قربانی شدن با توجه به زورگویی و تجاوز فیزیکی در مردان و زنان ساموآیی

Abstract In recent years, bullying has come into focus as a critically important social issue that demands empirical understanding to inform best practice regarding both intervention and prevention. In Western cultures, low physical aggression in boys, but high physical aggression in girls, predicts elevated victimization due to bullying, and we predicted that the same would be true cross-culturally. The present study sought to understand the role that physical aggression plays in victimization in Samoa, provide a prevalence estimate of the rate of bullying in the island nation, as well as validate the Forms of Bullying Scale (FBS; Shaw, Dooley, Cross, Zubrick, & Waters, 2013) in a cross-cultural context. In a sample of adult Samoan men and women (n = 214), men reported elevated rates of verbal, physical, and overall rates of victimization due to bullying in childhood compared to women, but no sex differences emerged in levels of physical aggression. Additionally, the FBS showed appreciable reliability, as well as a latent factor structure consistent with the findings of the scale's authors. Prevalence of victimization due to bullying in Samoa is comparable to that reported by other authors conducting cross-cultural research on this topic.

1. Introduction Bullying (i.e., repeated attempts by a group or individual to gain social advantage by the use of physical, verbal, or relational aggression against a target; Crick and Dodge, 1999 and Espelage and Swearer, 2003) has come to the forefront in recent years as a highly important social issue (Arseneault et al., 2010, Gini and Pozzoli, 2009 and Hawker and Boulton, 2000). Research indicates that bullying has both immediate and long-term negative impacts on physical and mental health (e.g. Copeland et al., 2013, Copeland et al., 2014, Fekkes et al., 2006, Gini and Pozzoli, 2009 and Hawker and Boulton, 2000). This has led the World Health Organization (WHO) to declare bullying to be a “major public health problem” (p.403) that necessitates immediate and widespread policy regarding prevention and intervention (Srabstein & Leventhal, 2010). Most bullying research has been conducted using WEIRD samples (i.e., those that are White, Educated, Industrialized, Rich, and Democratic; Henrich, Heine, & Norenzayan, 2010) even though the prevalence and incidence of bullying is known to differ across a variety of cultural contexts ( Craig et al., 2009, Due and Holstein, 2008, Due et al., 2009, Flemming and Jacobsen, 2010 and United Nations Children's Fund, 2014). Despite this cross-cultural variation, bullying behavior seems to be a relatively ubiquitous feature of human development ( Due & Holstein, 2008), and some argue a logical manifestation of childhood aggression aimed at hierarchy formation and maintenance ( Cillessen and Mayeaux, 2004 and Pellegrini and Bartini, 2000). Further cross-cultural research could help to elucidate the common unifying elements of bullying that are cross-culturally invariant. Because bullying is often characterized as one subset of aggressive behavior (e.g. Craig et al., 2009, Crick and Dodge, 1999 and Gini and Pozzoli, 2009), it is critical to understand the relationship that bullying shares with aggression more broadly. Indeed, some definitions of aggression (e.g. “any action undertaken with the apparent intent of causing physical or psychological harm” Burbank, 1987: 72) could easily function as operational definitions for bullying as well. In studies conducted on participants ranging from young children to middle-aged adults, it is widely reported that males tend to be more aggressive than females (e.g. Archer, 2004, Archer, 2009, Hyde, 1990 and Maccoby and Jacklin, 1980). This finding must, however, be evaluated in light of evidence that men and women tend to differ in the quality of their aggression, but not so much in quantity (Archer and Coyne, 2005 and Björkqvist, 1994). While males typically engage in more blatant and direct forms of aggression (Archer, 2009 and Craig et al., 2009) women exhibit styles that are more subtle and covert (Björkqvist et al., 1992, Crick and Grotpeter, 1995 and Salmivalli et al., 2000). Additionally, some cultural milieus seem to foster more uniform levels and forms of aggressive behavior in both men and women (Archer, 2004, Maccoby and Jacklin, 1980 and Whiting and Edwards, 1973). Indeed, while men tend to be more physically aggressive cross-culturally, sex differences can be variable for verbal aggression, and either nonexistent or reversed when considering