_“ ww- m; n 7533137 MES-‘1. ' :7; 7: -. tie. a, a; fite Usa‘é‘v fiffim l This is to certify that the thesis entitled THE DIFFERENTIAL INFLUENCE OF SOCIAL REPUTATION IN THE SOCIAL DEVELOPMENT OF REJECTED AND ISOLATED CHILDREN presented by Fred Arthur Rogosch has been accepted towards fulfillment of the requirements for M.A. degree in Psychology £044 MW Major professor Date 8-11'82 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution MSU LIBRARIES .—_—. RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. THE DIFFERENTIAL INFLUENCE OF SOCIAL REPUTATION IN THE SOCIAL DEVELOPMENT OF REJECTED AND ISOLATED CHILDREN By Fred Arthur Rogosch A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Psychology 1982 ABSTRACT THE DIFFERENTIAL INFLUENCE OF SOCIAL REPUTATION IN THE SOCIAL DEVELOPMENT OF REJECTED AND ISOLATED CHILDREN By Fred Arthur Rogosch The influence of social reputation on first-, third-, and fifth- grade boys' and girls' (N=182) social interaction and effectiveness was investigated. Each child completed a peer nomination sociometric, a class play sociometric, and predicted behavior for children of dif- ferent reputations in hypothetical social interaction vignettes. Diver- gence in mutual friendship involvement distinguished isoiated and rejec- ted children. Both groups lacked mutual friends in first grade. How- ever, fifth grade isolated children attained mutual friendships while rejected children remained limited in mutual friendship relations. Divergent social reputations were attributed to isolated and rejected children by peers. Rejected children were restricted by an interper- sonally rejecting social reputation and concomitant peer expectations fbr negative social behavior while isolated children evidenced a non- distinct reputation and less rigid expectations from peers. Social reputation's role in regulating social behavior and constraining effec- tive social development for rejected children and the need for peer group intervention were discussed. ACKNOWLEDGMENTS My thanks to . . . Dr. Andrew F. Newcomb for his diligent support, guidance, and enthusiasm; Dr. Ellen A. Strommen. Dr. Gary E. Stollak, and Dr. Lucy Rau Ferguson for their valuable input; ' William M. Bukowski, Julie Jueneman, and Judith C. Meister for their gen- erous assistance in interviewing the children; and The administration, faculty, and students of Fowlerville Elementary School fbr their cooperation. ii TABLE OF CONTENTS INTRODUCTION METHOD. Subjects. Procedure Sociometric Class Play. Vignettes . Measures. Sociometric Measures: Class Play Measures . Vignette Measures RESULTS Sociometric '. Mutual Friendships Class Play Percent of Nominations Acceptance and Dominance Dimensions. Sex of Nomination Consensus of Choice . Vignette Predictions Predicted Type Frequencies Prediction-Reputation Consistency DISCUSSION . REFERENCES . Table LIST OF TABLES Mean Number of Mutual Friendships of Social Effectiveness Groups by Grade . . . . Patterns of Friendship Nomination for Star, Average, Isolated, and Rejected Children by Effectiveness Group and by Grade . . . . . . Intercorrelations between Reputation Types Approximating Circumplex Structure . . . . Mean Percent of Nominations for Class Play Roles by Social Effectiveness Group . . . . Mean Percent of Nominations for Class Play Roles by Sex . . . Proportion of Male and Female Nominations Made by Boys and Girls for the Class Play Roles . Mean Consensus of Choice for the Class Play Roles by Grade . . . . . . Mean Number of Predictions for Each Prediction Type by Grade Level . . Mean Consistency between Reputation and Predicted Behavior by Reputation Type . . . iv Page 16 18 19 21 23 26 28 3O 31 Figure LIST OF FIGURES Page Circumplex Model Representing Relationship between Reputation Types as Conceptualized along the Dimensions of Dominance- Submission and Acceptance-Rejection . . . . 12 Relationship between Total Acceptance and Total Dominance Dimensions for Boys and Girls and for the Social Effectiveness Groups . . . . . . . . . . 24 -INTRODUCTION Although children's social reputations appear to be a significant determinant of their social interactions in the peer group, the process by which social reputations influence social relations has been largely overlooked. Social reputations have been conceptualized as operating through collective peer group expectations for the characteristic behavior of individual children (Dodge, 1980). These expectations may bias social interaction, limit the types of social experience available to targeted children, and restrict friendship relations. In turn, individual children present social behavior which is a fulfillment of the peer group's expectations. Evaluation of the characteristics which peers attribute to one another has traditionally been accomplished through nomination proce- dures such as Bower's (1960) class play and similar sociometric methods. These procedures require children to nominate peers who best character- ize a range of personality attributes. Gronlund and Anderson (1957), for example, used this method to assess personality characteristics of socially accepted, socially rejected, and socially neglected or isolated preadolescents. Socially