childrens perceptions of same- and cross- sex peers social behavior and social status noel a. card,...
TRANSCRIPT
Children’s Perceptions of Same- and Cross-Sex Peers’ Social Behavior and Social Status
Noel A. Card, University of Arizona
Acknowledgements:
Todd Little (informal co-author of present paper)
Patricia Hawley & Ernest Hodges (coauthors on IJBD article)
Introduction and rationale
• The importance of studying children’s interpersonal perception– Possible links to social goals, information processing, and
behaviors.
– Allows for understanding of development of later interpersonal perception.
Introduction and rationale• The importance of studying children’s interpersonal perception
• Previous research studying children’s interpersonal perception– Malloy, Sugarman, Montvilo, & Ben-Zeev (1995)
• Relative actor, partner, and relationship variances in different domains across 1st to 6th grades.
– Card, Romero, & Wiseman (2004)• Relative variances in perceptions of aggression and victimization
among middle school children.
– Card, Hodges, Little, & Hawley (2005)• Gender effects in perceptions of various types of aggression and
social status.
Introduction and rationale• The importance of studying children’s interpersonal perception
• Previous research studying children’s interpersonal perception
• Conceptual expectations regarding gender differences and interpersonal perception:– In-group / out-group processes
– ‘Separate worlds’ hypothesis (Maccoby)
Goals of study• Extend upon Card et al. (2005) by testing process model of
gender differences– Proposed mediational model in which gender differences in perceptions
of social behavior account for gender differences in perceptions of social status.
Gender
-Sex of perceiver
-Sex of target
-Within/across sex
Perceptions of
Social Behaviors
-Aggression
-Prosocial behavior
Perceptions of
Social Status
-Social preference
-Perc. popularity
Method• Sample
– N=374 6th graders in 17 classes (M age = 10.5 years)
– Approximately equal number of boys (n=194) and girls (n=180)
– Northeastern US: 68% White, 22% African American, 5% Hispanic
• Measures – only Peer Nomination Inventory relevant– Aggression (8 items from 4 form / function combinations; range 0 – 100):
– Prosocial behavior (2 items; range 0 – 100):
– Social preference (liking – disliking; range = -100 – 0 - 100):
– Perceived popularity (popularity – unpopularity ; range = -100 – 0 - 100)
Results – Group-level
• IV to DV (sex perceived social status)– Repeated-measures ANOVA (nom sex, targ sex, interaction) of mean
levels across 17 groups– Both aspects of social status showed similar sex differences:
• Nominator effects: No significant differences• Target effects: Girls were more liked and perceived as more popular than
were boys.• Interaction: Higher social preference and perceptions popularity within sex
than across
Boys’ nominations of… Girls’ nominations of…Nominator Target N X T
Boys Girls Boys Girls
Social Pref. 18.8 0.6 -1.9 23.5 ns girls within
Perc. Popularity 4.1 2.0 0.1 6.3 ns girls within
Note: Rightmost column show group with higher means when effect is significant (p < .05)
Results – Group-level
• IV to Mediator (sex perceived behaviors)– Repeated-measures ANOVA (nom sex, targ sex, interaction) of mean
levels across 17 groups– Both perceived behaviors showed similar (though not identical) sex
differences:• Nominator effects: Girls perceived higher rates of aggression and prosocial
behavior than did boys.• Target effects: Girls were seen as more prosocial than boys.• Interaction: Higher perceptions of both aggression and prosocial behavior
within sex than across
Boys’ nominations of… Girls’ nominations of…Nominator Target N X T
Boys Girls Boys Girls
Aggression 4.3 1.7 2.6 5.4 girls ns within
Prosocial 7.2 3.7 2.9 15.6 girls girls within
Note: Rightmost column show group with higher means when effect is significant (p < .05)
Results – Group-level
• Mediator to DV (social behaviors social status)– Regressions of group-mean social status onto group-mean behaviors
(across 17 groups)
– Group-levels of social preference not predicted by perceptions of social behaviors.
– Group-levels of perceived popularity positively predicted by both perceived aggression and prosocial behavior
Social Pref/ Perc. Popularity
Aggression ns .44
Prosocial ns .55
Note: Values are standardized regression coefficients of social status (top) onto two behaviors (left).
