using a virtual role-play environment to reduce bullying in primary schools: the fearnot! evaluation...
TRANSCRIPT
Using a virtual role-play environment to reduce bullying in primary schools:
The FearNot! evaluation
Moral Disengagement
University of Wuerzburg
Natalie Vannini & Wolfgang Schneider
Supported by the EU Framework VI under IST-4-027656-STP
Moral Disengagement - MD
• MD refers to socio-cognitive processes through which the average person is able to commit horrible acts against others (Bandura 1999, 2002)
• Four major psychological mechanisms:– cognitive restructuring of harmful behaviour– obscuring or minimizing one’s role in causing harm– disregarding or distorting the impact of harmful behaviour– blaming and dehumanizing the victim
• MD Scale (Hymel et al, 2005) including 11 items (plus Example item), i.e.– “Some kids get bullied because they deserve it.“ – “Getting bullied helps to make people tougher.”
MD Questionnaire
MD Analyses of Scale
• Factor analysis (principal components, varimax rotation) on the 17 moral disengagement items– failed to differentiate the four categories of moral
disengagement concept. – Instead most items (11 items*) loaded on a single
factor (explained 35,5% of the variance).
• Cronbach´s alpha =.69 - .80 (assessment 1-3) • Moral disengagement score = mean of the 11
items (higher scores = higher level of MD)
*items: 01,03,04,05,09,10,11,13,14,15,17
MD & Total Longitudinal Sample
• Longitudinal sample = 916/1129 (81,1%)– 443 in the intervention group – 473 in the control group children
• UK sub-sample = 500• German sub-sample = 416 • Mean age overall: 8,89 years (SD=0.74)
– Mean age sig. higher for UK pupils than German pupils: 9.36 (SD= 0.53) vs 8.34 (SD=0.55) years (F[1,909] = 793.9; p <.001).
MD & Total Sample: Gender & Bullying Roles: Means & SD
– Includes baseline MD as a covariateGroup N Mean s.d.
Bullies
Girls 30 1.47 .30
Boys 28 1.55 .33
Total 58 1.51 .31
Victims
Girls 82 1.35 .29
Boys 96 1.41 .34
Total 178 1.38 .31
Non-Involved
Girls 280 1.31 .23
Boys 304 1.40 .27
Total 584 1.36 .25
Bully-Victims
Girls 30 1.41 .22
Boys 22 1.67 .29
Total 52 1.48 .27
MD Effects of Gender & Bullying Roles
• Sig. main effect for the Bullying Roles– (F [3,871]=11.70, p<.001); Post hoc analyses (Bonferroni
comparisons):• Bullies significantly higher level of MD than victims and
non-involved children • Bully-victims significantly higher level of MD than
victims and non-involved children
• Sig. main effect for Gender – Girls scored significantly lower than boys. (F[1,871] =18.90
p<.001)
• No interaction effect for Roles x Gender• UK children sig. higher level of MD than
German children – (1,43 vs 1,33; t= 4.87 p <.001)
MD & Intervention: Results Total Sample
– Includes baseline MD as a covariateGroup N Mean s.d.
Baseline
Intervention 443 1.40 .28
Control 473 1.37 .28
Total 916 1.38 .28
Post-Test
Intervention 443 1.38 .31
Control 473 1.36 .30
Total 916 1.37 .30
Follow-up Test
Intervention 443 1.37 .33
Control 473 1.35 .31
Total 916 1.36 .32
– No between groups main effect of ‘Group’, F(1, 914)=2.25, p=.134.
– Sig. within subjects main effect of ‘Time’, F(2, 914)=3.04, p=.048.
– No interaction effect of Group vs Time, F(2, 914)=0.315, p=.730.
MD & Intervention: Girls Sub-sample
– Includes baseline MD as a covariateGroup N Mean s.d.
Baseline
Intervention 225 1.34 .25
Control 212 1.34 .26
Total 437 1.34 .26
Post-Test
Intervention 225 1.31 .24
Control 212 1.34 .26
Total 437 1.33 .25
Follow-up Test
Intervention 225 1.31 .28
Control 212 1.32 .26
Total 437 1.31 .27
– No main effect of ‘Group’, F(1, 435)=0,335, p=.563.
– Sig. within subjects main effect of ‘Time’, F(2, 435)=3,227, p=.040.
