Stranger distress: Mother and stranger as reinforcers

Stranger distress: Mother and stranger as reinforcers

INFANT BEHAVIOR AND DEVELOPMENT 2, 309-317 (1979) Stranger Distress: Mother and Stranger as Reinforcers* GREGORY FOUTS University of Calgat3' AND PAT...

426KB Sizes 0 Downloads 72 Views

INFANT BEHAVIOR AND DEVELOPMENT 2, 309-317 (1979)

Stranger Distress: Mother and Stranger as Reinforcers* GREGORY FOUTS University of Calgat3' AND PATRICIA ATLAS UniversiO' of Colorado

Six- and nine-month-old male and female infants were presented with their mothers, strangers, and a scrambled face contingent upon bar-pressing. For six-month-olds, the mother and a stranger were expected to possess reward and neutral reinforcer properties, respectively, as shown by an increase and no change in behavior compared with that in a scrambled face baseline. For ninemonth-olds, the mother and a stranger were expected to possess reward and punishing properties, respectively. The predictions were confirmed and discussed in terms of the operant paradigm providing a sensitive method of" assessing the differences in reinforcer properties of mothers and strangers, commonly labeled as "attachment" and "stranger distress."

Before the sixth or seventh month, the occurrence of stranger distress is relatively rare (e.g., Ainsworth, 1963, 1967; Morgan & Ricciuti, 1969; Scarr & Salapatek, 1970; Drcarie, 1974). However, after the seventh month and usually before the thirteenth month, stranger distress can often be demonstrated (e.g., Morgan & Ricciuti, 1969; Ddcarie, 1974). The variability of onset of stranger distress reflects the diversity of methods and criteria of assessment as well as the considerable variations in cognitive development (e.g., Piaget, 1954, 1972; Flavell, 1963), maturation (e.g., Ames, 1967), and/or differences in experience among infants during this period. Assuming that cognitive and maturational changes beyond the seventh mo.nth predispose infants to similar learning experiences, it was hypothesized that infants learn that the presence of a stranger is discriminative for fewer or no ,

*Portions of this study were presented to the American Psychological Association. New Orleans, September, 1974. Requests for reprints should be sent to Gregory Fours, Department of Psychology, Universityof Calgary, Calgary, Alberta. Canada, T2N IN4. 309

310

FOUTS AND ATLAS

rewards and/or punishment (e.g., Gewirtz, 1961); therefore, a stranger acquires punishment properties. The rationale are these: First, a stranger may be associated with qualitatively and/or quantitatively less positive reinforcement when the mother is present, e.g.', different kinds of affectional behaviors, less consistent feeding. Second, the presence of a stranger may signal the withdrawal of the agent of reinforcement (mother), e.g., a babysitter signals the leaving of the parents. Third, a stranger may produce noxious stimuli, e.g., awkward handling. Fourth, mothers reinforce negative behaviors in the presence of strangers, e.g., when infants cry in the presence of strangers, mothers typically hold or talk to them. Fifth, infants are more likely to be hurt in the absence of mother, e.g., hitting head, falling off bed, and to the degree that the presence of a stranger is discriminative for the absence of mother, a stranger may acquire negative reinforcing properties. In his study, an infant was allowed to control the amount of time of looking at either the mother or a stranger depending on how long a lever was depressed. A baseline operant rate was obtained with a scrambled schematic face being presented contingently in the absence of mother. A scrambled face was used rather than a normal face, since pilot work showed that with the former, a moderate and stable rate of responding could be maintained. Contingent presentation of the scrambled face was used so that any comparisons between the baseline and the contingent presentation of mother or stranger would not be confounded with the difference of the presence of a contingency, i.e., if responding with mother was greater than with the scrambled face, it would not be due to the control or contingency being in itself reinforcing. After contingent presentation of mother (or stranger), another baseline occurred, followed by the contingent presentation of stranger (or mother), with another baseline at the end. It was expected that there would be greater responding with the contingent presentation of mother than a scrambled face, i.e., discrimination, a finding which would be consistent with that of Wahler (1967) who used mother to deliver a sequence of social reinforcement contingent upon smiling. It was expected that for six-month-olds, there would be little difference in responding between the scrambled face and stranger conditions, since the several maturational, cognitive, and experiential factors would not have had their impact. But by nine months, the stranger condition was expected to suppress responding compared to the scrambled face, since the former would now possess punishing properties. Also, the decrease in responding with the stranger would not be due to the withdrawal of mother, a common confounding in many studies, since both the scrambled face and stranger conditions involved mother absence. Thus; for sixmonth-olds, mother-stranger discrimination was expected, due to the discrimination between mother and scrambled face but not between stranger and scrambled face. For nine-month-olds, mother-stranger discrimination was expected to continue, with the bases now being the continued attachment (as reflected by mother resulting in more responding than the scrambled face) and stranger distress (reflected by a stranger producing less responding than the scrambled face).

