of the control group which shows a gradual increase followed by a decline. The 5500 and 7500 msec. groups show a sharp rise in difference score. For the 7500 msec. group this increase is greater than the initial increase. The 5500 msec. group shows an additional increase following the second rise, which does not later decrease. In the case of the 7500 msec. group, the final increase drops Off to stabilize at a lower point. Individual differences. For each S, all trials from sessions 2, 3, and 4 were combined and divided into blocks of 10 trials. Each block of 10 trials contained 5 CS+ (including test trials) and 5 CS- trials. The mean response magnitudes of the CS+ trials were compared with their counterpart on 08- trials, and a difference score computed (a minus Sign 24 was assigned to all CS- responses, and the algebraic difference was used as the difference score for each trial). As can be seen in Table 4, seven of the 30 infants demonstrated a Significant level of differential responding to the CS+. Of these seven infants, 4 were in the 7500 msec. group, and the remaining 3 S3 were in the 5500 msec. group. This dis- tribution of § showing superior performance at longer ISIS supports the group finding of better conditioning at longer ISIS. Analysis of OR Magnitude and Conditionability Figure 2 Shows the performance of the high and low OR S_s collapsed across ISI groups. Both curves Show an initial gradual increase reaching a peak in responding, which is then followed by a decrease. For the high OR group, the decrease following the peak in responding is gradual ending on the 7th trial block, with an increase extending through the 9th trial block. On the other hand, the decrease following the peak of responding for the low OR group continues to zero at the 9th trial block with only a small increase occurring between the 7th and 8th trial blocks. The performance of the High OR S_s_ can be seen to be superior to the performance of the low OR _S_s_.. Also as seen in Table 2 the OR effect was significant (F (1,20) = 10.471, p <2.005). Figure 2 also shows that there were no significant differences between high and low OR _S_s on gate of habituation and Lag of extinction. However, there was a significant difference between the high and low OR _S_§_ on the mean difference magnitude during the extinction session (3 = 25 TAB LE 4 MEANS RESPONSE MAGNITUDE FOR EACH S TO THE CS+ AND 08- DURING DISCRIMINATION CONDITIONING Subject: mean response magnitudes (mv/mm) S ISI GROUP ' CS+ CS- (if t* 1 1500 msec. .155 . 177 8 . 181 2 . 268 . 233 8 . 238 3 .204 .231 8 .221 4 . 304 . 113 8 1 . 156 5 .088 .031 8 1. 240 6 .922 . 588 8 .998 7 3500 mSBC. 1.502 1.073 8 .813 8 . 340 . 380 8 . 175 9 1.270 .720 8 1.863 10 1.000 .444 8 1.932 11 .913 .457 8 1.682 12 .702 .488 8 1.034 13 5500 msec. .973 .429 8 1.854 14 .830 .344 8 1. 227 15 1.306 .511 8 1.475 16 2.325 1.003 8 2.357* 17 1.733 .833 8 2.868* 18 7500 msec. 1.942 . 831 8 3. 215** 19 .950 1. 210 8 . 670 20 1.758 . 787 8 2.932** 21 2.168 1.131 8 2. 336* 22 .996 . 386 8 2.803* 23 1.133 .433 8 3. 670*** 24 Control . 263 . 277 8 . 235 25 1 . 710 1. 653 8 . 342 26 . 478 . 499 8 . 229 27 .086 .053 8 . 618 28 .880 .753 8 . 392 29 .923 1. 235 8 . 571 +related measures t; *p<:.05; **p<102; ***p<.01. 26 Mean Mean Mean 3 . 0 - Response Difference Difference Magnitude (mv/ mm) (mv/mm) 2. 8 ~ (mv/ mm) High OR . Low OR A ‘N 0hIII lllll' III I I I 1 2 3 1 2 3 4 5 6 7 8 9 1 2 3 HABITUATION ACQUISITION EXTINCTION TRIAL BLOCKS Figure 2. The mean response magnitude for high and low OR _S_§_ during habituation and the mean difference score for high and low OR S_s during acquisition and extinction. 