A SHORT .- TERM. Lonemmsmnv 50F - HABITUATION "1 TWO- AND FOUR-MONTH v OLD INFANTS ‘ _ ' O Dissertation for the Degree of-Ph. D. MICH'GAN STATE UNIVERSITY THOMAS ROBERT CHIBUCOS 1974 Links» 1 "I‘ll“... u. A, u! '1'..." 1‘ QY t!- L Michxg; State . 1-2; Univcnty EV ./’-.\ " ABSTRACT A SHORT-TERM LONGITUDINAL STUDY OF HABITUATION IN TWO- AND FOUR- MONTH-OLD INFANTS BY Thomas Robert Chibucos A longitudinal design was used to investigate habituation of attention in human infants. Sixteen boys and sixteen girls participated in auditory and visual habituation tasks at two and at four months of age. The visual stimuli were blinking lights that described simple geometric patterns, and the auditory stimuli were simple pure tones of 500 and 1100 Hz. Heart rate deceleration, changes in general body activity, visual fixation time, skin conductance, and skin potential were used as indices of attention. A habituation-dishabituation paradigm was used to determine whether the response decrement obtained was specific to the stimulus that was repeatedly presented or was due to other factors such as fatigue. One-half the infants were presented the same stimulus throughout the testing session, while the remaining infants received a "dishabituation" (novel) stimulus on the last two trials. Thomas Robert Chibucos The infants who were presented the dishabituation stimulus exhibited significant recovery of attention on all vari- ables except heart rate (HR) deceleration (and HR decelera- tion was in the predicted direction). There were no age by treatment interactions. Thus, habituation was demon- strated for two- and four-month-old infants. There were several other major areas of concern in the study. Specifically, there was interest in (l) the cross-age correlation of habituation and dishabituation; (2) the extent to which habituation and dishabituation are specific to particular stimulus-response connections as opposed to being holistic phenomena; and (3) the degree of cross-modal correlation for habituation and dishabituation within ages. Little support obtained for the hypothesized positive cross-age and cross-modal correlations. In addi- tion, habituation was seen to be a specific process. That is, the correlations among the several indices of attention were generally not significant for habituation, though there were several significant correlations for dishabitu- ation in the visual task. The study included the following important methodo- logical features. First, the state of the infants was scrupulously monitored throughout testing. Second, the age at each testing and the interval between testing ses- sions was tightly controlled. Third, the auditory and visual stimuli were randomly chosen. Therefore, the Thomas Robert Chibucos results of the study may be generalized to other stimuli to a relatively greater extent than is usually the case. Fourth, the stimuli presented for response decrement and for recovery trials were completely counterbalanced across infants. Thus, the treatment effect for dishabituation cannot be attributed to differential response-eliciting power of the stimuli used. Finally, several indices of attention (and thus of habituation and dishabituation of attention) were used. The use of several indices provided a more general assessment of the questions posed in the study than would have been possible with a single index. A SHORT-TERM LONGITUDINAL STUDY OF HABITUATION IN TWO- AND FOUR- MONTH-OLD INFANTS BY Thomas Robert Chibucos A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1974 C) Copyright by THOMAS ROBERT CHIBUCOS 1974 This dissertation is dedicated to my wife, Mary. ii ACKNOWLEDGMENTS Many people were instrumental in getting this piece of work finished. First, I want to thank the mem- bers of my dissertation committee, Hiram Fitzgerald, Ellen Strommen, Elaine Donalson, and Andrew Porter for their guidance and encouragement. Hi Fitzgerald, in particular, was a source of wisdom and guidance, not just for this dissertation, but throughout my graduate career at Michigan State University. If I can be the model for my students that Hi is for his, then I will do important work. There are others who deserve acknowledgment. Bob Bundy assisted me in setting up the psychophysiology equipment and was a valuable resource person. Matt Wasiac, Steve Mangan, and Mary Chibucos all helped run subjects. Finally, Oscar Tosi made recording equipment easily accessible to me and gave me assistance in recording the auditory stimuli. The parents and infants who took the trouble to come to our laboratory deserve special mention. It was a joy to work with all those beautiful infants. iii LIST OF LIST OF Chapter I. II. III. IV. TABLE OF CONTENTS TABLES O C O O O O O O O O O . FIGURES . . . . . . . . . . . . INTRODUCTION . . . . . . . . . . Definitions . . . . . . . . . Overview of the Study . . . . . . LITERATURE REVIEW . . . . . . . Visual Stimuli . . . . . . . . . Auditory Stimuli . . . . . . . . Other Studies . . . . . . . . Conclusions . . . . . . . . . . PURPOSES AND HYPOTHESES OF THE STUDY . . Purposes O O O O O O O O O O I Hypotheses . . . . . . . . . METHOD . . . . . . . . . . . . Subjects . . . . . . . . . . Research Design . . . . . . . Operational Definition of Habituation Experimental Design . . . . . . Choice of Stimuli . . . Testing Apparatus, Visual and Auditory Stimuli . . . Recording Equipment: Polygraph, Elec- trodes, Electrode Paste, Observers . Procedure . . . . . Response Measures, Reliability and Scoring . . . . . . . . . . Dependent Variables: Data Reduction . Statistical Analysis . . . . . . . Summary . . . . . . . . . . . iv Page vi viii 23 26 26 29 31 34 38 41 42 47 48 50 Chapter V. RESULTS . . . . . . . . . . . . Descriptive Statistics . . . . . . Hypothesis Testing . . . . . . . Overall Decrement of Attention . . . Age Effects . . . . . . . . . Treatment/Control: Recovery of Attention . . . . . . . .