STRANGER DISTRESS

31 1

METHOD Subjects Twelve six-month-old infants (six of each sex; age range, 5.9-6.4 months; mean age, O. 1 months) and 12 nine-month-olds (six of each sex; age range, 8.8-9.5 months; mean, 9.3 months) served as subjects. All infants were normal, white, and home-reared.

Apparatus and Experimental Situation Two rooms were separated by a one-way glass. The experimental room contained a high chair and a lever which protruded from a box (which was between the chair and one-way glass). On the end of the lever was a brightly colored plastic cap (2.5 cm diameter) and the bottom edge of the cap was 5.1 cm above the tray on the highchair. The highchair was placed facing the one-way glass and infants could touch the plastic cap with either hand but could not put their mouths on the cap. Vertical or horizontal displacement (1.3 cm) of the end of the lever in either direction activated one of four microswitches, which in turn (a) activated a pen in an event recorder located in the adjacent control room, and (b) turned on two floodlights in the control room which allowed infants to see into the control room through the one-way glass (when both rooms were otherwise darkened); thus, as long as infants kept the lever displaced, they could see into the control room. When the lever was released, it automatically returned to a resting position, resulting in deactivation of the microswitch. The control room contained an event recorder, two floodlights attached to the wall on either side of the one-way glass (directed into the control room), a life-size black-and-white picture of a scrambled face attached to a panel which could be placed 15.2 cm from the one-way glass, a male experimenter, a mother, and another mother who served as a stranger.

Procedure An infant was placed in the highchair and a safety strap was tightened, after which (a) the room was darkened so that the lever was barely visible and there were no reflections of light on the one-way glass from an infant's point of vision; (b) the experimenter took the infant's hand and guided it to the plastic cap and depressed the lever for approximately 5 sec, resulting in the floodlights (in contrN room) transilluminating the glass and making visible the scrambled face (all infants immediately looked at the glass), the lever was then released for 5 sec, and again depressed and released for 5 sec each; and (c) the mother and the experimenter left the room and entered the control room. The procedure of having mothers leave when the room was dark and the fact that infants were attending to the scrambled face had the decided advantage of eliminating most of

312

FOUTS AND ATLAS

the distress often found in nine-month olds when mothers leave their presence. The experimenter turned on the event recorder and a sequence of five conditions was begun. Initially, all infants received a 3-minute baseline in which lever movement illuminated the scrambled face. After this initial baseline of contingent scrambled face, the face panel was removed and the mother or a stranger was directed to stand in front (15.2 cm away) of the one-way glass (the experimenter positioned the person). The mother or stranger was asked to remain in that position (3 minutes) and to keep a smile on her face; only her head was visible when illuminated. This manipulation constituted the experimental condition, after which another 3-minute scrambled face condition was begun, followed by another experimental condition (3 minutes), and finally another 3-minute scrambled face condition. There were two orders of baseline and experimental conditions: Scrambled Face-Mother-Scrambled Face-Stranger-Scrambled Face (SF-M-SF-S-SF) and Scrambled Face-Stranger-Scrambled Face-MotherScrambled Face (SF-S-SF-M-SF). Half of each age group was tested under each order; and within each order, there was an equal number of male and female infants. The stranger condition was manipulated by having each mother serve as a stranger for another infant in the subject sample (this, of course, necessitated several mothers returning a second day to serve as a "stranger"). Therefore, the 24 infants in the mother condition were exposed to their 24 respective mothers, and the same infants were exposed to 24 strangers. This procedure was used in order to eliminate the possibility that a preselected stranger used for all infants might have some unusual attributes which could confound the results, as well as to add generalizability to the effects of strangers.

RESULTS Besides the 24 infants whose results are presented below, nine six-month-olds and six nine-month-olds were tested but eliminated from the study because of (a) failure in = 6) to press the lever in the initial scramble face condition for a minimum of five times in the last two minutes (this criterion was imposed to ensure an adequate operant rate; the younger infants were typically eliminated for this reason); (b) uncontrollable crying (n = 4), more common for older infants in the Stranger condition; (c) the mother not being able to serve as a stranger for another infant (n = 3); and (d) infants (n = 2) falling asleep. For each of the 24 infants, the frequency of lever-pressing under each of the five conditions was determined. Statistical comparison of the two presentation orders revealed no significant differences (Mann-Whitney U test, p > .10) between the mother conditions, stranger conditions, and the scrambled face baselines. Therefore, the five scores for each infant were cast into two matrices, Scrambled Face-MotherScrambled Face (SF-M-SF) and Scrambled Face-Stranger-Scrambled Face (SF-S-SF), by taking the three scores associated with these three conditions

