Mother-Child Attachment and Peer Social Competence in Unacquainted and Acquainted Peers

The impact of mother-child attachment on the development of peer interactions of toddlers was examined longitudinally. To determine the level of attachment security, the mothers completed the Q-set assessment. The toddlers ranged in age from 2 years 3 months to 3 years 7 months and were initially unacquainted with each other. The toddlers were observed in the classroom setting at the beginning of the study and then were given 12 weeks to interact and become acquainted with each other. At the end of the 12 weeks, the toddlers were observed once again. Although attachment patterns did not predict the social interaction process of getting acquainted with peers, it did predict the development of the emotional regulation process when getting acquainted with peers. Secure attachment is related to less under regulation even when children do not know each other, and as they become acquainted, the more secure children are, the more likely they are to increase their emotion regulation. Gender differences were also analyzed. Attachment patterns were predictive of peer interactions in girls but not boys. When girls were first introduced to their peers, secure girls initiated more interactions. After girls were acquainted with their peers, secure girls were less under regulated and antisocial.

Introduction

Infants readily form an attachment to their primary caregiver, and according to Bowlby (1973), attachment bonds are not all the same. They vary in quality depending on the caregiver's responsiveness and behavior. The highest quality relationship is called a secure relationship and is formed when a caregiver is responsive and available. When a mother is responsive and sensitive to the needs of her child, the child is able to rely and trust in the caregiver for fulfilling needs. In contrast, when a child is neglected or rejected by mothers, an insecure attachment is formed. The child is not able to trust in her/his most fundamental relationship and as a consequence, may minimize contact with the caregiver (Cassidy, 1994). Based on Bowlby's internal working model theory, children internalize these affect and behavior patterns that they learn within the attachment relationship and apply them to other relationships outside the family. Therefore, a child who finds the relationship with his/her mother satisfying will feel other relationships are worthwhile, and will become more sociable. Research has found this to be true. Lieberman (1977) found that security in 3-year olds was associated with peer reciprocity and more sharing, effective requests, and social initiations. Thus, the attachment relationship is important in predicting the child's peer relationships.

The simple theory of internalizing and recreating the mother-child relationship outside the family does not completely explain the direct connection between attachment and peer relationships. Kerns et al (2000), argued that it is through the emotional regulation process that the primary attachment relationship can been thought of as a precursor to peer relationships and thus predict peer relationship outcomes. Emotional regulation is defined as including "both extrinsic and intrinsic processes responsible for monitoring, evaluating, and modifying emotional reactions, especially . . . intensive and temporal features to accomplish one's goals" (Kerns, Contreras, & Neal-Barnett, 2000). Emotional regulation is important in navigating through everyday life, whether it is a conflict or other social interaction. Children learn emotional regulation primarily through their parents. Researchers have found that parents mold their child's regulatory style early in infancy (Kerns et al, 2000). Children that are able to successfully regulate their emotions are better prepared at handling conflicts and other social interactions. Studies have found that securely attached children are less antisocial and more prosocial. Lieberman (1977) found that security was associated with less non-social behaviors and an absence of negative behavior. Sroufe (1983) found that 3-year-old children with secure attachment histories ranked higher in social competence and positive affect, and ranked lower in negative affect. Lastly, Waters, Wippman, and Sroufe (1979) found that securely attached children had significantly higher peer competence scores. Thus, children with parents who are responsive and sensitive will form a secure attachment in which healthy emotional regulation can be learned, and, in turn, children will form positive interactions with peers.

Bowlby also described a secure relationship as one with a secure base, built as a platform, in which a child is willing to explore. Research showed that attachment patterns have predicted the child's ability and capacity to explore unfamiliar environments (Ainsworth et al., 1978). Children who are securely attached to their mothers form a secure base from which they are able to comfortably explore an unfamiliar environment. So, it might be expected that a child differing in attachment security might also be expected to differ in their response to unfamiliar peers in an unfamiliar environment. When children interact in familiar groups with the same gender, they may elicit different behaviors then if they were in an unfamiliar group with both genders. Previous research showed that same sex dyads would optimally promote social interaction (McGrew, 1972). However, in a natural environment or in a classroom setting, there are typically interactions of both genders. So, one might expect that social interactions might differ if both genders are involved. The presence of both genders as it relates to attachment and peer relations has not been widely researched.