° . Sex Differences . . . . . . . . Decrement and Recovery: Cross—Age Correlations . . . . . . . . Decrement and Recovery: Cross-Modal Correlations . . . . . . . . Habituation: Global Versus Specific Nature . . . . . . . . . . Summary of Hypotheses and Analyses . . VI. DISCUSSION 0 O O O O O O O O I 0 Conclusions . . ._ . . . . . . . General Discussion . . . . . . . REFERENCES . . . . . . . . . . . . . APPENDICES O O O I O I O O I I O O O A. LETTER REQUESTING PARTICIPATION OF INFANTS IN THE STUDY AND POSTCARD TO BE RETURNED BY PARENTS O O O O O O I O O I O B. QUESTIONNAIRE FOR MOTHERS WHOSE INFANTS ARE PARTICIPATING IN THE ATTENTION STUDY OF 2- TO 4-MONTH-OLDS . . . . . C. PARENT CONSENT FORM . . . . . . . . Page 52 52 54 54 61 62 70 71 73 77 81 83 83 87 9O 98 99 102 105 10. ll. 12. 13. LIST OF TABLES Age and Time of Testing of Infants . . . . Example of Decrement and Recovery Scores Used in Data Analysis . . . . . . . . Means, Standard Deviations, and Standard Errors of the Mean Over All Subjects for the Sums of Trials 1+2, Trials 7+8, and Trials 9+lO . . . . . . . . Means,Standard Deviations, and Standard Errors of the Mean for the Two Difference Scores Used in Hypothesis Testing . . . . . . Analysis of Variance Summary Table for Skin Conductance Decrement Scores . . . . Analysis of Variance Summary Table for Skin Potential Decrement Scores . . . . . Analysis of Variance Summary Table for Heart Rate Change Decrement Scores . . . . Analysis of Variance Summary Table for ACthltY Change Decrement Scores . . . . Analysis of Variance Summary Table for Total Fixation Decrement Scores . . . . . Analysis of Variance Summary Table for Skin Conductance Recovery Scores . . . . Analysis of Variance Summary Table for Skin Potential Recovery Scores . . . . . Analysis of Variance Summary Table for Heart Rate Change Recovery Scores . . . . Analysis of Variance Summary Table for Activity Change Recovery Scores . . . . . vi Page 24 49 53 54 55 56 57 58 59 63 64 65 66 Table 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. Analysis of Variance Summary Table for Total Fixation Recovery Scores . . Means for Dishabituation for Treatment and Control Groups . . . . . . Between Age Correlations for Response Decrement for Each Index of Atten- tion (Auditory Stimulation) . . . Between Age Correlations for Response Recovery for Each Index of Atten- tion (Auditory Stimulation) . . . Between Age Correlations for Response Decrement for Each Index of Atten- tion (Visual Stimulation) . . . . Between Age Correlations for Response Recovery for Each Index of Atten- tion (Visual Stimulation) . . . . Cross-Modal Correlations for Response Decrement for Each Index of Atten- tion (Two-Month-Old Infants) . . . Cross-Modal Correlations for Response Recovery for Each Index of Atten- tion (Two-Month—Old Infants) . . . Cross-Modal Correlations for Response Decrement for Each Index of Atten- tion (Four-Month-Old Infants) . . Cross-Modal Correlations for Response Recovery for Each Index of Atten- tion (Four-Month-Old Infants) . . Correlations Among Indices of Attention for Two-Month-Old Infants . . . . Correlations Among Indices of Attention for Four-Month—Old Infants . . Summary of Hypotheses and Results . . vii Page 67 67 72 72 74 74 75 75 76 76 78 79 82 LIST OF FIGURES Figure Page 1. General Response Decrement/Response Recovery Paradigm . . . . . . . . . 27 2. Matrix Depicting the Design of the Study . . 30 3. Display of Visual Stimuli . . . . . . . 36 4. Treatment by Sex Interaction for Dishabitu- ation on the Activity Measure . . . . . 68 5. Age by Modality Interaction for Dishabitu- ation of Skin Conductance . . . . . . . 70 viii CHAPTER I INTRODUCTION It is certain that the human infant is not an insensate organism. It is also certain that William James was not accurate when he asserted that the infant's world is a "blooming grey mass of confusion." Several lines of research provide an empirical basis for these assertions, and for modern views of the capacities and capabilities of young humans. Research in several areas including attention (Fantz, 1956, 1958), conditioning and learning (Fitzgerald & Porges, 1971; Sameroff, 1971), and habitu- ation (Jeffrey & Cohen, 1971) indicates that considerable response modification occurs early in life, and that human infants from months of age and older are perceptually and cognitively capable organisms. The present work was an attempt to further explore the phenomena of habituation and attention in early infancy. Definitions Jeffrey and Cohen (1971) and Kessen, Haith, and Salapatek (1970) have observed that the term habituation continues to be used inconsistently. Thus, it must be made clear that habituation refers to response decrement to repeated stimulus presentation, such decrement not being due solely to peripheral processes such as sensory adapta- tion, effector fatigue, or changes in arousal (Kessen gt 31., 1970). Response decrement is a general term that is used without regard to the mechanisms of decreased responding. Habituation, on the other hand, refers to a specific class of mechanisms, i.e., central nervous system (CNS) processes, as explanations of decreased responding. Neuronal model (Sokolov, 1963), schema (McCall & Kagan, 1967), and cognitive process (Lewis, 1970) are all terms which implicate the CNS comparison mechanisms assumed to be involved in response decrement. Inextricably bound with the definition of habitu- ation is the thing "habituated." In the present study the -phenomenon of interest was attention. Attention, in a general sense, seems to be the process by which organisms direct their sensory and elaborating (cognitive) systems (Lewis, 1970). Researchers have attempted to operation- alize the construct by using physiological and behavioral measures as indices of attention. For example, decreases in activity (Collins, Kessen, & Haith, 1972; Kagan & Lewis, 1965), decreases in sucking rate (Kaye, 1966), heart rate deceleration (Graham & Clifton, 1966; Lewis & Spaulding, 1967), change in galvanic skin resistance (Crowell, Davis, Chun, & Spellacy, 1965), and cortical changes (Sharpless & Jasper, 1956) have all been used as Operational definitions of attention. It should be noted further that much of the Russian work on habituation of the orienting reaction (OR) can be regarded as studies of habituation of basic attentional processes (e.g., Sokolov, 1960, 1963). Overview of the Study The basic interest in the present research was on performing a well controlled study of habituation of atten- tion in human infants. Specifically, the study was designed to eliminate and/or control a variety of reasonable alternative explanations of response decrement in two- and four-month-old infants. This was done in order to inves- tigate whether habituation is indeed a major mechanism of response decrement in young infants. A within subjects longitudinal design was used to examine the stability of individual differences in habitu- ation of attention. In addition, the holistic versus specific nature of habituation was investigated by using several behavioral and physiological indices of attention for both the auditory and visual stimulus modalities. Finally, the relative durability of responsivity to visual and auditory stimulation was examined by between modality comparisons of habituation of attention. CHAPTER II LITERATURE REVIEW Visual Stimuli The available evidence strongly suggests that infants four months of age and older habituate to repeated visual stimulation (Jeffrey & Cohen, 1971). This conclu- sion is based mainly on the consistent finding of response decrement in virtually all studies rather than on the fact that peripheral processes have been eliminated as possible explanations of response decrement in many individual studies. For example, Lewis, Goldberg, and Campbell (1969) demonstrated decrement of visual fixation time to repeated presentations of a blinking light in infants aged 3, 6, 9, 12, and 18 months. Stimulus duration and inter-trial interval (ITI) were both 30 seconds. Nonetheless, the failure in their study to obtain recovery of the fixation response when a new stimulus was presented leaves the find- ing of response decrement open to several possible explana- tions in addition to the Lewis et_al. interpretation that it was habituation, e.g., fatigue and/or state changes may have caused the decrease in visual fixation. It should be noted that the original ITI of 30 seconds yielded little response decrement in the three-month-olds. Use of a zero ITI, however, produced some (though a nonsignificant degree) visual response decrement. Similarly, the findings of response decrement by Kagan and Lewis (1965) and McCall and Kagan (1970) are open to several alternative interpretations. In the Kagan and Lewis study, six-month-old babies were repeatedly presented with blinking lights-—the same stimuli that Lewis gt_al. (1969) had used. Visual fixation time decreaSed over trials, though little decrement in heart rate (HR) decel- eration occurred. Since no test for response recovery was attempted, the decrement in visual fixation is not clearly interpretable. McCall and Kagan (1970) did include a test for recovery in their study of four-month-olds; however, only those infants who exhibited response decrement were tested for recovery. Therefore, even though recovery of visual fixation obtained, there is a strong possibility that the recovery simply reflected a regression to the mean effect. Pancratz and Cohen (1970), however, obtained what appears to be a clear demonstration of habituation of visual fixation in four-month-old infants. Pancratz and [Cohen presented slides of simple geometric forms, each for ten fifteen-second trials with a .5 second ITI. In the test for recovery, the familiarized stimulus was alter- nately presented with a nonfamiliarized stimulus. Boys showed response decrement and recovery of visual fixation, while girls showed neither decrement nor recovery. Using the same stimuli, Cohen, Gelber and Lazaar (1971) replicated and extended the results of Pancratz and Cohen. Again boys, but not girls, habituated. Also, there was greater recovery of visual fixation time to stimuli that differed from the habituation stimuli in form and color as Opposed to form or color only. In other words, the boys' recovery was a direct function of the discrepancy between the standard (habituation) stimulus and the com- parison (recovery) stimulus. Demonstrations of habituation to visual stimuli in infants less than four months of age generally are less con- vincing than in older infants. For example, neonates (Haith, 1966) and two-txafour-month-olds (Haith, Kessen, & Collins, 1969) did not show a decrease in cessation of sucking with repeated presentations of patterns of blinking lights. Caron and Caron (1969) and Saayman, Ames, and Moffett (1964) did report response decrement of visual fixation as well as recovery to a novel stimulus. The infants in these studies were three and one-half and three months of age, respectively. Since presentations of the habituated and recovery stimuli were not counterbalanced in either study, however, response recovery may simply have resulted because the recovery stimuli were capable of producing higher levels of responding. Recently Friedman (1972) apparently has demon- strated habituation of visual fixation to checkerboard stimuli in neonates one to three days old. This finding needs replication, since it is the only demonstration to date of such an effect in neonates. Friedman's study appears to have been well controlled except for the strong possibility of statistical regression effects. Only 40 of the 90 subjects observed actually showed response decre- ment. Of these 40, only 29 percent (12) showed the recovery effect. The habituation criterion Friedman used was a fixation time of 8 seconds less than the mean of the fixa- tion time on the first three trials on which the stimulus was presented. A large difference between Friedman's study and other studies is the number and length of habitu- ation trials employed. Each trial was 60 seconds with an ITI of 5-10 seconds. The range of trials to the habitu- ation criterion was 8-2§_with a mean of 13.2 for the 40 babies who reached criterion. This degree of exposure to the stimuli far exceeds that reported for any other study of this nature in the literature. It was further reported