27 2.237, df = 14, p <1.05). Table 5 summarizes the analysis of response frequency to 081- and CS— by high and low OR SS, and for each ISI group. Extinction Figure 1 shows the extinction performance of all S_s. As seen in Figure 1 there is a significant decrement in responding from the last block of extinction trials collapsed across all .SE: i.e., means for §§ on the last block of conditioning were compared with the means for all _S_s on the last block of extinction trials (3 = 2.791, df = 29, p <:.02). However, when the above comparison was made within each group no Significant differences were found, i.e., the amount of decrement from the last conditioning trials block and the last extinction trial block was analyzed separately for each group (see Table 6). Figure 2 shows the extinction performance of both high and low OR §§_. A comparison of the means for the last block Of conditioning trials and the last block of extinction trials showed that a significant difference in magnitude of responding existed, which indicates the decrement occurr- ing over extinction trials was significant (t_= 2.791, df = 29, p<2.02) (see above) (the same comparison). Also, a comparison was made between high and low OR S_s on overall magnitude of the difference score across extinction trials. There is a significant difference between these two groups (t = 2. 237, df = 14, p <.05) which is probably accounted for by the large magnitude difference on the first block of trials. 28 TABLE 5 RESPONSE FREQUENCY TO CS+ AND CS- FOR HIGH AND LOW OR _S_s_ AND FOR ALL §§ N 08+ 08- df _t__ High OR 15 .52 .24 14 2.360* Low OR 15 .32 .52 14 .852 All _S_s 30 .84 .76 29 1.537 *p<.05. TABLE 6 THE MEAN MAGNITUDE OF THE MEAN DIFFERENCE ON THE LAST BLOCK OF CONDITIONING TRIALS AND THE LAST BLOCK OF EXTINCTION TRIALS (THE 9th AND 3rd BLOCK RESPECTIVELY) 9th Block of 3rd Block of Conditioning Trials Conditioning Trials GROUP (Mv/ mm) (Mv/ mm) df t 1500 . 34 . 10 5 1. 201* 3500 . 24 . 02 5 . 668* 5500 1.60 . 29 5 1.873* 7500 .87 -.01 5 1.805*+ Control .33 . 42 . 31 5 . 043* *Not significant +The lack Of significance in a comparison that appears as though it should Show significance may be explained by the extremely small number of degrees of freedom available. 30 Second Interval Response The second interval reSponse terminology refers to the response occurring in an interval beginning 1000 msec. after UCS onset and ending 5000 msec. later only on those trials when the UCS is omitted (test trial). The response was compared with a control to produce an algebraic difference score. The control was the first CS- trial occurring prior to the test trial. Table 7 summarizes the performance of all is except the 1500 msec. ISI group when the test trial response is used. The reason for excluding the 1500 msec. group is that the ISI is too short for solely a response to the UCS to occur, thus the response to the CS would overlap the response to the UCS. Therefore, in terms of the second interval response, a comparison between long and short ISI may be meaningless due to the lack of a comparable response. As seen in Table 7 only 1 _S shows a significant level of differential responding; this S is in the 7500 msec. group. In addition, the data indi- cates that the mean difference score increases as the length of the ISI in- creases. In Table 8 the Pearson ; shows significant correlations between the second interval response mean difference score (algebraic difference be- tween test and control of response magnitude) and trial blocks required for conditioning to take place. Also a significant correlation occurred be- tween the second interval response mean difference score and the response frequency score (the percentage difference between the percent of respond- 31 TABLE 7 THE MEAN RESPONSE MAGNITUDE TO TEST AND CONTROL STIMULI FOR THE SECOND INTERVAL RESiPONSE DURING DISCRIMINATION CONDITIONING Subject: mean response magnitude (mv/mm) S ISI GROUP Test Control (If t_ 1 . 3766 . 440 2 . 278 2 . 386 0.00 2 1.000 3 1500 msec. .886 .440 2 .909 4 * * 5 . 