The present study was conducted to examine the association between attachment and toddler peer competence when observing the toddler interact with unacquainted peers and then with the same peers when they become more familiar with them. By looking at how children interact at these two different time periods, one might be able to see if attachment also plays a role in the process of developing and forming social interactions. In addition, gender difference was also a subject of interest in this study. Since both genders were incorporated in each interaction, this study will try to determine if the relationship between attachment security and peer interactions differ from boys and girls.

Hypotheses.The first four hypotheses were related to attachment security to certain peer interaction constructs that have been known to correlate with attachment. The researcher predicted that there was a direct relationship between secure attachment and the prosocial behavior score. A more secure child will have a higher prosocial behavior score and a less secure child will have a lower prosocial behavior score. Also predicted was a direct relationship between secure attachment and overall social bid score. The higher a child's security score the higher their overall social bid score, and the less secure the child the lower the overall social bid score. Both of these hypotheses were based on the results of previous research (Lieberman, 1977) that found that secure attachment was associated with more effective requests and initiations as well as less non-social behavior. The researcher predicted an inverse relationship between secure attachment and under regulation. The more securely attached a child is the less they will be under regulated and more likely to be capable of regulating their emotions. The less securely attached a child is the more he/she will exhibit under regulating and incapable of emotional regulation. Also predicted was an inverse relationship between secure attachment and antisocial behavior. The higher a child's security score the lower their antisocial behavior score will be and the lower the child's security score the higher their antisocial behavior score will be.

Predictions were also based on the notion that secure children are able to explore unfamiliar environments at a higher capacity (Ainsworth, 1978). In the initial peer interactions, it was predicted that the more secure a child is, the more likely the child will initiate interactions (higher social bid score). And it was predicted that the higher a child's security score, the more likely he/she will be to improve peer interactions as the child became more familiar with peers. Thus, as the children became more familiar, it was predicted that the more secure a child is, the more likely he/she will become more prosocial, less antisocial and under regulated, and exhibit more social bids.

Lastly, it was predicted that secure girls, compared to boys, would show higher scores in prosocial behavior and lower scores in antisocial behavior and under regulation. These predictions were based on previous research that showed that aggression tended to increase with age, from two years through four years, and boys were found to be significantly more aggressive than girls (Walter, Pearce, & Dahms, 1957). A further prediction was that girls would have a higher overall social bid score than boys because research has shown that girls produce significantly more words than boys (Bouchard et al, 2009).

Method

Participants

Participants were composed of thirty-one male and twenty-one female toddlers, who were on an early childhood classroom waiting list in the Austin area. Their ages ranged from 2 years 3 months to 3 years 7 months, and the mean age was 3 years.

The toddlers were part of a longitudinal study assessing the effects of a positive guidance program (Saunders, 2009). Thus, the sample also included the toddlers' mothers. Their age range was 26 to 43 years, and the mean age was 34 years. The samples' ethnic distribution was made up of 75 percent Caucasian, 11.5 percent Latino, 11.5 percent Asian, and 2 percent African American. The mothers' education levels also varied, with most finishing college (57.7%), some attending graduate school (32.7%), and fewer only going to some school after high school (9.6%). The distribution of family income of the sample was: $0-20,000 (5.8%), $20,001-40,000 (9.6%), $40,001-60,000 (11.5%), $60,001-80,000 (25.0%), and >$80,000 (48.1%) (Saunders, 2009).

Initially, the sample consisted of 52 mother-child dyads, but after accounting for participant attrition, there were 49 mother-child dyads that completed the entire program, including the pre- and post-measure packets and observations. There were two mothers that had twin boys, however, since dyad measures were taken, only one of the boys, chosen at random, was used in the final analysis.

Procedure

The mothers were recruited from a daycare program waitlist to participate in a positive guidance parenting education 12-week program. The program consisted of teaching parents positive guidance through a seminar only or seminar and hands-on instruction. The program did not significantly affect peer interactions, therefore, this part of the procedure is not relevant in this particular study.

In the beginning of the program, the children were split up into eight interaction groups with five to seven children in each. The interaction groups were evenly formed based on gender and according to the mothers' program participation (e.g., seminar only or seminar and hands-on) to avoid classroom effects. Because there were 31 boys and 21 girls, there were three groups that had a significantly uneven ratio of girls to boys. These three groups had ratios of 2:7, 2:7, and 1:5. The other groups had even or roughly even number of girls to boys. One group did have one more girl than the number of boys. After gender was accounted for and the groups were formed, they were placed into separate classrooms. The children attended class two days a week for three hours each day. Peer interactions took place 6 hours a week for a total of 12 weeks. There were four different classrooms, so not all eight groups met at the exact same time and day.