– No interaction effect of Group vs Time, F(2, 435)=0,446, p=.640.
MD & Intervention: Boys Sub-sample
– Includes baseline MD as a covariateGroup N Mean s.d.
Baseline
Intervention 217 1.46 .29
Control 262 1.39 .30
Total 479 1.42 .30
Post-Test
Intervention 217 1.45 .35
Control 262 1.37 .32
Total 479 1.41 .34
Follow-up Test
Intervention 217 1.44 .36
Control 262 1.39 .34
Total 479 1.41 .35
– Sig. main effect of ‘Group’, F(1, 477)=7.180, p=.008.
– No within subjects main effect of ‘Time’, F(2, 477)=0.653, p=.521.
– No interaction effect of Group vs Time, F(2, 477)=0,313, p=.731.
MD & Intervention: German Sub-sample
– Includes baseline MD as a covariateGroup N Mean s.d.
Baseline
Intervention 204 1.34 .30
Control 212 1.33 .27
Total 416 1.33 .29
Post-Test
Intervention 204 1.31 .31
Control 212 1.32 .29
Total 416 1.32 .30
Follow-up Test
Intervention 204 1.29 .31
Control 212 1.32 .32
Total 416 1.31 .31
– No between groups main effect of ‘Group’, F(1, 414)=0.170, p=.680.
– No within subjects main effect of ‘Time’, F(2, 414)=2,858, p=.058.
– No interaction effect of Group vs Time, F(2, 414)=1.554, p=.213.
MD & Intervention: UK Sub-sample
– Includes baseline MD as a covariateGroup N Mean s.d.
Baseline
Intervention 238 1.45 .25
Control 262 1.40 .29
Total 500 1.43 .27
Post-Test
Intervention 238 1.44 .30
Control 262 1.39 .30
Total 500 1.41 .30
Follow-up Test
Intervention 238 1.45 .33
Control 262 1.38 .30
Total 500 1.41 .32
– Sig. main effect of ‘Group’, F(1, 498)=6,451, p=.011.
– No within subjects main effect of ‘Time’, F(2, 498)=0.686, p=.504.
– No interaction effect of Group vs Time, F(2, 498)=.212, p=.809.
MD & Intervention: Conclusions Total Sample
• No significant effect of intervention for total sample– MD decreases in both groups– Overall, control group scores lower than
intervention group scores
• Bullies scored higher at baseline than victims and non-involved children – Further analysis with bully sub-sample
Moral Disengagement & Bully Sub-sample
• Baseline Bullies (self-report) MD Sample = 58 / 916 (6,3%)– 25 in the intervention group – 33 in the control group children – 28 males & 30 females
• UK sub-sample = 54• German sub-sample = 4• Mean age: 9,26 years (SD=0.61)
MD & Intervention: Results Bully Sub-sample
Group N Mean s.d.
Baseline
Intervention 25 1.55 .33
Control 33 1.47 .29
Total 58 1.51 .31
Post-Test
Intervention 25 1.45 .36
Control 33 1.49 .32
Total 58 1.48 .34
Follow-up Test
Intervention 25 1.42 .35
Control 33 1.46 .31
Total 58 1.44 .32
– No between groups main effect of ‘Group’, F(1, 56)=.003, p=.956
– No within subjects main effect of ‘Time’, F(2, 56)=1.930, p=.150.
– No interaction effect of Group vs Time, F(2, 56)=.1,980, p=.143.
MD & Intervention: Conclusions Bully Sub-sample
• No significant effect of intervention for bully sub-sample– Trend: MD
decreases more for intervention group bullies
Intervention & sub-scale “Blaming the victim”
• Intention of FearNot!: (among others) changing attitude and promoting empathy towards victims of bullying
• MD-Scale is composed of a Bullying/ Victim Attitude Questionnaire (see Hymel et al study)
• “Blaming the victim” sub-scale: items 12-17 plus item 10 (“Some children need to be picked on just to teach them a lesson”)– Hymel et al subcategorised item 10 under “distorting of negative
consequences” but theoretically and statically (factor analysis) it also fits under “blaming the victim” sub-scale
• Factor Analysis (principal components): one factor explaining 29,3% of the variance
• Conbach´s alpha: AS1=.59; AS2=.68; AS3=.72
BV Sub-scale & Intervention: Results Total Sample
Group N Mean s.d.