200 .093 2 1.163 6 * * 7 2.886 .553 2 1.401 8 * * 9 3500 msec. * * 10 2.720 .110 2 1.447 11 1.163 .110 2 1.305 12 . 663 . 496 2 . 318 13 * * 14 * * 15 5500 msec. * * 16 .703 .303 2 1.397 17 2. 606 . 276 2 2. 206 18 1.850 .333 2 1.710 19 1.163 .386 2 1.950 20 1.930 .683 2 1.772 21 7500 msec. 22 1. 340 .166 2 1.944 23 .996 .386 2 1.571 24 .606 .076 2 3.826** 25 * * 2 26 * * 2 27 Control 1 . 740 1 . 523 2 . 305 28 * * 2 29 . 530 . 672 2 . 221 30 32 Table 7 (continued) *Subject did not reSpond to either the test or control trial at»: p . 05. l. The data from a response in the position of the second interval response was recorded for the 1500 msec. grOUp for possible comparison. 33 TABLE 8 INTERCORRELATIONS, PEARSON r 1 2 3 4 5 6 7 8 9 10 11 12 1 1.00 2 .00 1.00 3 -.08 .02 1.00 4 .09 —.34 -.01 1.00 5 -.08 -.21 -.01 .17 1.00 6 .15 1.44* .02 .22 .22 1.00 7 -.14 -.32 .00 .32 .39* .57* 1.00 8 —.17 .50* .02 -.31 -.15 -.79* -.42* 1.00 9 -.15 -.46* .31 .43* .04 .69* .40* -.67*1.00 10 .24 -.20 -.05 .00 .13 .19 .06 -.31 .05 1.00 11 .13 -.21 -.19 .24 -.01 .22 .16 -.12 .04 -.18 1.00 12 .00 -06 -.12 -.06 -.08 -.07 .02 .16 -.06 .06 -.03 1.00 *p .05 1. ISI 2. OR 3. Habituation (number of trials) 4. State 5. Extinction (number of trials) 6. Mean difference score 7. Mean difference score for test trial 8. Trial blocks to condition 9. Frequency difference score 10. Sex 11. Age 12. CS type (400 or 800Hz) 34 ing to CS+ and percent of responding to CS -). The latter correlation suggest a positive relationship between the amount of responding to CS+ and CS- to the magnitude of the second interval response mean difference score (the larger the magnitude of the mean difference score, the greater the response to test trials over the response to the CS- control trials). There was no significant relationship between the second interval response and the magnitude of the OR. DISCUSSION The results of the present study may be summarized as follows: First, young infants (3-4 months old) are capable Of learning a conditioned discrimination, as indicated by the autonomically mediated SPR. This finding strengthens previous assertions that autonomic conditioning is possible in infants (Fitzgerald 8: Brackbill, 1973). Second, the inter- stimulus interval has a definite influence on conditionability, such that conditionability increases as the length of ISI increases. Third, there are individual differences in OR magnitude. Fourth, these individual differences in OR magnitude are related to conditionability, where condi- tionability in this study refers to the rapidity with which conditioning occurs as well as to the attainment of set criteria (a Significant level of differential responding) concerning reSponse magnitude. The question arises as to whether or not any difference between the response magnitudes to CS+ and CS- can be considered sufficient to denote a conditional discrimination. The answer to this question is apparently yes. The general superiority of the experimental S_s over the control Se; indicates a trend toward conditioning for all except one group, the 1500 msec. ISI group. In addition, the overall significant decrement during extinction indicates that discrimination conditioning took place. The 35 36 significant decriment Shows that a Shift took place from greater magnitude responding to the CS+ over the CS- to equivalent responding to both CSs. However, the failure of some individual S_s to reach the criterion set for conditioning (a significant level of differential responding) indicates differences in the degree of conditioning or conditionability among S_s. The overall analysis of the data showing a Significant effect of ISI on conditioning, and the ranking of the ISIS from least significant to most significant tends to suggest a trend similar to ISI research with human adults; that is