Results – Group-level
• IV to DV, controlling mediators (sex social behaviors perceived social status)– Two effects met initial criteria for mediation:
• Target effects in perceived popularity (girls > boys) were mediated by target effects in perceptions of prosocial behavior (girls > boys).
• Interaction effects in perceived popularity (within > across) were mediated by interaction effects in perceptions of prosocial and aggressive behaviors (within > across)
Nominator Target N X T
Social Pref. -- -- --
Perc. Popularity -- ns ns
Results – Group-level
• Summary of effects– Only evidence of mediation is that perceptions of prosocial behavior
mediate prediction of perceived popularity from target sex (girls > boys) and sex interaction (within > across).
– Direct effects (not mediated by perceptions of social behaviors) on social preference from target sex (girls > boys) and sex interaction (within > across).
Nominator Sex
Target Sex
N X T (within/across
sex)
Perceived Aggression
Perceived Prosocial
Social Preference
Perceived Popularity
Results – Assimilation (actor variance)
• IV to Mediator (sex perceived behaviors)– Repeated-measures ANOVA (nom sex, targ sex, interaction) of actor
variances across 17 groups.– Both perceived behaviors showed similar sex differences as did social
status measures:• No main (sex of nominator or target ) effects.• Interaction: Greater degree of assimilation in views of cross-sex individuals’
social behaviors.
Boys’ nominations of… Girls’ nominations of…Nominator Target N X T
Boys Girls Boys Girls
Aggression 58 109 122 78 ns ns across
Prosocial 46 422 414 55 ns ns across
Note: Rightmost column show group with higher means when effect is significant (p < .05)
Results – Assimilation (actor variance)
• Mediator to DV (social behaviors social status)– Computed bivariate actor-actor covariances among variables, which were
then used to compute regression coefficients. Regression coefficients were then aggregated across 17 groups.
– Actor variance in social preference associated with actor variance in perceived prosocial behavior.
– Actor variance in perceived popularity associated with actor variance in both perceived aggression and prosocial behavior.
Social Pref/ Perc. Popularity
Aggression ns .13
Prosocial .35 .44
Note: Values are standardized regressions of social status (top) onto two behaviors (left).
Results – Assimilation (actor variance)
• IV to DV, controlling mediators (sex social behaviors perceived social status)– Two effects met initial criteria for mediation:
• Target effects in perceived popularity (girls > boys) were mediated by target effects in perceptions of prosocial behavior (girls > boys).
• Interaction effects in perceived popularity (within > across) were mediated by interaction effects in perceptions of prosocial and aggressive behaviors (within > across)
Nominator Target N X T
Social Pref. -- -- ns
Perc. Popularity -- -- ns
Results – Assimilation (actor variance)• Summary of effects
– Only sex of nominator X target interactions was related to magnitude of assimilation in social behaviors and status (greater assimilation across than within groups).
– Across-group (versus within-group) assimilation in perceptions of social behaviors accounted for (mediated) across-group assimilation in perceived popularity (but not social preference).
Nominator Sex
Target Sex
N X T (within/across
sex)
Perceived Aggression
Perceived Prosocial
Social Preference
Perceived Popularity
Results – Consensus (partner variance)
• IV to DV (sex perceived social status)– Repeated-measures ANOVA (nom sex, targ sex, interaction) of partner
variances across 17 groups.– Both aspects of social status showed similar patterns:
• Girls exhibited higher degrees of consensus than did boys.• Sex of target was not related to degree of consensus• For both measures of social status, there was greater consensus across sex
than within.
Boys’ nominations of… Girls’ nominations of…Nominator Target N X T
Boys Girls Boys Girls
Social Pref. 185 779 1041 206 girls ns across
Perc. Popularity 223 654 1931 216 girls ns across
Note: Rightmost column show group with higher means when effect is significant (p < .05)
Results – Consensus (partner variance)
• IV to Mediator (sex perceived behaviors)– Repeated-measures ANOVA (nom sex, targ sex, interaction) of partner
variances across 17 groups.– Both perceived behaviors showed similar sex differences:
• No significant nominator or target sex differences.• Interaction: Higher consensus of both aggression and prosocial behavior
within sex than across.