that state was monitored and that all babies were alert when tested. Several things are apparent from this brief review of visual habituation studies. First, a wider range of dependent variables should be sampled. It is true that several investigators have used responses other than visual fixation and heart rate as indices of attentional processes in infants (e.g., Lewis, Wilson, & Baumel, 1971; Stechler, Bradford, & Levy, 1966), but these efforts have been few and far between. Second, research is needed on visual habituation in the neonatal to four-month age range. The number of studies using infants in this age range is small, though three months may be a transition period for many aspects of development (e.g., EEG changes, the elimination of various reflexes, etc.). [See Lewis et al. (1969) for a brief review of physiological and behavioral changes that occur in the eight- to twelve-week age range.] Auditory Stimuli A detailed review of studies that have dealt with habituation to auditory stimulation can be found in Jeffrey and Cohen (1971). Studies of auditory habituation have overwhelmingly employed neonates as subjects, and virtually all have used heart rate responsivity as the dependent variable. Further, there is ample evidence from these studies to suggest that the discrimination of sounds of various kinds does take place (Kessen et al., 1970). Unfortunately, most of the results concerning response decrement of neonatal heart rate responsivity can be ade- quately accounted for in terms of arousal or state changes (e.g., Bartoshuk, 1962; Hutt, Von Bermuth, Lenard, Hutt, & Prechtle, 1968; Graham, Clifton, & Hatton, 1968). Conse- quently, evidence of human neonatal habituation to auditory stimulation is not very convincing. On the other hand, studies by Moffett (1968), Clifton and Meyers (1969), and Horowitz (1972) indicate that four- to six-month-old infants do habituate to auditory stimulation. It is important to note that the number of response measures used in auditory habituation studies is even more restricted than is the case in visual habituation studies. Specifically, heart rate responsivity (usually accelera— tion because of the ages of the subjects) has been the_ dependent measure for all practical purposes. It must be emphasized that other variables be used, particularly since the heart rate component of the orienting reaction (i.e., heart rate deceleration) is difficult to elicit in young infants (Jackson, Kantowitz, & Graham, 1971). Other Studies A study by Moreau, Birch, and Turkewitz (1970) is of direct concern, since it was one of the studies that prompted the present research. These experimenters tried to measure the relative durability of responsiveness of newborns to repeated presentation of initially equivalent auditory and somesthetic stimuli, and to determine whether habituation is a general. phenomenon <1r is specific to particular stimulus-response connections. Newborns were presented with 40 one-second applications of either an 10 auditory (90 db white noise) or a somesthetic (stroke on the cheek with a brush) stimulus. The ITI was eight seconds and the dependent variables were cardiac acceleration and eye movement. Among other difficulties, the most serious problems with the Moreau et a1. study are: 1. There were no controls for sensory adaptation, effector fatigue, or state changes as possible explanations of the response decrement; 2. The eight-second ITI is too short when cardiac responses are of interest, since heart rate cannot return to baseline in so short a time (Kagan & Lewis, 1965); and 3. The auditory stimulus had a rise-decay time of less than .05 milliseconds (rise-decay time is the total time it takes a tone to reach maximum intensity from onset and then to decrease to offset) and it has been suggested that fast rise times cause startle responses and are not conducive to habituation (Graham & Jackson, 1970). Because of these and other methodological problems, the results of the Moreau et a1. study must be cautiously considered. Nonetheless, their work leads to the sugges- tions that: 1. Response decrement of HR acceleration was more readily obtained to somesthetic stimulation than to audi- tory stimulation; and 2. "Changes in the frequency of occurrence of the two types of response (eye movement and HR acceleration) 11 cannot be considered as resulting from general and ubiqui- tous changes in the infants' 'responsiveness' to stimula- tion," since the cardiac response to repeated auditory stimulation was more persistent than the eye movement response, and since the two responses were independent of each other in individual infants. Moreau et a1.concludedthat habituation, rather than being a unitary phenomenon which affects all response systems that are activated by a given stimulus, is specific to the particular stimulus-response conjunction examined. They may be right, but it is just as reasonable to conclude that somatic responses may habituate in one fashion, while autonomic responses habituate in a different way. To choose between these alternatives, it is necessary to study habituation to two or more somatic responses and to two or more autonomic responses. Thus, heart rate change, skin potential, skin conductance, visual fixation, and general body activity were used as response measures in the present study. The first three responses are autonomically medi- ated, while the last two are mediated somatically. Another pertinent study is the investigation by Brotsky and Kagan (1971). They investigated the stability of the OR longitudinally at four, six, and thirteen months. But instead of measuring habituation of the OR, Brotsky and Kagan measured degree of attention to stimulus presen- tations across a variety of visual and auditory stimuli, 12 attention being defined as cardiac deceleration. The two major findings were (1) a significant correlation for girls between the degree of the OR at eight and at thirteen months; and (2) a significant correlation for both boys and girls between the degree of the OR to auditory and to visual stimulation. Wetherford