Baseline
Intervention 440 1.60 .38
Control 473 1.60 .37
Total 913 1.60 .38
Post-Test
Intervention 440 1.59 .39
Control 473 1.59 .40
Total 913 1.59 .40
Follow-up Test
Intervention 440 1.58 .42
Control 473 1.59 .40
Total 913 1.58 .41
– No between groups main effect of ‘Group’, F(1, 911)=.024, p=.877
– No within subjects main effect of ‘Time’, F(2, 911)=1.258, p=.284.
– No interaction effect of Group vs Time, F(2, 911)=..240, p=.787.
BV Sub-scale & Intervention: Results Bully Sub-Sample
Group N Mean s.d.
Baseline
Intervention 25 1.80 .48
Control 33 1.82 .36
Total 58 1.81 .41
Post-Test
Intervention 25 1.68 .39
Control 33 1.84 .45
Total 58 1.77 .43
Follow-up Test
Intervention 25 1.65 .45
Control 33 1.72 .39
Total 58 1.69 .42
– No between groups main effect of ‘Group’, F(1, 56)=.0.701, p=.406
– Sig. within subjects main effect of ‘Time’, F(2, 56)=3,333, p=.039.
– No interaction effect of Group vs Time, F(2, 56)=..1,082, p=.342.
BV Sub-scale & Intervention Conclusions
• No significant effect of intervention for total sample
• Sig. change over time for both bully
sub-samples: – Intervention group
bullies: BV mean decreases
– Control group bullies: BV mean increases at 2nd and decreases at 3rd Assessment
MD Summary I
• Total Sample at baseline:– Bullies sig. higher level of MD than victims and non-
involved children – Boys sig. higher level of MD than girls– UK children sig. higher level of MD than German children
• Total Sample & Intervention:– No significant effect of intervention on MD score
• Sub-samples Gender & Intervention– No significant effect of intervention on MD score– Girls MD level decreases over time;
• interesting pattern: interv. group MD scores decrease after FN!- intervention; control group MD scores decrease from post- to follow-up test
– Boys of intervention group sig. higher level of MD than control group boys
MD Summary II
• Sub-samples Country & Intervention– No significant effect of intervention on MD score– UK- sample: Control group lower level of MD than
intervention group
• Sub-Sample Bullies & Intervention– No significant effect of intervention on MD level for bully
sub-sample• Interesting trend: MD decreases more for
intervention group bullies
• Total Sample & Sub-scale “Blaming the victim” – No significant effect of intervention on `negative attitudes
towards victims`
MD Summary II
• Sub-Sample Bullies & Sub-scale “Blaming the victim” – No significant effect of intervention on `negative attitudes
towards victims` for bully sub-sample– Sig. change over time for both bully sub-samples:
• Intervention group bullies: BV mean decreases• Control group bullies: BV mean increases at 2nd and
decreases at 3rd Assessment
Using a virtual role-play environment to reduce bullying in primary schools:
The FearNot! evaluation
Moral DisengagementResults part Two
Uni Wuerzburg
Results MD part two Overview
Overview: • Slides 3-9:
– Analyses of Moral Disengagement (MD) score and “Blaming the Victim” (BV) subscale score regarding FN! intervention within the (self reported) Bully sub sample
• Bully sample: this time focus on self reported bully status regardless of any other self reported status, in other words this bully sub sample include also actually (self reported) “bully-victims”
• Slides 10 – 21– Analyses of MD score and BV sub scale score regarding FN!
intervention within two peer nominated bully sub samples • Sample 1: peer nominated pure bully sub sample • Sample 2: peer nominated bully ( regardless of any other
nomination in any other category) sub sample
Baseline: MD Mean Differences & Bullying roles (self-report)
Baseline Roles
N MDMean
s.d.
Bullies 110 1.51 .30
Victims 178 1.38 .32
Non-Involved584 1.36 .26
Total872 1.38 .28
• Sig. main effect for the Bullying Roles– (F [2,871]=14.41, p<.001); Post hoc analyses (Bonferroni comparisons):
• Bullies significantly higher level of MD than victims and non-involved children (p=.05)
Baseline: MD Mean Differences & Bullying roles (self-report)
Baseline: Blaming the Victim (BV) sub scale & Bullying roles (self-report)
Baseline Roles
N BV sub scaleMean
s.d.