superior autonomic conditioning is attained by using longer ISIS. All experimental groups with the exception of the 1500 msec. ISI group Show an increase in the differential responding to CS+ over CS-. The obvious speculation from these data is that some minimal time is needed for young infants to process the Stimulus input before an effective discrimination can be made. A similar interpretation of differential eyelid conditioning is offered by Hartman and Grant (1962) who, upon finding that the Optimal ISI is longer for differential eyelid condition- ing than for simple conditioning, interpreted this to suggest that the longer ISI permits a more complete "perceptual" reSponse to the nature of the CS. Either way a longer Optimal ISI for differential conditioning implies that Operations involved in responding to stimuli differentially are occurring during the ISI, and have definite temporal requirements. With respect to the temporal requirements for perceptual assimila- tion, the superior performance by the high OR S_s_, suggest that individually 37 unique levels of responding are related to differences in the degree that an individual can utilize the amount of time available for performing the operations involved in differential responding. The high OR S, regardless of the ISI, may demonstrate a more stable system of responding, which is reflected by the probability of an individual's ability to produce and in- hibit responses at given instances of stimulation. In other words, the high OR individual receives stimulation and performs whatever percep- tual Operations are necessitated by the nature of the stimulus, producing or inhibiting a response as required, in a Shorter amount of time. How- ever, for all individuals, the probability of a response decreases as the amount of time required for that response to occur decreases. In terms of the present data this can be seen in the superior performance of the high OR .SE at ISIS Of less than the optimal, where the probability of a response is not yet high enough to reach significance, but still higher than the low OR S'_s response probability. In addition, the significant Trials X OR magnitude interaction indicates that the mean level of differential responding to CS+ increases over trials. These data also sug- gest that the likelihood of responding appropriately, also increases over trials. Specifically the response probability interpretation suggests that the difference between the high and low OR _S_s_ on the mean difference score reflects a separation of the two groups on the probability of the CS- elicit- ing a response of equal strength as the CS+. Recall from Table 5 that 38 the frequency of response to the CS- is higher for Low OR S_s than it is for High OR __S§ Zeiner and Schell (1971) offer a similar interpretation of thier results obtained with human adults. They found that low OR S_s had as many as 4-7 times as many non-responses to the CS+ as did high OR 23' There are at least three theories which try to account fa: the fact that some S_S_ condition better than others. First, Spence (19541) suggested differences in drive level, with manifest anxiety as an indirect measure of drive. Second, Eysenck (1957, 1972) argued that introverts (from the introvert -extrovert dimension of the MPI) have less cortical inhibition, therefore, condition more rapidly. Third, Maltzman and his associates (Maltzman 8: Raskin, I965; Maltzman 8: Mandell, 1968) relate condition- ability to differences in attention, with the magnitude of the OR as the index of attention. Moreover, Sokolov (1963) suggest that OR elicitation facilitates conditioning. The second interval response, the response occurring when the US is omitted, is probably the most difficult of all the responses to interpret. For example, Grings, Lockhart, and Dameron (1962) found that in