and Cohen (1971) performed the only longitudinal investigation of habituation with infants three months of age and younger. They tested babies at six, eight, ten, and twelve weeks of age and found habitu- ation of visual fixation only at the two oldest ages. Wetherford and Cohen were interested in age differences and testing effects and did not examine the stability of individual differences in habituation. Finally, Lewis and his co-workers have reported two studies in which infant attention across two modalities was investigated. In the first of these (Lewis & Spaulding, 1967), there was greater heart rate deceleration to audi- tory than to visual stimulation for six-month-old infants. In this study, a variety of auditory stimuli (e.g., 70 db tones, mother's voice) and visual stimuli (blinking lights describing various patterns) were used. Interest in the study by Lewis, Baumel, and Groch (1971) was on response decrement, and is, therefore, more germane bathe present study. These investigators foundru>significant decrementixx visual fixation time (i.e., length of first fixation), 13 heart rate deceleration, or activity change over six 30- second presentations of the visual stimuli. Similarly, there was no response decrement in HR deceleration or activity change for the auditory stimuli. The babies were three months old in this study, the visual stimuli were the same as in Lewis' other studies, and the auditory stimuli were C-tones and C—chords. Of particular relevance to the present study were the findings of no between modality differences in decrement of heart rate deceleration and activity. These results obtained despite between modal- ity differences in overall heart rate deceleration (i.e., greater HR deceleration for the auditory stimuli) and in overall activity (i.e., greater activity for the auditory stimuli). Similarly, though there were no sex differences in response decrement, the sex by modality interaction was significant for overall levels of responding. That is, girls showed greater HR deceleration to the auditory stimuli than to the visual stimuli, while boys showed comparable degrees of HR deceleration across modalities. Conclusions It is clear from this review that most studies of infant response decrement have examined only one dependent variable. This variable has usually been cardiac activity in auditory habituation studies, and some measure of visual fixation in visual habituation studies. It is also clear that too little is known about visual habituation in the l4 neonatal to four-month-old age range, while the same can be said for ages beyond the neonatal period in studies of response decrement to repeated auditory stimulation. The present study was designed to contribute to reducing these deficiencies by using several dependent variables to assess response decrement to auditory and to visual stimu- lation for infants at two and at four months of age. CHAPTER III PURPOSES AND HYPOTHESES OF THE STUDY Purposes The first and most basic purpose of the study was to perform a well controlled investigation of habituation of attention in human infants. Though many researchers in infancy have interpreted the results of their response decrement studies in terms of neuronal models (Lewis et_al., 1969), schema formation (McCall & Kagan, 1970), and other constructs which imply some form of "memory" and comparison process (i.e., habituation), the response decrement obtained in the overwhelming majority of cases can just as easily be ascribed to peripheral factors such as those previously mentioned (e.g., Bartoshuk, 1962; Kagan & Lewis, 1965; Lewis, Goldberg, & Rausch, 1967; Moreau et al., 1970). Even in those studies where the investigators were working within an explicit theoretical framework that defined habituation in a specific way, e.g., as acquisition of a neuronal model (Lewis et al., 1969), the empirical demonstration of response decrement could often be attribu- ted to any one of a number of potential mechanisms. It must be pointed out that no attempt was made in the design 15 16 of the present study to differentiate between theoretical positions that implicate cortical mechanisms of habituation (e.g., Sokolov, 1963) and positions that implicate other central mechanisms (e.g., Horn, 1967). Rather, what was emphasized is the difference between this type of explana- tion of response decrement and those explanations that rely on changes in peripheral receptor/effector systems. The distinction drawn between habituation and other decremental processes (such as sensory adaptation or fatigue) is that in habituation, the sensitivity of receptors and effectors is thought to be unaltered (Kling, 1971). Changes are assumed to be occurring in the central nervous-system. The second purpose of the study was to assess the stability of individual differences in habituation over a short age span in infancy. There have been several longi- tudinal investigations of habituation (e.g., Lewis e£_al., 1969; Wetherford & Cohen, 1971) but none has systematically addressed this issue (Ratner, 1970). A longitudinal exami— nation of the stability of individual differences in habituation is important also because of the suggestive relationships found between response decrement and other individual difference variables. For example, response decrement at 44 months was found to correlate significantly with performance on a two-choice discrimination task (r = .39) and with performance on a concept formation task 17 (r = .37; Lewis et al., 1969). Also response decrement at 12 months was related to Stanford Binet I.Q. scores at 44 months (r = .46 for girls and r = .50 for boys; Lewis, 1967). The third purpose of the study was to initiate investigation of the features of the environment to which human infants remain responsive. This line of research is important, though often unrecognized as being so, in that the identification of the infant's effective environment is a necesSary precursor to analyzing the infant's psycho- logical level of functioning (Hess, 1970; Schneirla, 1957). The general question concerns the