relational forms of aggression (see especially Archer, 2004). Sex differences in styles of aggression are echoed in much of the bullying literature. Boys suffer the ill effects of physical bullying more often, whereas girls tend to be victimized in less obvious, but equally damaging social ways (Craig et al., 2009, Prinstein et al., 2001 and Wang et al., 2012). It has also been demonstrated that boys who are unlikely to use physical aggression tend to be especially likely targets of bullies (Craig, 1998, Smith et al., 2004 and Young and Sweeting, 2004). The opposite is true of girls, where a tendency to employ physical aggression (among other gender-atypical traits) is associated with elevated victimization (Young & Sweeting, 2004). This may be reflective of the broader social context in which bullies operate, namely that gender-atypical behavioral expressions (i.e. low physical aggression in boys or high physical aggression in girls) provide salient cues, which bullies use to target victims. Although numerous measurement instruments have been used by bullying researchers, there is little consensus on which one is best, and even less certainty regarding their respective validities and psychometric properties (Cornell and Bandyopadhyay, 2010 and Felix et al., 2011). In response to this, a group of Australian researchers (Shaw, Dooley, Cross, Zubrick, & Waters, 2013) constructed and validated a multi-item measure of bullying victimization and perpetration in adolescence, the Forms of Bullying Scale (FBS), which drew extensively from the work of pioneers in the field (e.g. Olweus, 1996 and Rigby, 1998). The current study sought to utilize the FBS in a sample of men and women from Samoa in order to assess its cross-cultural validity and provide a prevalence estimate of victimization due to bullying in this country. UNICEF released information regarding the prevalence of bullying throughout the developing world (UNICEF, 2014), including Samoa, where 74% of youth aged 13–15 reported having experienced bullying in the previous 12 months. Although this figure suggests that bullying is a salient social issue in Samoa, the measures employed did not fully capture the types or severity of victimization that researchers gain when using multi-item inventories such as the FBS. Additionally, we sought to understand the connections between physical aggression and bullying in the Samoan context. Specifically, we anticipate that men and women will differ in their reported levels of physical aggression, and that low physical aggression in men, but high physical aggression in women, will significantly predict reported victimization due to bullying in childhood.

3. Results Biographic variables were compared between men and women using independent sample t-tests. Men and women did not differ in their age (p = .85), education level (p = .45), or income (p = .22). Age did not correlate with adult levels of physical aggression in either men or women (both p > .50). Means (±SD) were calculated separately for men and women for both childhood and adult levels of physical aggression, the five types of bullying recommended by Shaw et al. (2013), as well as Overt/Direct (questions 1, 4, 5, 6, and 8) and Covert/Indirect (questions 2, 3, 7, 9, and 10) bullying (see Crick and Bigbee, 1998, Espelage and Swearer, 2003 and Prinstein et al., 2001). These values were then compared using independent sample t-tests, which indicated that men did not differ from women in levels of physical aggression, but were more likely to report significantly higher levels of childhood verbal, physical, overt and overall victimization due to bullying than were women ( Table 1). Table 1. Comparisons of various forms of bullying between men and women. Men Women M SD M SD t df p Cohen's d Childhood AQ 19.06 6.43 18.49 7.13 0.62 212 .538 – Adult AQ 19.17 6.26 18.22 7.04 1.05 212 .297 – Type of victimization: Verbal 3.86 1.93 3.08 1.41 2.62 199.4a .009⁎ .46 Threatened 4.12 2.06 3.64 1.84 1.77 212 .078 – Physical 3.95 1.93 3.30 1.71 2.63 212 .009⁎ .36 Relational 4.27 2.36 3.92 2.01 1.17 209.7b .246 – Social 4.56 2.17 4.35 2.15 0.74 212 .463 – Overt/direct 9.74 3.88 8.24 3.51 2.95 212 .004⁎ .41 Covert/indirect 10.85 4.57 10.05 4.06 1.36 212 .174 – Overall bullying 20.59 7.63 18.29 6.66 2.35 212 .020⁎ .32 a Df adjusted based on Levene's test for equality of variances: F = 10.33, p = .002. b Df adjusted based on Levene's test for equality of variances: F = 4.81, p = .03. ⁎ p < .05. Table options In line with previous research (e.g. Skrzypiec, Slee, Murray-Harvey, & Pereira, 2011), individuals were