accepted boys and girls were characterized by being friendly, likable, enthusiastic, cheerful, and having a good sense of humor. Socially rejected children were noted as being restless, talk- ative, and not liked as well as lacking the positive characteristics attributed to accepted children. In contrast, socially neglected 2 children seemed to be overlooked by peers receiving only rare nominations for any of the attributes except for some mention as to being quiet. Interestingly, undersirable aggressive behavior, while not being attrib- uted to girls, did not discriminate among boys of the three sociometric groups. The pattern of results from investigations employing nomination procedures (e.g., Bower, 1960; Asher and Hymel, 1981) has typically been interpreted as an assessment of the personality attributes or behavioral styles of individual children. While the validity of this interpreta4 tion is plausible, an aIternative interpretation can be that the pattern of results is also indicative of the social reputations maintained by the peer group about individuals. This alternative viewpoint changes the focus from an examination of what individual children are like to the manner in which the peer group collectively appraises individual children. As a result of the peer group's differential characteriza- tion of children at varying levels of social effectiveness, the social environment in which children of these different groups interact may be regulated by different sets of expectatiOns for interpersonal behavior based on social reputations. These differential expectations could substantially influence the types of social interaction experienced by children and markedly influence their Social development. In a recent review by Barley and Fazio (1980), a model from adult social psychology is presented which has relevance to expectation proc- esses which may be operative within children's social environments. The model proposes that a perceiver's expectations about an individual may be confirmed in a number of ways independent of the behavior of the individual. For example, expectations may guide decisions on whether 3 interaction with an individual is initiated. If the perceiver main- tains negative expectations about an individual, interaction with this individual may be avoided or quickly terminated. This has the result of preventing disconfirmation of the perceiver's expectation because * the individual does not have the opportunity to interact with the per- ceiver in order to alter the expectation which will be operative in later encounters. In actual interactions between the perceiver and the individual, direct confirmation of the expectation by the individ- ual's behavior is possible, and if confirmed, the perceiver's convic~ tions about the individual would be strengthened. However, the per- ceiver's expectations themselves may result in him/her behaving differ- ently toward the individual, and the individual's response may be a reaction to the perceiver's unintentional cues rather than an indica- tion of characteristic behavior. The perceiver's expectations may also be maintained when an individual behaves in an ambiguous manner. Here the perceiver is likely to be biased by previous expectations and judge the individual's behavior as confirmatory. Similarly, the individual may even act in ways which are contradictory to the perceiver's expec- tations. Yet, the perceiver may maintain the expectation by attributing the individual's behavior to transient situational determinants rather than to true dispositional qualities. The results of these interactions have consequences for the indi- vidual. Avoidance because of the perceiver's expectations will limit the frequency with which soCial interactions are available. Further, while outcomes from actual interactions with the perceiver are probably less than favorable, the individual may come to expect social situations to be less rewarding and behave more in accord with what is expected of him/her. In this way an individual's behavior may be channeled toward agreement with that which is expected of him/her. These expectation processes have been demonstrated empirically among college undergraduates in manipulations of a "perceiver's" expec- tations about the friendliness/unfriendliness of an individual (Snyder, Tanke, and Berscheid, 1977). In a social encounter, the perceiver was seen to unwittingly modify his behavior toward the individual in accord with the contrived expectations. In turn, the individual was seen to respond to the perceiver in a manner consistent with the perceiver's expectations (friendly/unfriendly), Thus, the individual's behavior was a reaction to cues produced by the perceiver's expectations rather than a demonstration of personal characteristics. However, for the per- ceiver, previous expectations for the individual had been objectively confirmed. In future interactions, the perceiver's maintained expec- tations would again be assumed to operate thereby increasing the like— lihood of similar interactions. The expectations shared by the peer group which constitute a social‘ reputation may also cause children to experience a generalized bias in social