Boys’ nominations of… Girls’ nominations of…Nominator Target N X T
Boys Girls Boys Girls
Aggression 148 34 37 277 ns ns within
Prosocial 730 21 132 308 ns ns within
Note: Rightmost column show group with higher means when effect is significant (p < .05)
Results – Consensus (partner variance)
• Mediator to DV (social behaviors social status)– Computed bivariate partner-partner covariances among variables, which
were then used to compute regression coefficients. Regression coefficients were then aggregated across 17 groups.
– Partner variance in social preference associated with partner variance in perceived prosocial behavior.
– Partner variance in perceived popularity associated with partner variance in both perceived aggression and prosocial behavior.
Social Pref/ Perc. Popularity
Aggression ns .33
Prosocial .45 .48
Note: Values are standardized regressions of social status (top) onto two behaviors (left).
Results – Consensus (partner variance)
• IV to DV, controlling mediators (sex social behaviors perceived social status)– Two effects met initial criteria for mediation:
• Interaction effects predicting consensus in social preference and perceived popularity remained significant after controlling for consensus in aggression and prosocial behavior.
• This might be expectable given that sex interaction had opposite direction of effects for consensus in social behavior (within > across) versus social status (across > within)
Nominator Target N X T
Social Pref. -- -- Sig
Perc. Popularity -- -- Sig
Results – Consensus (partner variance)• Summary of effects
– Only sex of nominator X target interactions was related to magnitude of assimilation in social behaviors and status (greater assimilation across than within groups).
– Across-group (versus within-group) assimilation in perceptions of social behaviors accounted for (mediated) across-group assimilation in perceived popularity (but not social preference).
Nominator Sex
Target Sex
N X T (within/across
sex)
Perceived Aggression
Perceived Prosocial
Social Preference
Perceived Popularity
+
++
+
+-
-
Results – Uniqueness (relationship variance)
• IV to DV (sex perceived social status)– Repeated-measures ANOVA (nom sex, targ sex, interaction) of
relationship variances across 17 groups.– Girls exhibited higher degrees of unique perception than did boys in both
social preference and perceived popularity.– For social preference (but not perceived popularity), there was a greater
degree of uniqueness in social preference within than across sexes.
Boys’ nominations of… Girls’ nominations of…Nominator Target N X T
Boys Girls Boys Girls
Social Pref. 1128 517 934 1430 girls ns within
Perc. Popularity 468 372 544 589 girls ns ns
Note: Rightmost column show group with higher means when effect is significant (p < .05)
Results – Uniqueness (relationship variance)
• IV to Mediator (sex perceived behaviors)– Repeated-measures ANOVA (nom sex, targ sex, interaction) of
relationship variances across 17 groups.– Unique perceptions of aggression were not related to sex.– Unique perceptions of prosocial behavior showed more variability when
girls were nominators, when girls were targets, and for perceptions within sex.
Boys’ nominations of… Girls’ nominations of…Nominator Target N X T
Boys Girls Boys Girls
Aggression 88 51 85 115 ns ns ns
Prosocial 317 309 252 618 girls girls within
Note: Rightmost column show group with higher means when effect is significant (p < .05)
Results – Uniqueness (relationship variance)
• Mediator to DV (social behaviors social status)– Computed bivariate intrapersonal relationship covariances among
variables, which were then used to compute regression coefficients. Regression coefficients were then aggregated across 17 groups.
– Relationship effects of perceptions of aggression were uniquely linked to relationship effects in perceived popularity.
– Relationship effects of perceptions of prosocial behaviors were uniquely linked to relationship effects in social preference.
Social Pref/ Perc. Popularity
Aggression ns .12
Prosocial .35 ns
Note: Values are standardized regressions of social status (top) onto two behaviors (left).
Results – Uniqueness (relationship variance)
• IV to DV, controlling mediators (sex social behaviors perceived social status)– Two effects met initial criteria for mediation:
• Both the main effect of nominator sex and the interaction of nominator X target sex on unique variance in social preference were mediated by unique perceptions of prosocial behavior.
Nominator Target N X T
Social Pref. ns -- ns
Perc. Popularity -- -- --
Results – Uniqueness (relationship variance)• Summary of effects
– Although there were several effects, only two pathways met criteria for moderation
• Nominator sex and N X T interaction were mediated by prosocial behavior in predicting unique perceptions of social preference.
• Note that target sex was not initially associated with relationship variance in social preference.
Nominator Sex
Target Sex
N X T (within/across
sex)
Perceived Aggression
Perceived Prosocial
Social Preference
Perceived Popularity