Observation of Peer Interaction. The first day, the children were unfamiliar with each other and were placed in the classrooms with their respective groups. The children were also new to the classrooms, and had not been inside them prior to the study. The classroom was carpeted with posters on the walls and sectioned into different activity areas. There were a few tables and chairs and many developmentally appropriate toys. During this first period of interaction, they were videotaped for 25 minutes. Then the children continued to interact and attend class 6 hours a week for 12 weeks. The toys in the classroom were randomly switched out and into the classroom. This method was used to provide more opportunity for children to get involved in interactions with their peers due to the curiosity and demand of new toys. The purpose was to elicit the children's social skills with peers and sharing, as well as to initiate conflict and show their lack of social skills. By switching out toys or limiting the amount of people that could play with the toys (e.g., a small tent) children were given the opportunity to interact, get into conflicts, share and take turns, and show aggression. At the end of the 12 weeks, the children were videotaped again for 25 minutes. All observational videos were collected and coded to determine the child's social behaviors and competence level during peer interactions before they were familiar with their peers and after they were familiar with their peers. The coders for the videos were undergraduate research assistants, who were unaware of the child's security score, and achieved an excellent inter-rater reliability (r= .94). They also attended weekly reliability meetings.

Attachment Q-set.The mothers were given the Q-set assessment for their children. They were handed 90 different cards that each had a description of a child, and they had to sort the cards into piles that range from most characteristic of their child to least characteristic of their child. For example one card stated, "Child sometimes signals mother (or gives the impression) that he wants to be put down, and then fusses or wants to be picked right back up" (Waters and Deane, 1987). If the mother believes this statement is exactly how her child behaves she would put it in the first pile. In order to designate 90 cards from most descriptive to least descriptive, several steps had to be taken. First, the mother organized the cards into three piles. Then the mother organized each of the three piles into nine piles. The nine piles were only allowed a certain number of cards so that the piles had a unimodal distribution (e.g., 5, 8, 12, 16, 18, 16, 12, 8, and 5) (Waters and Deane, 1985). Then the mothers handed back the cards in their particular order. Depending on where each card was placed in order by the mother, the child received a security score. The scores were then correlated to a security criterion. This criterion is a score of the most perfectly secure child and it was created based on previous research. The criterion incorporated all the characteristics which had been found to correlate with security (Waters and Deane, 1985).

Measures

Assessing attachment security. The Attachment Q-set (Waters & Deane, 1987) was used to determine the level of security in attachment between each mother-child relationship. The Q-set was composed of 90 descriptive items that were placed into categories based on what was most characteristic of the child to which was least characteristic of the child. Each descriptive item distinguished individual differences by using three domains: personality, attitudes, and behavior. Based on the order in which the descriptions were designated to the child, Waters and Deane (1987) showed validity and reliability in determining the child's security level of attachment to his/her mother. In this study, the mothers completed the Q-set assessment for her child. This data were then used to achieve a security attachment composite score for each child, after correlating each score to a criterion. This criterion was based on what researchers who were experts in attachment theory had found to be characteristics of the most secure child (Waters & Deane, 1985).

Observational coding.The toddler peer interactions were rated using a 7-point scale. The rating of each child was based on general types of social interactions (e.g., prosocial or antisocial behaviors, appropriate social bids) and on their emotional orientation (e.g., empathetic behavior, emotional regulation). The coders were given a list of constructs that were clearly defined and also explained. For example, "appropriate social bid" was defined as "a positive or friendly attempt to interact with another child either verbally or nonverbally" (Heaton, 2010). After the construct was defined, the guide explained the difference between a strong and a weak interaction. For example, the guide stated that "strong appropriate bids include verbally inviting a child to play, offering a toy, or otherwise including another child in ongoing activity, whereas a weak appropriate bid would be simply joining other children's play in a non-disruptive way" (Heaton, 2010). Once the coders were familiar with the constructs, the 7-point scale was used in order to determine a number value for the degree in which the interaction and behaviors were exhibited. A score of 1 meant that the behavior was rarely or never demonstrated, whereas, a score of 7 meant the behavior was demonstrated frequently. Also, depending on the behavior, 1 could mean the child performed the interaction inappropriately or lacked social skills, and a 7 could mean the interaction was positive and friendly. In this study, the researcher used the codes for overall social bids (e.g., how the child initiated an interaction), prosocial behavior (e.g., sharing), antisocial behavior (e.g., not cooperating), and under regulation (e.g., crying, screaming, unable to control their emotions). The peer interaction ratings manual is located in Appendix B (Heaton, 2010).