interaction. Children of varying reputations have been shown to differ in the amount and type of behavior which they direct toward others and which others direct toward them (Campbell and Yarrow, 1961; Yarrow and Campbell, 1963). Children of positive reputations (e.g., leader, helpful) were allowed greater freedom in the type of behavior they could perform. For example, these children could perform aggressive behavior without being evaluated negatively. As a result children of positive reputation may have the opportunity to gain greater skill in a range of behavior maintaining and fOrtifying their social competence. On the other hand, children of negative reputations (e.g., afraid, angry) were more restricted in initiated behavior and in behavior directed to them. This could lead to less experience in social interaction, continued retardation of social skills, and increasingly poor social competence. Reputations tended to be quickly formed and retained greater stability than did behavior. Children's descriptions of their peers showed con- siderable discrepancy with actual observed behavior. Winder and Higgins (1964) found a similar discrepancy for the behaviors and reputations of aggression and dependency. Moreover, rather than behavior leading to formulation of one's reputation, there were indications that over time reputation tended to elicit behaviors which were consistent with the reputation. Dodge (1980) has shown that the reputation for being aggressive has implications for interaction with peers. A sociometric procedure was used to identify boys in grades two, four, and six who were aggressive and non-aggressive. These boys responded individually to hypothetical stories in which they were involved in a situation with a negative out- come as a result of the behavior of a peer known to them. The intention of the peer was left ambiguous. The subjects were more likely to inter- pret the situation as motivated by hostile intentions if the other child had an aggressive reputation. This confirmed their expectations of the child as being aggressive, and they were more likely to retaliate with aggressive behavior which they felt was justified. Subjects expected continued aggression from the child and also mistrusted.him. Having more aggression directed against him, the child with the aggressive reputation was more likely to view his peers as hostilely motivated and to continue to retaliate with aggression himself which he felt was 6 justified. A cyclical self-perpetuating relationship develops leading to greater aggressive reputation, increasing aggressive behavior, and social rejection. The negative results of this label were shown to increase over time. These investigations support the view that social reputations are operative in children's social environments and result in differential influences in the course of social development. The adverse effects associated with negative social reputations appear to result in pro- gressive social rejection by peers, a significant warning sign of psy‘ chological risk. Roff, Sells, and Golden (1972) found a higher preva- lence of later delinquency among those children who were rejected by their peers. Similarly, Cowen, Pederson, Babigian, 1220, and Trost (1973) found that negative evaluation from peers in a class play was the most powerful predictor of later psychological disturbance. Addi- tionally, Cowen, et al. speculated that peer group perceptions (i.e.. social reputations) could have resulted in different behavior being directed toward these children, exacerbating their difficulties in social adjustment, and contributing to their later psychiatric dif~ ficulties. Investigation of the process of social reputation thus seems para- mount. It is important to include in this investigation examination of a wide range of reputational types rather than focusing on one or two specific social reputations. Higgins (1979) points out the advan- tage of circumplex conceptualizations of interpersonal behavior. Cir- cumplex models have the capacity to tap a broad domain of interpersonal characteristics while illustrating the relationship between the differ- ent attributes by ordering them within a two-dimensional framework, 7 Campbell and Yarrow (1961) attempted to use reputational categories derived from Leary's (1957) circumplex model, However, gaps were present in the range of reputations which they selected. A more complete representation of social reputations appears warranted. The present study utilized a fuller circumplex model of reputational categories similar to Leary (1957) to investigate the development of social reputation in the elementary school years. A sociometric proce- dure was used to evaluate the social effectiveness of participating boys and girls as well as to determine the extent of their integration into the peer group by examining the degree of involvement in mutual friend- ships. A class play nomination was used to examine the social reputa- tions which the peer group maintains for boys and girls of different social effectiveness levels. Measures of how the peer group made their reputational assignments were also determined. Finally, predicted behav- ioral expectations for children