relative durability of the effectiveness of stimulation in different sensory modalities. This translates in the present study to a comparison of habituation between the auditory and visual modalities. .The final purpose of the experiment was to inves- tigate whether habituation is a unitary phenomenon or is specific to particular stimulus-response conjunctions. This was the major reason for using multiple indices of attention in the present study. In addition, as Lewis (1970) observed, it is inappropriate to use only one response measure to indicate a behavior process, particu- larly when infants are the subjects of study, since the meaning of responses changes over time. For example, crying at nine months of age may be a very aggressive response 18 (e.g., when a toy is taken fromaanine-month-old there is not much else the infant can do but wail). On the other hand, if taking a toy away from a three-year—old elicits crying, the reaction may well be considered a nonaggressive or dependent one. Second, it was desirable to use a variety of responses, since habituation has not been clearly demonstrated in two-month—old infants (the young— est infants in the study). Finally, there have been very few attempts to use more than one measure of habituation in a particular study. Moreau et a1. (1970) and Lewis et al. (1971) used two or more response measures, but the need for multiple response measures in habituation research was not met because of problems in research design and differences in the purposes of these studies. Hypotheses The first hypothesis was based on the consistent findings from many studies that have used neonates, and infants four months of age and older as subjects: H1: There will be decreases in response magni- tude w1th repeated stimulus presentations for both two- and four-month-old infants. The hypothesis seems straightforward, but it was actually an "extrapolation" from two sides. It is quite possible that the decrement obtained with neonates is qualitatively different than that obtained with four- to six-month-olds. If this is the case, and remembering Lewis' failure to obtain decrement for a variety of responses in 19 three-month-olds, as well as the variety of cerebral reorganization that occurs around three months of age, it is possible that there would be an age difference in the decrease in attention. These considerations and the Lewis et a1. (1969) finding of age differences in response decre- ment led to the second hypothesis: H2: There will be greater decreases in response magnitude to repeated stimulus presentation for four-month-olds than for two-month-olds. Given response decrement (i.e., decreased attention) to repeated stimulation, it becomes meaningful to talk about response recovery. Therefore, hypothesis three was: H3: Following familiarization as a result of repeated stimulation, such familiarization being indicated by response decrement, there will be response recovery when a nonfamiliarized stimulus is presented. As indicated previously, and as will be made clearer when details of experimental design are considered later, this was the crucial hypothesis. If evidence obtained which supported the hypothesis, then the conclusion that habitu- ation is a viable and operative cause of response decrement in human infants would be empirically based. It was also hypothesized, on the basis of the sex differences in duration of first visual fixation presented by Cohen and his co-workers, that males would show greater habituation than females. The hypothesis was: H : There will be greater decreases in response magnitude and greater response recovery for males than for females. 20 The next hypothesis was truly of the hunch variety, since there is no previous work on which it could be based. It is true that Wilson and Lewis (1971) performed a longi— tudinal study of attention, but their study did not address the issue of habituation of attention (and they used infants six months of age and older). H There will be positive correlations between ages for both response decrement and response recovery. 5: The hypothesiskusbased on the work of Lewis who found significant correlations between performance in a response decrement task at 12 months of age and I.Q. performance at 44 months of age. This hypothesis was for each modality of stimulation separately. The sixth hypothesis was: H6: There will be positive correlations, Within age, between modalities for both decrement and recovery of responsivity. This prediction was derived mainly from the neural systems theory of habituation which implicates the reticular acti- vating system as a major mechanism in habituation (Groves & Lynch, 1972). A basic feature of the reticular activating system (RAS) is that it receives collateral input from all sensory modalities. Therefore, at least to some extent, responsivity of the RAS is independent of type of stimula- tion. The final purpose of the study was to examine, within each type of stimulation, habituation of several 21 indices of attention. This type of investigation is needed from a purely descriptive standpoint (Kagan & Lewis, 1965); however, information also obtained that is relevant to Sokolov's (1963) neuronal model of the orienting reflex (OR) and its habituation (Lynn, 1966). The OR, or "what- is-it" reaction, was described by Pavlov as follows: It is this reflex which brings about the immedi- ate response in man and animals to the slightest changes in the world around them, so that they immediately orientate their appropriate receptor organ in accordance with the perceptible quality in the agent bringing about the change, making a full investigation of it. The biological sig- nificance of this reflex is obvious. If the animal were not provided with such a reflex its life would hang at any moment by a thread. In man this reflex has been greatly deve10ped with far reaching results, being represented in its highest form by inquisitiveness--the parent of that scientific method through which we hope one day to come to a true orientation in knowledge of the world around us. [Pavlov, 1927.] Do