کارآزمایی کنترل شده مداخله خانواده برای کودکان مورد آزار واقع شده توسط همسالان

Abstract This study examined the effects of a family intervention on victimization and emotional distress of children bullied by peers. The intervention, Resilience Triple P, combined facilitative parenting and teaching children social and emotional skills relevant to developing strong peer relationships and addressing problems with peers. Facilitative parenting is parenting that supports the development of children’s peer relationship skills. A randomized controlled trial was conducted with 111 families who reported chronic bullying of children aged 6 to 12 years. Families were randomly allocated to either an immediate start to Resilience Triple P (RTP) or an assessment control (AC) condition. Assessments involving children, parents, teachers, and observational measures were conducted at 0 (pre), 3 (post) and 9 months follow-up. RTP families had significantly greater improvements than AC families on measures of victimization, child distress, child peer and family relationships, including teacher reports of overt victimization (d = 0.56), child internalizing feelings (d = 0.59), depressive symptoms (d = 0.56), child overt aggression towards peers (d = 0.51), acceptance by same sex and opposite sex peers (d = 0.46/ 0.60), and child liking school (d = 0.65). Families in both conditions showed significant improvements on most variables over time including child reports of bullying in the last week reducing to a near zero and indistinguishable from the normative sample. The intervention combining facilitative parenting and social and emotional skills training for children produced better results than the comparison assessment control condition. This study demonstrated that family interventions can reduce victimization and distress and strengthen school efforts to address bullying.

Results Preliminary Analyses To check the effectiveness of randomization, we analyzed between-group differences at Time 1. There was a significant difference on one demographic measure with AC parents tending to be older, F(1, 107) = 11.20, p = .001. There was an initial difference between groups on 2 of the 26 outcome variables. The AC group reported higher scores on Problem of Sibling Agonism, F(1, 107) = 4.27, p = .041, and RTP children were liked less by peers of the opposite sex than were AC children, F(1, 107) = 4.27, p = .041. A missing data analysis (including that due to attrition) revealed 10.59% of total values were missing. The proportion missing in teacher data was higher (17.82%). Little’s test indicated that data points were missing completely at random, χ2 (4642) = 3342.93, p > .999. Families who dropped out after the initial assessment were comparable with other families on 24 of 26 variables but had higher ratings on child reports of TKD Bullying, F(1, 106) = 4.43, p = .038, and on parent reports of Frequency of Sibling Warmth, F(1, 105) = 5.64, p = .019. We checked comparability of the GS with trial families on key demographic measures. The samples were comparable on children’s age (F [1, 268] = 1.30, p = .255), grade (F [1, 280] = 1.91, p = .168), and gender (χ2 [1] = 1.42, p = .246) and on family income (F [1, 256] < 0.01, p = .994). The trial sample had a higher proportion of children with preexisting diagnosis than the GS, (χ2 [1] = 21.76, p < .001). Parents in the trial were