mentally subnormal boys the earliest evidence of conditioning occurred with the second interval response. However, one problem occurs with the second interval response when one attempts to compare very short (under 3000 msec. ISIS) and long ISIS (over 4000 msec.) using an autonomic response such as the GSR, and 39 SPR. The problem is with latency and time course, which are in the area of from 1000 to 5000 msec. This long time course and latency would produce overlapping responses, thus, making a comparison with long ISI situations difficult to interpret meaningfully. The long ISI would allow the reSponse to return to base level, thus subsequent responses would more likely be completely new responses rather than a summation of two responses. In the long ISI situations the second interval response has been considered a disparity response, that is, a response to the perceived absence of the US. It has also been considered to be a temporal condi- tioned response to the length of the ISI. A third possibility is that the response may be a conditioned OR to the Offset of the CS when the CS terminates at the same time as the US. Regardless of the nature of this response it shows characteristics of conditioning, i.e., fig demonstrate differential levels of responding to differentially reinforced stimuli. Despite that many S_s did not respond on test trials, a sufficient number did to indicate that this response has some significance in conditioning. What this significance is can be determined only through further research ‘ | into classical conditioning and ISI effects on the formation of conditioned , I | reflexes . LIST OF REFERENCES LIST OF REFERENCES Adams, C.K., Hardesty, D.L. 8: Noble, M. On the Optimum CS-UCS Interval, Paper read at the Midwestern Psychological Association, St. Louis, 1960. Beecroft, R.S. Classical Conditioning, 1966, Golita, Calif.: Psychonomic Press. Brackbill, Y., Fitzgerald, H.E., 8: Lintz, L.M.. A developmental study of classical conditioning. Monograph of the Society for Research in Child Development, 1967, 38, While No. 8. Brackbill, Y. 8: Fitzgerald, H.E. Development of the sensory analyzers during infancy. In L.P. Lipsitt 8: H.W. Reese (Eds.) Advances in Child Development and Behavior, Vol. 4, New York: Academic Press, 1969. Brackbill, Y. 8: Fitzgerald, H.E. Stereotype temporal conditioning in infants, PsychOphysiology, 1972, 9, 569-77. Berstein, A.L. Temporal Factors in the Formation of Conditioned Eyelid Reactions in Human Subjects. Journal of Genetic Psychology, 1934, 10, 173-97. Bierbaum, W.B. The Temporal Gradient in GSR Conditioning, Journal of Genetic Psychology, 1958, 59, 97-103. Caldwell, D. F. 8: Werboff, J. Classical Conditioning in Newborn Rats. Science, 1962, 136, 1118-19. Coppock, W. 8: Bitterman, M.E. Learning in Two Marine Analids, American Psychologi_sts, 1955, 10, 501 (Abstract). Deane, G.E. Cardiac Conditioning in the Albino Rabbit Using 3 CS*UCS Intervals. Psychonomic Science, 1965, 3 119-20. Dykeman, R.A. On the Nature of Classical Conditioning. In C.C. Brown (Ed.) Methods in Psychophysiology, Baltimore: Waverly Press, 1967. 41 42 Ebel, H.C. 8: Prokasy, W.F. Classical Conditioning as a Function of Sustained and Shifted Interstimulus Intervals. Journal of Experimental Psychology, 1963, 65, 52-58. Eccles, J .C. The Physiology of Synapses. New York: Academic Press, 1964. Fitzgerald, H.E., Lintz, L.M., Brackbill, Y., 8: Adams, G. Time Per- ception and Conditioning of an Autonomic Response in Young Infants. Perception and Motor Skills, 1967 , 24, 479-486. Fitzgerald, H.E. 8: Brackbill, Y. Interstimulus Interval in Classical Pupillary Conditioning, Psychological Reports, 1968, 23, 369-70. Fitzgerald, H.E., 8: Brackbill, Y. Tactile Conditioning of an Autonomic and Somatic Respose in Young Infants, Conditional Reflex, 1971, 6, 41-51. Fitzgerald, H.E. 8: Porges, S.W. A Decade of Infant Conditioning and Learning Research. Merrill-Palmer Quarterly, 1971, 17, 79-117. Fitzgerald, H.E., 8: Brackbill, Y. Stimulus-Response Organization, State, and Conditionability During Early Infancy. Paper presented at the meeting of the Society for Research in Child Development. Philadelphia, 1973. Fitzwater, M.E. 8: Reisman, M.N. Comparison of Forward, Backward, Simultaneous, and Pseudoconditioning, Journal of Experimental Psycholog, 1952, 44, 211-14. Fitzwater, M.E. 8: Thrush, R.S. Acquisition of a Conditioned Response as a Function of Forward Temporal Contiguity, Journal of Experi- mental Psychology, 1956, 51, 59-61. Gerall, A.A. 8: Woodward, J .K. Conditioning of the Human Pupillary Response as a Function of the CS-UCS Interval, Journal of Experi- mental Psychology, 1958, 55, 501-507. Grings, W.W., Lockhart, R.A., 8: Dameron, L.E. Conditioning Auto- nomic Responses of Mentally Subnormal Individuals, Psychological Monographg, 1962, 76 (Whole No. 558). Hall, J. F. The Psychology of Learning. Philadelphia: J .B. Lippincott Co., 1966. 43 Hansche, W.J. 8: Grant, D.A. Onset Versus TErmination of a Stimulus in Eyelid Conditioning, Journal of Experimental Psychology, 1960, 59, 19-26. Hartman, T.F. 8: Grant, D.A. Effects of Pattern of Reinforcement and Verbal Information on Acqiusition, Extinction, and Spontaneous Recovery of the Eyelid CR, Journal of Experimental Psychology, 1962, 63, 217-226. Hastings, S. 8: Obrist, P.A. Heart Rate Conditioning in Humans: Effects of Varying the Inter -Stimulus Interval. Journal Of Experimental Psycholog , 1967, 74, 431-442. Hilgard, E.R., Campbell, R.K., 8:Sears, W.N. Conditioned Discrimina- tion: The Effect of Knowledge of Stimulus Relationships, American Journal of Psychology, 1938, 39,, 498-696. Hilgard, E.R., Dutton, G.E., 8: Helmick, J.S. Attempted PUpillary Conditioning at Four Stimulus Intervals, Journal of Experimental Psychology, 1949, 39, 683 -689. Hull, C. Principles of Behavior: An Introduction to Behavior Theory. New York: Appleton-Century-Crofts, 1943. Ingram, E.M. 8: Fitzgerald, H.E. Individual Differences in Infant v Orienting and Conditioning. Paper read at the Society for Psycho- physiological Research, Boston, Mass. 1972. Jones, J.E. Contiguity and Reinforcement in Relation to CS -UCS Intervals in Classical Conditioning, Psychological Review, 1962 , 69, 176 -186. Jones, J.B. The CS -UCS Interval in Conditioning Short and Long Latency Responses. Journal of Experimental Psyphology, 1961, 62, 612- 617. Kakigi, S. Classical Conditioning of Human Pupillary Constriction, Epanese Psychological Research, 1964, 6, 155-162. Kimble, G.A. Conditioning as a Function of the Interval Between the Conditioned and the Unconditioned Stimuli, Journal of Experimental Psychology, 1947, 37, 1-15. Kimble, D.A., Hilgard, 8: Marquis. Conditioningand Learning. New York: Appleton-Century-Crofts, 1961. 44 Kimble, G.A., Mann, L.T., 8: Dufort, K.H. Classical and Instrumental Eyelid Conditioning, Journal of Experimental Psychology, 1955, 49, 407-14. Kimmel, H.P. 8: Pennypacker, H.S. Differential GSR Conditioning as a Function of the CS-UCS Interval, Journal of Experimental Psycholog, 1963, 65, 559-63. Leonard, C. 8: Winokur, G. Conditioning Versus Sensitization in the Galvanic Skin Response, Journal of Experimental Psychology, 1963, 66, 169-70. Little, A.H. Eyelid Conditioning in the Human Infant as a Function of the ISI, Paper read at the meeting of the Society for Research in Child Development, Minneapolis, 1971. Lintz, L.M., Fitzgerald, H.E. 8: Brackbill, Y. Conditioning the Eyeblink to Sound in Infants, Psychonomic Science, 1967, 7, 405-406. Lockhart, R.A. Comments on "An Analysis of GSR Conditioning," Psychological Review, 1963, 70, 562-64. Lockhart, R.A. 8: Grings. Comments Regarding Multiple Response Phenomena in Long Interstimulus Interval