several autonomic and somatic indices of atten- tion habituate as a group? Or as Moreau et a1. (1970) put it: Is habituation a characteristic of the input system per se or is it dependent on particular constella- tions of input—output relationships? Evidence supporting a single arousal mechanism underlying the various vegeta- tive and central nervous system (CNS) response systems would obtain if habituation in each modality occurred for all indices of attention. Such an outcome would support Sokolov's notion of a nonspecific orienting response. On the other hand, differential habituation for the several indices of attention would indicate different arousal 22 properties and habituation features for specific stimulus- response relationships. The latter result would be consistent with the conclusion of Moreau et a1. (1970) that different responses habituate with differential ease to the same type of stimulation. Thus, the final hypothe- sis, investigative in nature, was: H : There will be positive correlations among the indices of habituation (i.e., decrement and recovery) of attention within each input modality for each age. CHAPTER IV METHOD Subjects Four hundred and five letters requesting partici- pation in the study were sent to parents who had made announcements in a local newspaper of their child's birth. Seventy-nine (19.5 percent) returned the postcard that had been enclosed with the letter to indicate their permission to allow their baby to take part in the study. [Copies of the letter and postcard appear in Appendix A.] Complete sets of data were obtained for 32 (16 boys and 16 girls) of the 79 infants who initially came to the laboratory. Data from 47 of the initial group who participated in the experiment have not been included in the present study for the following reasons: (1) six babies failed to meet pre-established criteria for inclusion in the study; (2) seventeen of the subjects were affected by experimenter error and/or equipment failure, or were part of the pilot study; and (3) twenty-four babies experienced state prob- lems that were judged to interfere with the purposes of the study. The criteria for inclusion in the study were (1) assurance from the parents that the baby would be 23 24 brought back for retesting at four months of age; (2) birth weight 3 2,500 grams (Tanner, 1970); (3) normal spontaneous or low forceps delivery; (4) no genetic defects or congenital abnormalities; (5) good maternal obstetrical history; and (6) normal postnatal medical history. A questionnaire which addressed these criteria was filled out by the mother. [A copy of the questionnaire is presented in Appendix B.] Other characteristics about the testing of infants appear in Table 1. Note particularly the restricted range of time between testing, and the fact that all babies were tested between 1:00 P.M. and 6:00 P.M. The latter pro- cedure was used in an attempt to control for the potential impact of differences in diurnal rhythms on state and TABLE 1 AGE* AND TIME OF TESTING OF INFANTS Range E First testing 51-69 60.18 Second testing 116-128 123.75 Interval between Testing Sessions 53-70 61.25 Time of day tested 1:00-6:00 P.M. *Numbers in rows 1-3 represent ages in days. 25 responsivity of the infants (see Luce, 1971, for a read- able compendium of "biorhythm" research). All babies were tested within nine days of their two-month birthdays, and within eight days of their four—month birthdays. The influence of "state" on a variety of infant behaviors is well documented (e.g., Ashton, 1971). State refers to the infant's overall level of functioning at any given period of time on a continuum ranging from deep sleep to awake, alert, and active (Brackbill and Fitzgerald, 1969). State was carefully monitored in the present study by an experimenter who remained in the testing room with the infant. If the baby was judged to have state problems (sleep-drowsiness or fussy-crying) on trials one and/or two, the baby was removed from the apparatus and handled by the mother, then testing was begun anew (n = 6). If handling by the mother did not induce a testable state in tflmainfant,cm'if state problems occurred on or after the third trial, the baby was eliminated from the study (n = 24). Finally, if the baby normally used a pacifier he was allowed it at any time during the testing seSsion. In these cases (n = 12) the experimenter presented the pacifier to the baby during a trial interval or the mother gave the baby the pacifier prior to starting the experi- ment. The twelve babies who used pacifiers at one time or another were distributed by chance among the groups in the 26 study. Further, there were not systematic differences on any of the dependent variables between "pacifier" and "non- pacifier" babies. For example, the mean heart rate change scores (i.e., pre-stimulus minus post-stimulus heart rate in beats/minute) for trials one plus two were 5.06 and 5.42 for pacifier and non-pacifier babies. Research Design Operational Definition of Habituation The general response decrement/response recovery paradigm employed in the study (often referred to in the literature as a habituation/dishabituation paradigm) had two key features. First, all infants were repeatedly pre- sented a stimulus (the decrement stimulus) for 8 trials. This was the habituation or response decrement phase of the procedure. Second, on trials 9 and 10 half of the infants were presented a different stimulus (the recovery stimulus) while the remaining infants continued to experience the decrement stimulus. This was the dishabitu- ation or response recovery phase. A general representation of the basic paradigm is depicted in Figure 1. It should be noted that the paradigm was identifical for both habituation tasks at both ages. The comparison between the treatment and control groups on trials 9 and 10 was the crucial comparison in 27 Groups Trials 1 2 3 4 5 6 7 8 9 10 T A A A A A A A A B B C A A A A A A A A A. A T: Treatment group presented recovery stimulus on trials 9-10 C: Control group presented decrement stimulus on all 10 trials A: Any stimulus in a given modality B: Any stimulus, other than A, in the same modality as A Figure 1.