of different reputational categories were obtained in hypothetical situations to investigate the degree of consistency between reputation and expected behavior. METHOD Subjects One hundred eighty-two children, enrolled in a large elementary school in semirural Michigan, participated in the investigation. The sample was comprised of approximately equal numbers of boys and girls in three classrooms each at grade levels first, third, and fifth (n_= S9, 63, 60, respectively). The mean ages for subjects in the three grades were 7-0, 8-11.10-11. In a sociometric procedure described below, four social effectiveness groups were delineated. Five boys and five girls at each grade level were selected for each group reducing the final experimental sample used in the data analysis to 120. Procedure Each child was escorted to a private room at the school for an individual interview session in which the child was assured that all responses would be kept confidential. Three tasks were completed during the interview: 1. a free-choice/forced choice sociometric nomination; 2. a modified class play nomination; and 3. a simulation procedure in which the child predicted behavioral responses in hypothetical vignettes portraying social interaction sequences. Additionally, free response descriptions of a random group of the childis Classmates were obtained. However, this procedure is not considered in the present investigation. Sociometric. Names of same-sex classmates were printed on separate cards. These cards were displayed to the child who was asked to read each name aloud to verify that the child recognized his/her classmates' 8 names. The child was then asked to select his/her three best friends, The cards representing these children were removed, and the child was asked to select three classmates with whom he/she would least like to play. The remaining cards were collected and presented to the child one at a time. The child was asked to indicate whether or not he/she would like to play with the child whose name was presented. Class Play. Next, in a modification of Bower's class play method- ology (1960), the child was instructed to pretend that his/her class was going to put on a play, ant that he/she needed to decide who could best portray the various roles in the play. The child was told that the best way to make decisions was to choose classmates whom most chil- dren would think would play the role best. The eight roles used in the modified class play included someone who: 1. would be a good leader; 2. tries to help everyone; 3. is nice and follows directions; 4. is quiet and shy; 5. is afraid to do most things; 6. is often angry and complaining; 7. gets into fights often; and 8. is mean and bossy. The order of the item presentation was randomized across participants. Vignettes. The final task required that the child pretend that he/she was engaged in a series of social interaction situations with different children. Each of the subject's choices for the eight class play roles was randomly assigned to One of eight hypothetical vignettes‘ portraying a conflict situation between the subject and the chosen child. The eight vignettes were as follows: Pretend that you and (chosen child's name) - 1. are playing records and one gets broken; 2. are eating candy bars, and your candy bar falls on the ground and gets dirty; 3. are play- ing a game, and (child's name) loses; 4. are riding bikes. You fall off your bike; 5. had a fight. Later that day you see him/her; 6. are 10 painting pictures and paint gets Spilled on his/her picture; 7. are play- ing a game and (child's name) wins; and finally 8. Pretend that a group of older kids are making fun of you on the playground and (child's name) comes by. The subject was asked to predict what the other child would be expected to do in the vignette. Responses were recorded verbatim, and the order of the presentation of the vignettes was randomized across subjects. Measures Each of the three tasks completed by the child -- the sociometric nomination, the class play nomination, and the vignette prediction -- yielded different groups of measures as discussed below. Sociometric Measures. The number of times each child was chosen by a classmate as a friend and the number of times he/she was chosen as disliked were compiled. A binomial probability model of sociometric classification (Newcomb and Bukowski, in press) was used to distinguish four social effectiveness groups, star, average, isolated and rejected children, based on rare total scores for the friends and disliked meas- ures. A criterion value of p .¢.10 was used to determine rare combina- tions of friend and disliked total scores. Stars were children who had many friends and few or no one who disliked them as determined by the probability criterion; rejected children had many who disliked them and few or no friends; isolated children had both few friends and few who disliked them; and average children had some friends and some who dis- liked them with neither total score being rare.. Five boys and five girls at each grade level were selected for each of the four social effectiveness groups. These children best satisfied the definitional requirements of the group to which they were assigned. 