STRANGER DISTRESS

313

regardless of the order of testing, e.g., for the SF-M-SF matrix, the first three scores from the SF-M-SF-S-SF order were used, as well as the last three scores in the SF-S-SF-M-SF order. It should be pointed out that in this way the scores of each infant are represented in both matrices, thus making possible within-subject comparisons within each matrix and between matrices. Due to (a) heterogeneity of variance in responding under the various conditions and (b) numerous zero scores which prevented data transformations, nonparametric tests were used throughout the study. The median responding for the six-month-olds is presented in Figure 1. A comparison of responding in the scrambled face conditions preceding the mother and stranger conditions (comparing between matrices) revealed no significant difference (T = 37, N = 12, p > .05; Wilcoxon test for related samples). Young infants responded significantly more when mother was contingent upon leverpressing than when stranger was presented (T = 3, N = 12, p < .01; Wilcoxon). In order to determine the reasons why the difference between mother and stranger conditions occurred, e.g., was it due to an increase in responding with mother and/or a decrease with stranger, the responding in each of the Scrambled FaceMother-Scrambled Face and Scrambled Face-Stranger-Scrambled Face matrices was separately examined. In the Scrambled Face-Mother-Scrambled Face matrix, the Friedman test (Siegel, 1956) revealed a significant difference among the Scrambled Face and Mother conditions (X" = 9.38, df = 2, p < .01). Follow-up pairwise tests (Wilcoxon) revealed (a) a significant increase from the initial scrambled face condition to the mother condition, T = 1, N = 1 l , p < .01; (b) a significant decrease from the mother to the following scrambled lace condition, T = 14, N = 12, p < .05; and (c) a significant increase in responding from the initial to the following scrambled face condition, T = 13, N = 12, p < .05. The Friedman test applied to responding in the Scrambled Face-Stranger-Scrambled Face matrix revealed no significant difference among the three conditions (X2 = 4.3, df = 2, p > . 10); therefore, no following tests were used. These findings show that the difference in responding between the mother and the stranger conditions was due to an increase in responding under the mother condition, not to changes in responding when a stranger was presented. Identical analyses were performed on the median responding of ninemonth-olds. Comparisons of responding between the scrambled face conditions preceding the mother and stranger conditions revealed no significant difference (T = 24, N = 12, p > .05; Wilcoxon); nor was a significant difference found between the scrambled face conditions following the mother and stranger cgnditions (T = 17, N = 11, p > .05). An examination of responding under the mother and stranger conditions revealed significantly more responding under the mother than the stranger condition (T = 00, N = 12, p < .01 ). The Friedman test applied to the data in the Scrambled Face-Mother-Scrambled Face matrix revealed a significant difference among the three conditions (X2 = 14.0, df = 2, p < .001), and follow-up pairwise tests (Wilcoxon) revealed (a) a significant increase from the initial scrambled face condition to the mother condition (T = O,

ILl

~I..L

IIIIIIIIIIIIIIIIIIII!111111111111111111111111111111~

-- ~- ~0

i "E .E -a

9 o E

IIIIIIIMI

~L.I-

"E o

OJ

~J

--

D.

u.J _J

Ii

IIMI

I

o~

I

I

•~ -

e~ L~

LI..I

IIIIIIIIIIIIII1~

-

I

~'