Results

Relating Attachment and Pre/Post Interactions.

The first four hypotheses were related to predicting the relationship of peer interactions based on secure attachment. With all four hypotheses to determine if the prediction held true, two different correlation techniques were use: Pearson correlations for the initial interaction (pre) and the partial correlations for the second interactions (post). The researcher examined the quality of the mother-child attachment security related to the quality of the interaction between unacquainted peers using Pearson correlations (correlating the security score with each pretest peer interaction score). Next, the researcher examined the relationship between security in attachment and the quality of social interaction once they got to know the peers, after controlling for the child's behavior in the initial encounter. In determining this relationship, Partial correlations were used by correlating the security score with each posttest peer interaction score, after controlling for the pretest peer interaction score. A standard alpha level of .05 was adopted. With regard to the prediction that there would be a direct correlation between attachment and social bids, there was not a significant difference, and therefore, security attachment scores were not significantly related to the amount of overall social bids a child initiated. The second prediction was that there would be a direct relationship between attachment security score and prosocial behavior. Again, the results were not significant and therefore attachment security was not directly related to prosocial behavior. Thirdly, it was predicted that security attachment would be inversely related to under regulation. The pretest was marginally significant, r (37) = -0.211, p = .095, and the posttest was significant, r (37) = -0.277, p = .044. Thus, secure attachment is related to less under-regulation even when children do not know each other; as they become acquainted, the more secure children are , the more likely they are to increase their emotion regulation. Lastly, it was predicted that security attachment was inversely related to antisocial behavior. The results did not show a significant difference so security scores were not inversely related to antisocial behavior.

Gender differences

To examine gender differences the same two tests were used as before: Pearson correlations and partial correlations. However, when running each test the researcher filtered out the boys when looking at the girls' results and filtered out the girls when looking at the boys' results. It was predicted that the there would be a gender difference in that the girls would show higher scores in prosocial behavior and lower scores in antisocial behavior and under regulation compared to the boys. The results showed no significant difference in the relation of attachment security and peer interaction, either when the children were unacquainted or after they become acquainted, for boys. However, there were significant findings for the girls. The more securely attached the girls were. the more social bids they tended to initiate in the first interaction (0.384, p<.10). After the girls were familiar with the peers, the more securely attached the girl was, the less under regulated (-0.411, p<.10) and the less antisocial she became (-0.500, p<.05). These results show that there are gender differences in the relation between peer interactions and attachment, and this is particularly true after children become acquainted. The security level of boys did not significantly predict a difference in the amount of peer interactions. In contrast, the security level of girls significantly predicted that, initially, the more secure the girl was the more social bids she exhibited, and once she knew her peers in the post test, the more secure she was the less under regulated and antisocial she became.

Gender Differences when Correlating Security Score with Social Bids, Under Regulation, Prosocial Behavior, and Antisocial Behavior

Discussion

In analyzing the data, little significant differences were found in relating attachment to peer interactions. This could be attributed to the constructs (e.g., sharing, conflict management) that were used to determine social competence. These constructs might have caused such a low outcome of significance in the results if the children were too young to exhibit these particular characteristics or skills. Because children develop social skills at different ages, these children, who were 2-3 years old, might not have developed the skills for which they were coded. For example, one act of prosocial behavior was sharing. Sharing was defined strictly so that the coder had to know that the child intended to give a toy to the other child in the sense of "sharing" (letting the other child have a turn). However, if the child placed a toy beside her and then allowed another child to grab it, this was not coded as prosocial behavior. During most of the observational data, the children participated in parallel play (playing next to each other with the same materials, but not involved in joint play) and most actions were playing with toys, putting down toys and grabbing them. According to Lieberman (1977), this is seen as a form of peer interaction typical of a three-year-old's social capacities and development. However, in this present study, this type of interaction was not coded as prosocial, since the intent to share had to be shown. This usually required the prosocial child to verbalize, (e.g., "Okay, you can have a turn"), or to respond positively to another child's verbal request (e.g., "I want a turn". Children this age seldom use words in the exchange of toys, so prosocial intent was rarely clear. That is, if the child had gotten angry because another child grabbed the toy, then it would have been coded as antisocial, but if the child does not protest to the toy being taken it would not have be coded as positive. Therefore, it could be that findings were due to the fact that the prosocial behavior measure did not account for the ages and social capabilities of the children.