11 The sociometric data were also examined for the number of mutual friendship choices between children. A mutual friendship was defined as a pairing of friendship choices in which two children nominated each other as friends. The total number of mutual friendships in which each child was involved was determined, and a score ranging from zero to three was given. Class Play Measures. The percent of nominations each child received for each of the class play roles was computed. The content of the class play roles had been determined to be understandable to children in a pilot investigation. The eight class play roles were selected as descriptbrs for the octants of Leary's (1957) circumplex system. This model varies the class play roles along the two interpersonal dimensions of acceptance-rejection and dominance-submission. The theoretical arrangement of the class play roles is presented in Figure 1. Leader, helpful, nice, and quiet-shy represented the accepting half of the circumplex and differed to the extent to which they were relatively dominant or submissive. Likewise, mean-bossy, fights, angry—complain- ing, and afraid represented the rejecting half of the circumplex while being varied on the dominance-submission dimension. The circumplex was similarly divided into a top dominant half and a bottom submissive half with the corresponding categories varying along the acceptance- rejection dimension. The relative positioning of children on this two-dimensional space was assessed to counterbalance the strength of each of the eight class play roles in relation to one another. A total acceptance score was derived by subtracting the percentage of nominations received on the rejecting half of the circumplex from the percent received on the 12 Dominant Mean-Bossy Leader (MB) (L) Fi hts .F) Helpful (H) Rejecting *‘Accepting Nice Angry-Complaining (AC) (N) Afraid QuieteShy (A) (05) Submissive Figure 1 Circumplex Model Representing Relationship between Reputation Types as Conceptualized along the Dimensions of Dominance-Submission and Acceptance-Rejection 13 accepting half. Likewise, a total dominance score was derived by sub- tracting the percentage of scores on the submissive half from the per- centage on the dominant half. Comparison of the total acceptance and total dominance scores allowed for locating the region of the circumplex most characteristic of the peer group's perception of the child when all of the reputational categories were considered together. Two additional measures were derived from the class play data. These measures assessed how subjects made their choices for the class play roles. The first of these measures was the sex of the child nomi- nated for each role. A female choice was scored zero and a male choice was scored one. The second measure examined whether or not the subject chose a classmate for each role who was also often chosen by other children. Binomial probability distributions were used to determine that a target needed at least four nominations for there to be a sig- nificant degree of consensus among peers (p < .05). Each child received a score of one or zero for each of the eight reputation categories depending on whether or not he/she chose a child about whom there was consensus. Vignette Measures. Trained raters who were blind to the reputation category of the nominated child scored the vignette predictions. Each response was evaluated for the relative strength of the acceptance- rejection and dominance-submission dimensions implied by the predicted behavior. The response was then matched to the octant of the circumplex which most closely characterized the prediction. The eight resulting prediction categories corresponded to the eight class play roles. Agree- ment between raters was 96%. Examples of responses which were scored in the different octants for the vignette involving a broken record were as 14 follows: 1. leader - decide how to replace the record; 2. helpful - help me clean up the mess; 3. nice - smile and say it was too bad that the record broke; 4. quiet-shy - wouldn't really matter to him; 5. afraid - would hide the broken record; 6. angry-complaining - start yelling and swear; 7.fights - beat me up; and 8. mean-bossy‘— tell me that I was stupid for breaking the record. The frequency with which responses were made in each of the eight prediction categories was compiled for each subject. This measure was used to study variations in the types of behavior characteristically expected. A second measure assessed the degree of consistency between the reputation category of each nominated child and the subject's predicted behavior for this nominee. The eight class play reputational categories and the eight prediction categories were designed to have the identical circumplex ordering. Predicted behavior matching the octant of the reputational category comprised high consistency whereas predictions made in octants more distant on the circumplex indicated less consistency between reputation and predicted behavior. A means of scaling the degree of consistency was devised. The intercorrelations between the class play reputation types represented the empirical