parsed into groups of low, medium, and high victimization on the FSB-V (i.e., low victimization = 10–19 points; medium = 20–29 points, high ≥ 30 points) in order to calculate a prevalence estimate for childhood victimization due to bullying in Samoa. Overall, 57.5% (68 women, 55 men) of individuals reported low, 29.9% (24 women, 40 men) medium, and 12.6% (12 women, 15 men) high victimization as children. The distribution of victimization intensity between these groups did not differ significantly by gender (χ2(2, n = 214) = 5.54, p = .063). Because different types of victimization do not necessarily occur independently, intercorrelations among the various styles of bullying, as well as childhood physical aggression, were calculated. Separate correlations for men and women are displayed in Table 2. The high intercorrelation among victimization subtypes indicates that individuals targeted in one way also tended to be victimized in others. Correlations were similar for both men and women indicating that victimization subtypes covaried in both sexes, consistent with the findings of Shaw et al. (2013). Table 2. Intercorrelations among childhood physical aggression and styles of victimization. 1 2 3 4 5 6 7 8 9 Adolescent AQ .69 .35 .37 .37 .25 .37 .45 .39 .46 Verbal .23 .33 .41 .38 .50 .41 .81 .54 .73 Threats .35 .62 .33 .39 .38 .39 .66 .61 .70 Physical .40 .44 .64 .52 .42 .35 .77 .49 .68 Relational .33 .38 .48 .40 .56 .49 .54 .85 .78 Social .45 .37 .37 .17ns .39 .47 .49 .81 .73 Overt .39 .78 .82 .84 .49 .32 .67 .63 .89 Subtle/relational .48 .52 .65 .42 .79 .81 .55 .68 .92 Total bullying .50 .73 .83 .70 .74 .66 .86 .90 .79 Note: Correlations for men appear above the diagonal, and women below. All correlations significant at p < .01. The diagonal displays reliability estimates (α) of each subdivision of the scale. Table options Factor analysis was performed in order to elucidate the underlying factor structure of the FBS-V in this sample. Because Shaw et al. (2013) indicated a single underlying factor in the FBS-V, the same structure was imposed on this data, rather than relying on eigenvalues > 1, or examination of the scree plot. 1 Maximum likelihood extraction of a single factor solution accounted for 27.8% of item variance (all loadings > .40; see Table 3). Consistent with Shaw et al. (2013), these results indicate that a single underlying factor, namely, general victimization due to bullying, is appropriate for this scale. Table 3. Principal components analysis of the Forms of Bullying Scale-Victimization. Item Factor 1 1 I was teased in nasty ways .43 2 Secrets were told about me to others to hurt me .49 3 I was hurt by someone trying to break up a friendship .50 4 I was made to feel afraid by what someone said he/she would do to me .53 5 I was deliberately hurt physically by someone and/or by a group ganging up on me .50 6 I was called names in nasty ways .60 7 Someone told me he/she wouldn't like me unless I did what he/she said .52 8 My things were deliberately damaged, destroyed or stolen .53 9 Others tried to hurt me by leaving me out of a group or not talking to me .68 10 Lies were told and/or false rumors spread about me by someone, to make my friends or others not like me .44 Table options Adolescent levels of physical aggression were found to strongly correlate (r(214) = .62, p < .001) with adult physical aggression in both men and women, in line with the longitudinal stability for intra-individual aggression described by Olweus (1979). Regression analysis was performed to understand the relative contribution of biographic variables, sex, and level of physical aggression on victimization due to bullying in Samoa. Given the fact that the correlations between physical aggression in childhood and overall victimization due to bullying were equivalent across the sexes (see Table 2), regression analysis was performed with both groups combined. All variables were simultaneously entered into the model, which significantly predicted (F(5, 207) = 15.54, p < .001) general victimization due to bullying. Physical aggression, education, and gender were the only significant predictors, and accounted for 25.5% of the variance ( Table 4). Table 4. Regression predicting overall victimization due to bullying. Regression coefficients (Β) Standardized regression coefficients (β) Squared semi-partial correlations (sri2) Childhood AQ 0.49** .46 .222 Education − 2.39* − .15 .029 Gender 2.04* .14 .026 Income − 0.22ns – – Age − 0.02ns – – Constant 15.76** R = .52**, adjusted R2 = .26; *p < .05, **p < .001. Note: Gender is coded as 0 = female, 1 = male.