Conditioning, Psych - physiology, 1966, 3, 108-14. Lockhart, R.A. 8: Grings, W.W. Interstimulus Interval Effects in GSR Conditioning, Journal of Experimental Ppycholom 1964, 67, 209- 214. Maltzman, I. 8: Raskin, D.C. Effects of Individual Differences in the Orienting Reflex on Conditioning and Complex Processes. Journal of Experimental Research in Personality, 1965, 1, 1-16. Moeller, G. The CS-UCS Interval in GSR Conditioning, Journal of Ex— perimental Psychology, 1954, 48, 162-66. Morgan, J .J .B. & Morgan, S.S. Infant Learning as a Developmental Index, Journal of Genetic Psychology, 1944, 65, 281-89. Naito, T. 8: Lipsitt, L.P. Two attempts to Condition Eyelid Responses in Human Infants, Journal of Experimental Child Ifiycholog, 1968, 6, 181-93. NOble, M., Gruender, A. 8: Meyer, D.R. Conditioning in Fish as a Function of Interval Between the CS and UCS, Journal of Compara- tive and Physiological Psycholgy, 1959, 52, 236-39. 45 Noble, M., Adams, C.K. The Effect of Length of CS-US Interval as a Function of Body Temperature in a Cold Blooded Animal, Journal of Genetic Psychology, 1963, 69, 197—201. Noble, M. 8: Adams, C.K. Conditioning in Pigs as a Function of the Interval Between CS and US. Journal of Comparative and Physio— logical Psychology, 1963a, 56, 215-19. Noble, M. , Harding, G.E. Conditioning in Rhesus Monkeys as a Function of the Interval Between CS and US, Journal of Comparative and Physiological Psycholoy, 1963, 56, 220-24. Ohrlirch, E.A. 8: Ross, L.E. Acquisition and Differential Conditioning of the Eyelid Response in Normal and Retarded Children, Journal of Experimental Child Psychology, 1968, 6, 181-93. Pavlov, I.P. Conditioned Reflexes, London: Oxford University Press, 1927. Prescott, J .W. Neutral Timing Mechanisms, Conditioning and the CS- UCS Interval, Psychophysiology, 1965, 2, 125-31. Prokasy, W.F., Fawcett, J.T., 8: Hall, J.F. Recruitment, Latency Magnitude and Amplitude of the GSR as a Function of Inter- stimulus Interval, Journal of Experimental Psycholog, 1962, 64, 513-180 Rendle-Short, J. The Puff Test, Archives of Diseases of Childhood, 1961, 36, 50-57. Raskin, D.C., Kotes, H., 8: Bever, J. Autonomic Indicators of Orienting and Defensive Reflexes. Journal of Experimental Psychology, 1969, 80, 423-33. Reynolds, B. The Acquisition of a Trace Conditioned Response as a Function of the Magnitude of the Stimulus Trace. Journal of Eyerimental Psychology 1945, 35, 15-30. Spence, K.W. Behavior Theory and Learnipg, Englewood Cliffs: Prentice— Hall, 1960. Spooner, A. 8: Kellogg, W.N. The Backward Conditioning Curve. American Journal of Psychology, 1947, 60, 321-34. Stewart, M.A., Stern, J.A., Winokur, G. 8: Fredman, S. An Analysis of GSR Conditioning. Psychological Review, 1961, 68, 60—67. 46 Smith, J .C. 8: Baker, H.D. Conditioning in the Horseshoe Crab. Journal of Comparative and Physiological Psychology, 1960, 53, 379-94. Wenger, M.A. An Investigation of Conditioned Responses in Human Infants. In M.A. Wenger, J.M. Smith, C. Hazard and O.C. Irwin (Eds.), Studies in Infant Behavior III. University of Iowa Studies in Child Welfare, 1936, 12, No. 1, 7—90. White, C.T. 8: Schlosberg, H. Degree of Conditioning of the GSR as a Function of the Period of Delay. Journal of Experimental Psycholog, 1952, 43, 357-62. Wickens, D. D. 8: Cochran, S.W. Conditioned Stimulus Flash Rate and Efficiency Of Conditioning, Journal of Comparative and Physiological Psychology, 1960, 53, 341-45. Wolf, H. Time Factors in Conditioning Finger Withdrawal. Journal of Genetic Psychology, 1930, 4, 372-78. Wolf, H. Conditioning as a Function of the Interval Between the Condi- tioned and the Original Stimulus. Journal of Genetic Psycholog, 1932, 7, 80-103. Zeiner, A.R. 8: Schell, A.M. Individual Differences in Orienting, Conditionability, and Skin Resistance Responsivity, Psychophysiolog, 1971, 8, 612-22.