relationship between the circumplex octants. The consistency of the prediction for each reputation type was scored using the correlation coefficient between the octant representing the class play reputation and the octant repre- senting the prediction type. Each of the subject's predictions was thus given a separate consistency score. RESULTS Each of the measures of the three tasks was subjected to an analysis of variance with grade, sex and social effectiveness group as the between- subjects factors. The dependent measure consisted of a single measure or a repeated measure depending on the nature of the variable being examined. Multiple comparisons were evaluated with the Scheffe test. All significant contrasts between means reported in the text were sig- nificant at the .05 level or better. Sociometric Mutual Friendships. The analysis of variance for the number of mutual friendships revealed a significant main effect for grade, f(2,96) = 14.08, p,< .001. A significant increase was evidenced across the age range in the number of mutual friendships indicating continued movement toward friendship pairings in which both members considered each other as friends. A significant main effect also resulted for social effectiveness group, §(3,96) = 66.24, p_< .001. Stars were involved in mutual friendship relations significantly more often than each of the other social effectiveness groups. Isolated children and rejected children did not differ in their mutual friendship relations and had significantly fewer mutual friendships as compared to average children. However, as shown in Table 1, a significant grade by effEC‘ tiveness group interaction occurred, {(6,96) = 3.14, p_< .01. Stars maintained a high level of mutual friendship involvement across the age 15 16 Table 1 Mean Number of Mutual Friendships of Social Effectiveness Groups by Grade Social Effectiveness Group 1 Star 2.70 Average 1.10 Isolated .20 Rejected .30 Grade 3 2.70 1.50 .60 2.70 2.10 1.80 17 range. In contrast, rejected children of different ages did not sig- nificantly differ in their minimal mutual friendship involvement. Both average and isolated children, however, significantly increased their participation in mutual relations between first and fifth grade. First grade isolated children were not distinguishable from first grade rejec- ted children, both groups having minimal involvement in mutual friend- ships. However, the number of mutual friendships evidenced by fifth grade isolated children dramatically increased so that they were no longer differentiable from average children but distinctly different from rejected children. The patterns of friendship selection by grade and social effective- ness group as shown in Table 2 indicated that the isolated children did not have different friendship selection patterns at the three grade levels. Instead at the older grade levels, star and average children tended to increase their number of nominations of isolated children as friends, and more of these nominations were mutual friendship choices. Class Play Each of the eight class play items chosen to represent the differ- ent octants of the interpersonal circumplex tended to evidence the strongest positive correlations with its adjacent octants and the most negative correlation with each octant's polar opposite. (See Table 3.) While the relationships between variables did not perfectly confirm“ the circumplex model, the pattern of intercorrelations tended to verify this conceptualization. The average of intercorrelations for all adja- cent octants tended to decrease and become negative as more distant octants were considered. The average correlations of adjacent octants proceeding to polar opposite octants were as follows: .25, .08. -.10, and 18 .zpe>vueeemeg .eeeem spew» use .ecwgp .umEFw gee om.mm .mw may mezegm eeaeenee so .eeue—em_ .geum ea eecmpmme we: cogeppse ppe we mumpmeee eaeem eaegm>wueeepe some cpsupz mcepeecwee: ye xuceeeeew euee_ecp me_gu=e epeep ammmm HooHH HNV N HNHH HONH e HHNH Hmme eH Hoe HONH e m Home HNHH e Hoe HNHH e ANNH HNmH NH Hoe HNHV m N Hoe HNH H Hee HNHH e HmHV Hmee NH Hme Hose NH H eaHeeHmm HeeHv Ame H HNeV HeHH m Hmev HNmH NH Heev Home m m - Hoe o HNeH HoHH m Hmme Home NH HHHH Home a m Hoe Ame H Hoe HNH N Hoe HHmH NH HoHe Hmme oH H eeHaHoWH Home Hoe m Heme HeHH NH HeNH Hose He HNeH Heme Hm m HHNH Hey S Home HNH N Hmee Has. Ne HNee Heme mm m HooHH Hmv e Home HNH N Hmee Hmee Ne Hmee Hmee an H a oma1a>< - Hoe o HeeHH HNHH e HHN. Hmme eH HeoHH HNNH oH m - Hoe o HooHH HoHH m Heme Hmee NH . HeoHH HNNV N m HeeHe HNH H Home HNV N Home HoNH HN HeeHv HONH e . H teem eeueenea team—emu eemeee>< seam eeeew geese mmeee>Pueewwm Leuec_5ez geese mme=e>weeewem weepsez eeegw an ece geese mmeco>vaeeewm >5 :eceppgu eeueemmm eee .eoaepemH .emece>< .geum Lew eepue:_5ez ewgmecevsm we mcgeauee N «Pee» 19 oo.H m2 arrow. oo.H arrfic. «ream. oo.~ u< eo.u «may: «am~.- NH.- mo. mo.i «may: rm~.. rmH.u cc. mo.r soar; «r-.s aro~.u Ho. oo.H ram“. mo.u no.1 arom.i oo.~ ramfi. rmH. co. oo.H «ramm. «arem. cc.H «Wren. < me 2 z eceeesgum xepessegpu mcwues—xegee< meezp cepueaaeem awesome mcewue—eeceeeeueH m mpneh oo.~ Hoe. v.mxr« mo. v.mra mo. v.mx Amzv ammemlceez Hue ne‘eHL Au