also older (F [1, 254] = 78.39, p < .001), had a higher educational level (F [1, 254] = 4.32, p = .039), and included less parents born outside Australia (χ2 [1] = 4.99, p = .036) than the GS. Effects on Victimization by Peers and Child Distress4 Table 1 shows means, SDs and LMM results for the primary outcome variables of peer victimization and child emotional distress. There are significant main effects for time across conditions for most variables. Simple effect sizes in the RTP group ranged from medium for the teacher report of Relational Victimization (d = 0.56) to very large for the child report of Internalizing Feelings (d = 1.34), and for the AC group, from no change for the teacher report of Overt Victimization to a medium effects for child measures of TDK Upset and Internalizing Feelings. There were significantly greater improvements in RTP than AC children over time in teacher ratings of Overt Victimization F(1, 247.07 5) = 4.47, p = .036, 6 child ratings of Internalizing Feelings over time than AC children, F(1, 116.46) = 6.77, p < .010, and parent ratings of Child Depression, F(1, 161.68) = 7.52, p = .007. Table 1. Effects of Time and Intervention on Child Bullying and Internalizing Problems Dependent Variable Intervention (n = 56) Active Control (n = 54) Main Effect of Time Time X Condition Interaction Simple effect of time for RTP from time 1 to 3 d Simple effect of time on AC group from time 1 to 3 d Treatment Effect (RTP vs AC) Times 1 to 2 d [95% Confidence Interval] Treatment Effect (RTP vs AC) Times 1 to 3 d [95% Confidence Interval] Time 1 Time 2 Time 3 Time 1 Time 2 Time 3 F p F p M (SD) M (SD) M (SD) M (SD) M (SD) M (SD) Overt Victimiztn 1.73 (0.72) 1.38 (0.60) 1.39 0.58) 1.63 (0.68) 1.57 (0.71) 1.69 (0.88) 3.52 .062 4.47 .036 0.47 - 0.08 0.40 [− .001, 0.81] 0.56 [0.13, 1.00] Relational Victimiztn 2.63 (0.92) 2.36 (0.92) 2.11 (0.75) 2.57 (0.93) 2.42 (0.92) 2.28 (1.06) 10.01 .002 0.85 .358 0.54 0.30 0.13 [− 0.28, 0.53] 0.24 [− 0.18, 0.67] TKD Bullying 0.99 (0.73) 0.56 (0.59) 0.50 (0.59) 1.24 (.84) 0.87 (0.78) 0.74 (0.72) 45.14 < .001 0.01 .937 0.66 0.58 - 0.07 [− 0.32, 0.46] 0.01 [− 0.41, 0.38] TKD Upset 2.30 (1.43) 1.16 (1.25) 0.78 (1.23) 2.33 (1.52) 1.56 (1.35) 1.34 (1.49) 61.05 < .001 2.50 .117 1.05 0.64 0.25 [− 0.14, 0.65] 0.36 [− 0.04, 0.76] Internalizing Cognitions 0.39 (0.19) 0.24 (0.17) 0.23 (0.18) 0.44 (0.22) 0.32 (0.24) 0.30 (0.23) 34.32 < .001 0.00 .959 0.79 0.64 0.12 [0.27, 0.52] 0.06 [− 0.34, 0.45] Internalizing Feelings 1.07 (0.51) 0.52 (0.43) 0.38 (0.36) 0.95 (0.62) 0.63 (0.58) 0.58 (0.56) 89.97 < .001 6.77 .010 1.34 0.58 0.42 [0.02, 0.82] 0.59 [0.18, 0.99] Child Depression 5.05 (2.82) 3.33 (2.17) 3.04 (2.74) 5.05 (2.96) 4.46 (2.97) 4.65 (2.76) 20.39 < .001 7.52 .007 0.70 0.13 0.39 [− 0.01, 0.78] 0.56 [0.15, 0.96] Note: Overt Victimiztn = Overt Victimization (PPVM-T) teacher report; Relational Victimiztn = Relational Victimization (PPVM-T) teacher report; TKD = Thing Kids Do child report. Table options Table 2 includes means, SDs and ANOVAs for children’s and parents’ reports of overall change in victimization and emotional distress at Times 2 and 3. RTP children reported greater reductions in bullying than AC children at Times 2 and 3, F(1, 94) = 8.14, p = .005, and RTP parents reported greater reductions than AC parents, F(1, 94) = 16.18, p < .001. The mean Time 3 rating for AC parents was 0.83 and for RTP parents