Another factor that supports the theory that the children were too young for the constructs is the result showing that there was a significant finding in the correlation between security and under regulation. Under regulation is when a child cannot control his/her emotions. Emotional regulation is developed in younger children even before they become involved with peers. Learning to regulate emotions is an infant skill that is developed in the context of the parent-child relationship. A child that has a secure attachment with mother is more trusting and confident, and will be more influenced and learn to regulate emotions better than less secure children. Children learn to regulate their emotions through their parent's behaviors and actions, and researchers have argued that parents begin molding their child's regulatory process as early as infancy (Kerns, 2002). As predicted, secure children were able to significantly show less under regulation in the pre- and posttests. In conclusion, data did find that secure children exhibited less under regulation (infant skill), but did not find any other results, supporting the theory that the children might have been too young to elicit other social skills that were coded in the videos.

Another possibility for the outcome in the results is that the children were not observed for enough time or given the opportunity to elicit some of the constructs that were coded. This can be seen in the data. Many children were not given scores for certain constructs, because the child may not have experienced situations in which these skills or traits would have been used. For example, if the child was not given the opportunity to show aggressive behavior, then they did not receive a score. Or, if during the recorded observation there was not a situation in which the child needed to show empathy, then they were not given a score. Therefore, a secure child could have been given no score in the area of empathy or prosocial behavior, and we would not have been able to see if there was a correlation between prosocial behavior and security score. In conclusion, because of the limited time of observation, we would not have been able to find that prosocial behavior does in fact correlate with the mother-child attachment, a finding that previous studies have found.

As for the gender differences, the findings showed that secure girls had significant correlations with certain peer interactions. This is reflective of previous research findings, in which it has been shown that girls acquire language earlier than boys. Bouchard et al (2009) found a linguistic superiority of girls over boys at the age of 30 months. The results found that girls produce significantly more words than boys, with a greater number of grammatical forms and complexity. Because girls are more capable in communicating what they want compared to boys, girls could have been able to address certain issues or conflicts in a more socially appropriate manner than boys. Because communication in the form of speech is universal, a child with speech can understand and be understood with less frustration than a child unable to speak. For example, if two girls capable of speech wanted to play with a toy, they could express that in words. In contrast, a boy incapable of speech might just point and make a noise signaling that he wanted to play with the toy, but if the other child does not understand pointing and noises he will not share, the first boy might get very frustrated and resort to getting the toy by grabbing it. For this reason, secure boys may not by able to elicit each construct deemed as prosocial behaviors, because they do not have the capacity to communicate yet, and this results in frustration and slight aggression. Also, research has shown that at this age, boys were significantly more aggressive than girls (Walter, Pearce & Dahms, 1957). This also supports the theory on why security in boys did not tend to correlate with less antisocial interactions and more emotionally regulated interactions, whereas, security in girl did tend to have significant correlations to positive peer interactions.

A limitation that might have caused a bias in the results is the fact that the observer used for the Q-set was the mother. Waters and Deane (1985) noted that to get the most reliable and valid security scores, an outside observer should be used. They also explained that it is best to use two observers and obtain an inter rater reliability score to ensure reliability and validity of the security scores. However, using mothers as the observer has been proven to correlate with security scores but not as strongly. Because the mother was used in determining the child's security score, this might have biased the security scores. Also the Q-set was used instead of the strange situation (Ainsworth, 1970). The strange situation is an experiment used to distinguish the type of mother-child attachment relationship existed between the dyads. This method has been the gold standard in determining attachment. Therefore, this could also pose a reliability issue in the results.

In conclusion, the present study supported past research findings relating attachment security to peer relationships. It also extended the scope of previous research by incorporating the peer interaction process by looking at the development of social competence before peers were familiar with each other and after they became familiar with each other. Additionally, this study included both genders into the interaction groups and looked at gender differences. By looking at both genders, the present study contributed to research in that it showed secure girls differed significantly in social interactions when placed in the presence of boys, whereas, secure boys did not differ significantly in social interactions when place in the presence of girls. Previous research found no significant gender differences among peer interactions (Lieberman, 1977). Further research could further explain this phenomenon.