--Genera1 Response Decrement/Response Recovery Paradigm. the present study. Assuming that response decrement in "attention" from trial 1 to trial 8 obtained (an assumption which was tested), the decrement was defined as habituation only if there was a difference between treatment and con- trol groups on the recovery (n: dishabituation trials. If there was no treatment effect for dishabituation, then the initial decrement on trials 1 to 8 was not stimulus- specific. On the other hand, if recovery of attention was greater for the treatment group, it was likely to be for one reason (besides Type I error). Namely, the decre- ment in attention was specific to the stimulus that was 28 presented on trials 1 to 8 and was not the result of receptor fatigue, drowsiness, sensory accommodation or other peripheral factors. Therefore, two conditions were necessary for habituation to be demonstrated. First, there must be a decrease in response magnitude with repeated "experiences" or presentations of the stimulus. Second, it must be shown that the decrement was stimulus-Specific. The latter requirement was addressed by including the test for response recovery. An important concern here had to do with potential differential response- eliciting powers of the decrement and recovery stimuli. That is, an apparent dishabituation effect might simply obtain because the recovery stimulus was a more powerful response elicitor than the decrement stimulus. If this were the case, then the response decrement over trials 1-8 could not be clearly interpreted as being habituation. One method of handling this possible confounding would have been to pretest the pool of possible stimuli to determine which ones were comparable response elicitors. This was not done for one important reason. Namely, there was no basis for choosing the response with which to com- pare the response-eliciting powers of the stimuli. Since there were multiple responses for both habituation tasks it seemed reasonable not to make what would have to be an arbitrary decision on this matter. Rather, the decrement and recovery stimuli in each habituation task were simply 29 counterbalanced across subjects. Thus, each stimulus was used equally often as the decrement and as the recovery stimulus. Experimental Design The study was longitudinal with all infants taking part in visual and auditory habituation tasks at two and at four months of age. The order of task was counter- balanced across subjects, and the groups were equally composed of males and females. Thus the full design was a 2 (treatment/control) x 2 (sex) x 2 (order of habituation task) x 2 (age) x 2 (modality of stimulation). The last two factors were repeated measures, and there were four subjects per cell. The design is represented in Figure 2. Referring to Figure 2, the three factors across the tOp of the design matrix were between subjects factors, while factors listed along the left of the matrix were within or repeated measures factors. The only restriction on random assignment was the desire to have equal numbers of both sexes in all cells. Thus equal numbers of males and females were randomly assigned to combinations of treatment/control and order of habituation task. The treatment group (decrement/recovery) received one stimulus on trials 1-8 and a different stimulus on trials 9-10. Con- trol group infants were presented the same stimulus for all ten trials. Order of habituation task connotes simply that the sequence of testing (visual followed by auditory 30 Treatment (D/R) Control (D) 81 S2 S1 82 O O $3233 :3 ubLaJNH 25 29 26 30 27 31 28 32 13 17 n10 14 18 n22 n11 15 19 n12 16 20 n24 :3 :5 :3 :3 ooqmm :5 :5 :3 :3 :1 :3 .‘3 5 33:3 :3 5 n 1 NOTE: D/R: ...n See text for complete explanation of the figure. Represents groups presented a decrement stimulus on trials 1-8 and a recovery stimulus on trials 9-10 Represents groups presented only the decrement stimulus for all 10 trials Order of task (0 : visual then auditory; 02: auditory then visual) Male Ml: Visual habituation task Female M2: Auditory habituation task 2 months of age 4 months of age Represent each infant in the study Figure 2.--Matrix Depicting the Design of the Study. 31 for 01’ or vice versa for 02) was counterbalanced across infants. Thus, there were four infants in each of the eight cells which resulted from the combination of treat- ment, sex, and order of habituation task. The within subjects factors were age and modality of stimulation. In short, all infants took part in both visual (M1) and auditory (M2) habituation tasks at two (Al) and at four (A2) months of age. Therefore, though not specified in Figure 2, the same_32 subjects that appeared in the A M combination of factors were also 1 1 present in AlMZ’ AZMl’ and A2M2° Finally, it should be specified that once an infant was randomly assigned to a particular combination of conditions 11m“: was his placement for the entire study. For example, as can be seen in Figure 2, assignment to the treatment condition was for both tasks at both ages. Similarly, once order of task was specified for a subject, the same order was used at both ages. Choice of Stimuli Two auditory and two visual stimuli were chosen randomly from pre-selected pools of possible stimuli. The pool of auditory stimuli consisted of two sets of six pure simple tones exufli with ranges of 200-700 Hz and 1000’ 1500 Ina. The tones increased in steps of 100 Hz. Random numbers were assigned to all tones and one was selected from each set of six. In this manner it was determined 32 that a 500 Hz and an 1100 Hz tone would be used in the present study. The tones were played at 72 db since this intensity appears to be appropriate for eliciting heart rate deceleration (Berg, Berg, & Graham, 1971). It should be noted also that tones in the two pre- selected ranges have been shown to be discriminable in six- month-old infants (e.g., Horowitz, 1972). Further, by the restricted random selection of tones, i.e., setting two nonoverlapping .ranges