was 0.37, between “less” (1) and “much less” (0) bullying. Treatment effect sizes on child and parent measures were medium (d = 0.44; d = 0.52). For child distress, at Times 2 and 3, RTP children reported greater reductions than did AC children, F(1, 94) = 12.79, p = .001, and RTP parents rated greater improvements in child coping than AC parents, F(1, 94) = 21.22, p < .001. Mean ratings were 2.97 for AC and 3.37 for RTP parents where “3” means “coping better” and “4” means “coping much better” compared to Time 1. Treatment effect sizes for child distress were medium to large for both child (d = 0.63) and parent (d = 0.74) reports. Table 2. Parent and Child Reports of Change on Child Bullying and Internalizing Problems Dependent Variable Respond Time 2 (3 months) Time 3 (9 months) Difference Between Conditions Treatmt effect Time 2 d Treatmt effect Time 3 d M (SD) M (SD) F p RTP AC RTP AC Amount bullying Child 0.18 (0.56) 0.43 (0.65) 0.20 (0.57) 0.51 (0.80) 8.14 .005 0.41 0.44 Feeling better Child 1.73 (0.60) 1.43 (0.83) 1.82 (0.52) 1.38 (0.82) 12.97 .001 0.41 0.63 Amount bullying Parent 0.39 (0.49) 0.98 (1.00) 0.37 (0.72) 0.83 (1.02) 16.18 < .001 0.75 0.52 Child coping better Parent 3.37 (0.63) 2.68 (1.01) 3.58 (0.64) 2.97 (0.98) 21.22 < .001 0.82 0.74 Note: Respond = Respondent, Treatmt effect = Treatment Effect. Table options Table 3 shows normative comparison data including means and SDs, ANOVAs and post hoc tests of differences between RTP and AC groups at Times 1 and 3 and the GS. Time 1 RTP and AC scores were poorer than the GS across all primary outcome measures (listed in top half of Table 3). By Time 3, neither the AC group nor RTP groups were different to the GS on child reports of TKD Bullying or Internalizing Cognitions. By Time 3 the AC group was equivalent to the GS and the RTP reported significantly lower levels of distress than the GS for child reports of TKD Upset and Internalizing Feelings. Despite RTP parents reporting significantly less Child Depression than AC parents, both RTP and AC groups remained higher than the GS at Time 3. Table 3. Normative Comparison of Pre and Post RTP and AC Scores10 Measure General sample GS n = 198 RCT time 1 (0 months) RCT Time 3 (9 months) Time 1 overall difference Post hoc difference at Time 1 Time 3 overall difference Post hoc differences at Time 3: RTP n = 39 AC n = 46 RTP n = 50 AC n = 47 F p RTP vs GS p AC vs GS p RTP vs AC p F p RTP vs GS p AC vs GS p RTP vs AC p Mean (SD) Mean (SD) Mean (SD) Mean (SD) Mean (SD) TKD Bullying 0.66 (0.71) 1.04 (0.75) 1.32 (0.92) 0.48 (0.62) 0.78 (0.74) 15.40 < .001 .005 < .001 .289 1.79 .170 .416 > .999 .204 TKD Upset 1.59 (1.50) 2.39 (1.45) 2.54 (1.52) 0.80 (1.23) 1.56 (1.54) 10.23 < .001 .003 .001 > .999 4.88 .008 .006 > .999 .083 Internalizing Cognitions 0.25 (0.18) 0.37 (0.18) 0.47 (0.21) 0.21 (0.17) 0.29 (0.24) 27.67 < .001 < .001 < .001 .055 1.77 .173 .569 .795 .184 Internalizing Feelings 0.85 (0.50) 1.09 (0.49) 1.07 (0.62) 0.39 (0.37) 0.72 (0.57) 6.44 .002 .010 .037 > .999 14.68 < .001 < .001 .454 .015 Child Depression 1.63 (2.25) 4.97 (2.88) 4.93 (3.08) 3.02 (2.74) 4.42 (2.82) 49.88 < .001 < .001 < .001 > .999 21.49 < .001 .003 < .001 .039 Reactive Aggression 0.45 (1.01) 0.63 (1.14) 0.97 (1.46) 0.05 (0.22) 0.59 (1.21) 3.83 .023 0.951 .021 .459 3.62 .028 .048 > .999 .050 Friendedness 4.23 (0.69) 3.61 (0.83) 3.39 (0.86) 4.16 (0.57) 3.83 (0.80) 26.66 < .001 < .001 < .001 .554 4.73 .010 > .999 .007 .126 Likes School 4.19 (1.14) 3.62 (1.28) 3.56 (1.52) 3.98 (1.08) 3.15 (1.44) 6.94 .001 .016 .012 > .999 11.45 < .001 .884 < .001 .008 Facilitative Parenting 3.82 (0.36) 3.83 (0.30) 3.82 (0.33) 4.03 (0.31) 3.93 (0.36) .015 .985 > .999 > .999 > .999 6.21 .002 .003 .310 .695 Table options Table 4 displays outcomes for children with clinically elevated scores at Time 1. On the child report of TKD Bullying, by Time 3, 74% of RTP children and 57% of AC children had moved outside the clinical range, with the difference between conditions not significant, p = .171. For all measures of child distress a significantly greater proportion of RTP than AC children moved out of the clinical range: this included 79% of RTP compared with 53% of AC children for TKD Upset; 86% of RTP compared with 56% of AC children for Internalizing Feelings; 67% of RTP compared with 43% of AC children for Internalizing Cognitions; and 65% of RTP compared with 38% of AC children for Child Depression. Table 4. Significance of Clinical Change: Comparison of RCT Groups With General Population Sample Measure Clinical improvement at Time 2 Clinical improvement at Time 3 Proportion of cases clinically improved % age (n/n) Difference between conditions Proportion of cases clinically improved % age (n/n) Difference between conditions RTP AC χ2 p RTP AC χ2 p Primary Outcome Measures TKD Bullying 57.69% (15/26) 42.42% (14/33) 1.36 .184 73.91 % (17/23) 57.14 % (16/28) 1.56 .171 TKD Upset 62.07% (18/29) 45.16% (14/31) 1.72 .146 78.57 % (22/28) 53 % (16/30) 4.08 .040 Internalizing feelings (SPBI) 80.00% (24/30) 61.54% (16/26) 2.33 .110 86.21 % (25/29) 56 % (14/25) 6.16 .015 Internalizing cognitions (SPBI) 70.97% (22/31) 47.22 (17/36) 3.86 .042 66.67% (20/30) 42.86% (15/35) 3.69 .047 Child depression 45.71 % (16/35) 34.29% (12/35) 0.95 .232 64.71% (22/34) 20.59% (7/34) 13.53 < .001 Secondary Outcome Measures Reactive aggression (SPBI) 71.43% (10/14) 38.10% (8/21) 3.74 .055 61.65 % (8/13) 30.00% (6/20) 3.21 .076 Friendedness 60.71% (17/28) 23.53% (8/34) 8.82 .003 51.85 % (14/27) 48.48 % (16/33) 0.67 .500 I like school 16.67% (4/24) 21.74% (5/23) 0.20 .471 62.50 % (15/24) 13.64% (3/22) 11.51 .001 Facilitative Parenting 42.31% (11/26) 22.72% (5/22) 2.06 .130 61.54% (16/26) 38.10% (8/21) 2.56 .096 Table options Effects on Child Social Behavior and Relationships Table 5 shows means, SDs, and LMM analyses for secondary outcomes. There were significant main effects for time over both conditions for 63% of variables. For 47% of variables there were significantly greater improvements over time for RTP than AC children. Table 5. Effects of Intervention on Secondary Outcome Variables Dependent Variable Intervention (n = 56) Active Control (n = 54) Main Effect of Time Time X Condition Interaction Simple effect of time for RTP from time 1 to 3 d Simple effect of time for AC from time 1 to 3 d d treatment effect time 1 to 2 d treatment effect time 1 to 3 Time 1 Time 2 Time 3 Time 1 Time 2 Time 3 F p F p M (SD) M (SD) M (SD) M (SD) M (SD) M (SD) Overt aggression 1.47 (0.75) 1.30 (0.63) 1.21 (0.37) 1.36 (0.69) 1.43 (0.86) 1.69 (0.88) 1.91 .169 8.61 .004 0.33 - 0.17 0.33 0.51 Relational aggression 1.78 (0.82) 1.60 (0.69) 1.56 (0.65) 1.61 (0.67) 1.66 (0.74) 2.28 (1.06) 1.32 .252 0.79 .376 0.26 - 0.03 0.28 0.30 Reactive aggression 0.59 (1.07) 0.16 (0.46) 0.16 (0.42) 0.88 (1.33) 0.50 (0.97) 0.55 (1.08) 14.37 < .001 0.12 .735 0.39 0.24 0.04 0.08 Friendedness 3.58 (0.85) 4.10 (0.67) 4.12 (0.58) 3.41 (0.88) 3.62 (0.84) 3.86 (0.82) 43.44 < .001 0.14 .707 0.62 0.50 0.36 0.10 I Like School 3.40 (1.37) 3.43 (1.28) 3.96 (1.01) 3.48 (1.44) 3.27 (1.01) 3.13 (1.35) 0.37 .543 10.44 .002 0.40 - 0.24 0.17 0.65 Liked by peers of same sex 3.17 (0.97) 3.40 (1.01) 3.60 (1.01) 3.43 (1.05) 3.46 (1.24) 3.39 (1.05) 4.08 .055 5.19 .032 0.43 - 0.04 0.20 0.46 Liked by peers of opposite sex 2.79 (0.92) 3.09 (1.09) 3.37 (1.09) 3.21 (1.14) 3.18 (1.22) 3.18 (1.00) 4.64 .032 6.72 .010 0.62 - 0.03 0.32 0.60 Frequency of Sibling Warmth 3.31 (0.56) 3.34 (0.54) 3.50 (0.56) 3.40 (0.56) 3.41 (0.53) 3.41 (0.48) 8.46 .004 5.10 .025 0.32 0.02 0.04 0.30 Problem of Sibling Warmth 1.28 (0.34) 1.18 (0.27) 1.16 (0.28) 1.31 (0.43) 1.25 (0.33) 1.27 (0.38) 9.13 .003 1.50 .223 0.36 0.09 0.07 0.21 Freq Sibling Agonism 2.81 (0.57) 2.66 (0.59) 2.58 (0.56) 2.99 (0.69) 2.91 (0.65) 2.92 (0.77) 6.36 .013 2.29 .133 0.40 0.09 0.12 0.25 Prob Sibling Agonism 1.71 (0.54) 1.54 (0.51) 1.43 (0.46) 1.95 (0.67) 1.81 (0.68) 1.84 (0.75) 17.93 < .001 3.90 .050 0.51 0.16 0.03 0.28 Encourages bullying 7.57 (4.43) 4.25 (3.57) 2.50 (3.15) 8.52 (4.52) 7.27 (5.05) 3.90 (0.35) 56.63 < .000 10.65 .001 1.13 0.40 0.46 0.72 CRPA Child Uses Words 1.50 (0.93) 2.04 (0.96) 1.92 (1.03) 1.17 (1.08) 1.21 (0.99) 1.00 (1.02) 1.52 .220 5.65 .019 0.44 - 0.15 0.49 0.58 CRPA Child Uses Tells 0.79 (1.07) 0.35 (0.72) 0.42 (0.78) 0.74 (0.97) 0.54 (0.68) 0.70 (0.81) 5.04 .026 2.48 .117 0.37 0.04 0.34 0.32 CRPA - coding Assertiveness 2.26 (0.81) 2.88 (0.88) 2.42 (0.84) 2.21 (0.82) 2.18 (0.80) 2.11 (0.71) 0.31 .576 1.45 .230 0.19 - 0.11 0.79 0.31 CRPA – coding Provocativeness 2.02 (0.92) 1.78 (0.89) 1.77 (0.88) 2.28 (1.05) 2.27 1.08 2.34 (1.10) 0.69 .407 2.15 .145 0.26 - 0.05 0.22 0.30 Facilitative Parenting 3.82 (0.30) 3.95 (0.32) 4.03 (0.32) 3.82 (0.33) 3.92 (0.34) 3.93 (0.37) 30.27 < .001 4.57 .035 1.12 0.56 0.08 0.29 Warm responsiveness 6.02 (1.51) 6.74 (1.09) 6.57 (1.27) 5.84 (1.46) 5.96 (1.50) 6.26 (1.43) 10.02 .002 0.04 .838 0.36 0.28 0.40 0.08 Intrusive demandingness 3.48 (1.50) 3.49 (1.23) 3.47 (1.29) 3.56 (1.55) 4.14 (1.47) 4.23 (1.51) 3.61 .060 3.64 .059 0.00 - .42 0.37 0.43 Note: Freq Sibling Agonism = Frequency of Sibling Agonism; Prob Sibling Agonism = Problem of Sibling Agonism, Encourages bullying = Actor assessment of how much child’s response encourages bullying. Table options Children’s Social Behavior and Peer Relationships Teachers reported greater reductions in Overt Aggression towards peers for RTP than with AC children (p = .004), 7 and greater improvements in acceptance of RTP children by both same sex peers (p = .032), and opposite sex peers (p = .010). Treatment effect sizes were all in the medium range. There were no significant changes in Relational Aggression. Children’s reports of Friendedness and Reactive Aggression showed improvements over time across conditions but not between conditions. Normative comparisons in Table 3 show both AC and RTP groups scored lower than the GS on Friendedness at Time 1 (p < .001, p < .001), but by Time 3, the RTP group was similar to the GS (p > .999). Table 4 shows a higher proportion of RTP (61%) than AC children (24%) who were elevated at Time 1 were out of the clinical range by Time 2, but by Time 3, similar proportions of RTP (52%) to AC children (48%) were in the normal range. For Reactive Aggression, Table 3 also shows that RTP children at Time 3 reported significantly lower reactive aggression than the GS (p = .048), with the RTP mean near “0” and SDs constricting over time, suggesting a floor effect. Table 4 shows that, for children elevated on Reactive Aggression at Time 1, there was a trend towards a higher proportion of RTP (62%) than AC (30%) children moving outside the clinical range by Time 3 ( p = .076). The CRPA measures in Table 4 show that RTP children reported significantly greater increases over time than AC children in using adaptive words to solve peer problems, d = 0.58. CRPA Tells shows that across both conditions children reported they would tell the teacher less often over time, p = .026. Table 5 shows that actors rated significantly greater improvements in RTP than AC children in CRPA Encourages Bullying, with a medium to large treatment effect size, d = 0.72. Coding of Child Assertiveness and Provocativeness produced no significant results for change over time. Child Assertiveness of RTP children spiked at Time 2 (d = 0.79) before decreasing at Time 3 (d = 0.31). Liking School Table 5 shows greater improvements over time for RTP than AC children’s ratings on “I like school,” p = .002, d = 0.65. Table 3 shows that both AC and RTP children liked school less than GS children at Time 1 (p = .012; p = .016), but by Time 3, RTP children were no different to the GS (p = .884). Table 4 shows a greater proportion of RTP (63%) than AC children (14%) moved out of the clinical range by Time 3, p < .001. Parenting and Sibling Relationships Table 5 shows RTP families had greater improvements than AC families over time on facilitative parenting (p = .035), beyond improvements for across conditions over time (p < .001). Table 3 shows that at Time 1, the GS, RTP, and AC groups were no different on facilitative parenting, but by Time 3 RTP families scored significantly higher than the GS. 8 Table 4 shows, for parents low in facilitative parenting at Time 1, a trend towards greater improvements for RTP than AC families, p = .096. For coding of Intrusive-Demandingness in Table 5, there was a strong trend towards significantly differential change over time between RTP and AC families (p = .059), resulting mainly from increasing means for AC, but not RTP families. Warm Responsiveness showed significant increases across both conditions over time (p = .002). Table 5 shows significant main effects of time on all four sibling relationship measures with greater increases in Frequency of Sibling Warmth (p = .025) and marginally greater reductions in Problem of Agonism (p = .050 9), for RTP than AC families over time. Family Satisfaction with Program Mean ratings of parents across all questions ranged from 5.98 (SD = 1.03) to 6.70 (SD = 0.61) out of a maximum of 7 with a grand mean of 6.30 across questions. Parents gave a mean rating of 6.46 (SD = 0.92) for their overall satisfaction. Children gave a mean rating of 3.10 (SD = 0.98) for the program, between “very” (3) and “extremely helpful” (4). Contextual Changes Over Period of Monitoring There was one significant difference between groups in contextual changes which had occurred during the 9 months: significantly more AC families (22%) than RTP families (6%) reported that “the child who was bullying left the school,” t(1, 91) = 24.23, p < .001.