I.,U _,,I o ~ I..U

i~l II~

0:~ I.L. o9

I

I

I.~ C~I

0 OJ

SS3Wd-W3A37 314

o~

I I_Q ~

30 13N3n~3W3

I C

NVlO3N

o~ iZ

STRANGER DISTRESS

31 5

N = 11, p < .01); (b) a significant decrease from the mother to the following scrambled face condition (T = 1, N = 12, p < .01); and (c) no significant difference between the initial and following scrambled face conditions (T = 17, N = 11, p > .05). The Friedman test applied to the data of the Scrambled Face-Stranger-Scrambled Face matrix revealed a significant difference among the three conditions (X2 = 11.4, df = 2 , p < .01 ); and follow-up tests (Wilcoxon) revealed (a) a significant decrease from the initial scrambled face condition to the stranger condition (T = 8, N = 12, p < .02); (b) a significant increase in responding from the stranger to the following scrambled face condition (T = 0, N = 11, p < .01); and (c) no significant difference between the initial and following scrambled face conditions (T = 25, N = 12, p > .05). The results for the nine-month-olds indicate that the difference in responding between the mother and stranger conditions is due to a significant increase in the mother condition and a significant decrease in the stranger condition. DISCUSSION The results indicate that at six months of age, infants can discriminate between mother and a stranger, with the mother being a positive reinforcer and a stranger having no more positive or negative reinforcing properties than a scrambled face. This latter trend suggests that a stranger may have some positive reinforcing properties in the sense of maintaining the operant, or it may merely be the case that the contingency, the control of the environment, was reinforcing independent of the stranger. These findings are consistent with the view that motherstranger discrimination does not necessarily imply the existence of stranger anxiety commonly indexed by the absence and presence of positive and negative behaviors, respectively. For the nine-month-olds the mother continued as a positive reinforcer, but a stranger had acquired punishing properties as evidenced by the significant decrease in responding when a stranger was presented. Although it could be argued that the decrease in responding in the stranger condition may reflect less reward value than a scrambled face, rather than a punishing effect, this position appears unlikely since informal observation revealed (a) a precipitous decline in behavior following the first two presentations of a stranger, and (b) greater occurrence of sobering and fussing in the stranger condition than either the mother or scrambled face conditions, with the occurrence of smiling in the stranger condition being rare and occurring frequently in the mother and scrambled face conditions. One implication of the findings and the procedure is that the operant paradigm is sensitive procedure for assessing some of the dimensions associated with stranger distress; it also serves as a converging operation which adds credence to attachment and stranger distress. The operant paradigm may be used to assess the differences in reinforcing properties associated with mothers, strangers, and other familiar persons (e.g., fathers, siblings), thereby providing a means by which the attachment and stranger distress literature may be integrated.

316

FOUT$ AND ATLAS

To illustrate, an examination of Figure 1 reveals that for six-month-olds, the increase in responding between the baseline and mother conditions was 100%, with the increase for nine-month-olds being approximately 41%. This may suggest that the attachment is waning (although it could be a ceiling effect), and perhaps other familiar persons are acquiring reinforcing properties, i.e., dependency relations are developing. During this research project, there were two occasions when a father accompanied the mother and their nine-month-old infhnt; the infants were tested in an identical manner as in the present study, but with the fathe r serving as the "'stranger" rather than another female. Both infants exhibited greater increases in behavior with the father than with the mother, indicating that by this age and in this situation, the father had more reinforcing properties than the mother. This trend seems inconsistent with that found by Cohen and Campos (1974); however, direct comparison of the two studies is difficult, since the latter used response preference between mother and father, which perforce confounds the investigation of the separate contribution of the effects of each stimulus, e.g., was the mother more rewarding than father, the mother a neutral reinforcer witfi the father a punishing stimulus, the mother a rewarding stimulus and father a neutral reinforcer'? Another implication of this demonstration concerns the labeling of behaviors as indices of stranger "anxiety" or "distress"; these behaviors may be but behavioral and emotional concomitants of the differential reinforcing properties of the mother and a stranger. For infrahuman organisms, interfering and emotional behaviors are quite common when positive reinforcers are eliminated (extinction) and/or negative reinforcers are presented; therefore, it is not surprising to find similar reactions in very young humans.

ACKNOWLEDGMENTS The authors wish to thank Deborah Walder for her help in pilot work which led to the refinements in this study and Joseph Campos for his suggestions in the preparation of this manuscript.

REFERENCES Ainsworth, M. The development of infant-mother interaction among the Ganda. In B. loss (Ed.), Determhumts of it~mt behavior II. London: Methuen, 1963. Ainsworth, M. h~mc3" h~ U.gtttltltt: h~mt ctwe and the growth of love. Baltimore: John Hopkins University Press, 1967. Ames, L. The sequential patterning of prone progression in the human infant. P.s:vchology Monographs, 1937, 19,409--460. Cohen, L., & Campos, J. Father, mother, and stranger as elicitors of attachment behaviors in infancy. Developmental Psychoh)gy, 1974, I0, 146--154. Ddcarie T. The infimt's reaction to strangers. New York: International Universities Press, 1974. Flavell, J. The developmental psychology of Jean Piaget. New York: Van Nostrand, 1963.

STRANGER DISTRESS

31 7

Gewirtz, J. A learning analysis of the effects of normal stimulation, privation and deprivation on the acquisition of social motivation and attachment. In B. Foss (Ed.), Dett, rmhttln/.~' of it~m/ behavior Ill. London: Methuen, 1961. Morgan, G., & Ricciuti, H. Infant's response to strangers during the first year. In B. Foss (Ed.), Determhumts of infimt hehaviour IV. London: Methuen, 1969. Piaget, J. The construction of reality hi the child. New York: Basic Books, 1954. Piaget, J. The chifd and reality. New York: Grossman, 1972. Scan-, S., & Salapatek, P. Patterns of fear development during infancy. MerrilI-Pahner Quarterly of Behavior and Development, 1970, /6, 53--90. Siegel, S. Nonparametric statistics for the behavioral sciences. New York: McGraw-Hill, 1956. Wahler, R. Infant social attachments: A reinforcement theory interpretation and investigation. Child Development, 1967, 38, 1079-1088.