In future research to find more of an effect in the way boys and girl interact with unfamiliar peers versus when they become familiar with them, the importance of the process in developing social skills and the importance of the child's development capabilities should be taken into account. To achieve a better understanding of the process of developing social skills, more observation periods or more observational time would be beneficial. If weekly observations were coded, one could learn the process of how a child develops social competence as it relates to secure attachment. Because this study was conducted on toddlers, a longitudinal study of more than one year would also show the child's process of gaining social competence. To achieve a better understanding of the relationship of attachment and peer relationship using the present study, constructs that better consider child's development level and skills would need to be considered to enhance the results. Future research could evaluate social competence by defining constructs more loosely to assign behaviors and skills in more developmentally appropriate definitions.

References

Ainsworth, M, D, S, The development of infant- mother attachment. In B. M. Caldwell & H. N. Ricciuti (Eds.), Review of child development research. Vol, 3. Chicago: University of Chicago Press,1973.

Ainsworth, M., & Bell, S. (1970). Attachment, Exploration, and Separation: Illustrated by the Behavior of One-Year0Olds in a Strange Situation. Child Development, 41(1), 49-67. doi: 10.1111/1467-8624.

Ainsworth, M., Blehar, M., Waters, E., & Wall, S. (1978). Patterns of attachment: A psychological study of the strange situation. Oxford England: Lawrence Erlbaum. Retrieved from PsycINFO database.

Ainsworth, M., & Marvin, R. (1995). On the Shaping of Attachment Theory and Research: An Interview with Mary D. S. Ainsworth (Fall 1994). Monographs of the Society for Research in Child Development, 60(2-3), 3-21.

Bouchard, C., Trudeau, N., Sutton, A., Boudreault, M., & Deneault, J. (2009). Gender Differences in Language Development in French Canadian Children between 8 and 30 Months of Age. Applied Psycholinguistics, 30(4), 685-707. Retrieved from ERIC database.

Erickson, M., Sroufe, L., & Egeland, B. (1985). The relationship between quality of attachment and behavior problems in preschool in a high-risk sample. Monographs of the Society for Research in Child Development, 50(1-2), 147-166. doi: 10.2307/3333831.

Heaton, C. (2010). Maternal Empathy and Changes in Mothers' Permissiveness and Depression as Predictors of Toddlers' Early Social Competence with Peers: A Parenting Intervention Study. Unpublished 2nd Year Graduate Project. The University of Texas at Austin, Austin, TX.

Kerns, K.A., Contreras, J.M., & Neal-Barnett, A.M. (2000). Family and peers. Westport, CT: Praeger Publishers.

Lieberman, A. (1977). Preschoolers' competence with a peer: Relations with attachment and peer experience. Child Development, 48(4), 1277-1287. doi:10.2307/1128485.

McGrew, W. (1972). An ethological study of children's behavior. Oxford England: Academic Press. Retrieved from PsycINFO database.

Saunders, R. C. (2009). Stability and change in parenting attitudes and behaviors regarding discipline: The effectiveness of a hands-on training in positive guidance. Unpublished doctoral dissertation, The University of Texas at Austin, Austin, TX.

Walters, J., Pearce, D., & Dahms, L. (1957). AFFECTIONAL AND AGGRESSIVE BEHAVIOR OF PRESCHOOL CHILDREN. Child Development, 28(1), 15-26. Retrieved from Academic Search Alumni Edition database.

Waters, E. (1987). Attachment Q-set (Version 3). Retrieved (09/14/10) from http://www.johnbowlby.com.

Waters, E. & Deane, K. (1985). Defining and assessing individual differences in attachment relationships: Q-methodology and the organization of behavior in infancy and early childhood. In I. Bretherton & E. Waters (Eds.), Monographs of the Society for Research in Child Development, 50, nos. 1-2. pp. 41-65.

Waters, E., Wippman, J., & Sroufe, L. (1979). Attachment, Positive Affect, and Competence in the Peer Group: Two Studies in Construct Validation. Child Development, 50(3), 821. doi:10.1111/1467-8624.ep7251021.

Mother-Child Attachment and Peer Social Competence
in Unacquainted and Acquainted Peers

Elissa Prado Gonzalez

University of Texas at Austin

Abstract

Introduction

Method

Participants

Procedure

Measures

Results

Relating Attachment and Pre/Post Interactions.

Gender differences

Discussion

References