Bullies 110 1.78 .36
Victims 178 1.58 .42
Non-Involved583 1.57 .35
Total871 1.60 .38
• Sig. main effect for the Bullying Roles– (F [2,870]=14.64, p<.001); Post hoc analyses (Bonferroni comparisons):
• Bullies significantly higher level of MD than victims and non-involved children (p=.05)
Baseline: Blaming the Victim (BV) sub scale & Bullying roles (self-report)
MD MEAN & Intervention: Results Bully (self report) Sub-Sample
Group N Mean s.d.
Baseline
Intervention 47 1.55 .28
Control 63 1.48 .30
Total 110 1.51 .30
Post-Test
Intervention 47 1.51 .37
Control 63 1.49 .32
Total 110 1.50 .34
Follow-up Test
Intervention 47 1.50 .40
Control 63 1.46 .31
Total 110 1.48 .35
– No between groups main effect of ‘Group’, F(1, 108)=.0.551, p=.460
– Non Sig. within subjects main effect of ‘Time’, F(2, 108)=0.855, p=.427.
– No interaction effect of Group vs Time, F(2, 108)=..0.464, p=.630.
BV Sub-scale & Intervention: Results Bully (self-report) Sub-Sample
Group N Mean s.d.
Baseline
Intervention 47 1.78 .38
Control 63 1.78 .36
Total 110 1.78 .37
Post-Test
Intervention 47 1.76 .38
Control 63 1.79 .40
Total 110 1.78 .39
Follow-up Test
Intervention 47 1.73 .47
Control 63 1.73 .40
Total 110 1.73 .43
– No between groups main effect of ‘Group’, F(1, 108)=.0.005, p=.946
– No sig. within subjects main effect of ‘Time’, F(2, 108)=1.091, p=.338.
– No interaction effect of Group vs Time, F(2, 108)=..0.094, p=.911.
BV Sub-scale & Intervention: Results Bully (self-report) Sub-Sample
MD & Bullying – Bullying roles peer nomination
• Longitudinal sample = 916/1129 (81,1%)– 443 in the intervention group – 473 in the control group children
• UK sub-sample = 500• German sub-sample = 416 • Mean age overall: 8,89 years (SD=0.74)• Bullying roles by peer nomination:
– to account for the different class sizes children were classified as bully/ victim/ bully-victim/ defender or non-involved children who were nominated one SD above the mean nomination within the class :
• at baseline: 69 (7,5%) pure bullies; 99 (10,8%) pure victims; 54 (5,9%) bully-victims; 64 (7,0%) defenders; 630 (68,8%) non-involved children
Baseline: MD Mean Differences & Bullying roles (peer-nomination)
• Sig. main effect for Bullying Roles (peer nomination)– (F [2,915]=2.593, p=.035); Post hoc analyses (Bonferroni
comparisons):• Bullies significantly higher level of MD than non-involved children (p=.05)
MD MEAN & Intervention: Results pure Bully (peer nomination) Sub-Sample
Group N MD Mean s.d.
Baseline
Intervention 31 1.54 .35
Control 38 1.42 .27
Total 69 1.47 .31
Post-Test
Intervention 31 1.59 .47
Control 38 1.44 .36
Total 69 1.51 .42
Follow-up Test
Intervention 31 1.57 .42
Control 38 1.44 .34
Total 69 1.50 38
– No between groups main effect of ‘Group’, F(1, 67)=.3.331, p=.074
– No Sig. within subjects main effect of ‘Time’, F(2, 67)=0,392, p=.677.
– No interaction effect of Group vs Time, F(2, 67)=..0.109, p=.897.
Baseline: Blaming the Victim (BV) sub scale & Bullying roles (peer nomination)
Baseline Roles
N BV sub scaleMean
s.d.
Pure Bullies 69 1.66 .41
Bully-Victim 54 1.65 .38
Pure Victims 99 1.63 .41
Pure Defender 64 1.61 .35
Non-Involved 630 1.59 .37
Total 916 1.60 .38
• No Sig. main effect for the Bullying Roles (peer nomination) – (F [4,915]=1,184, p=.316)
BV Sub-scale & Intervention: Results pure Bully (peer nomi) Sub-Sample
Group N BV sub scale Mean
s.d.
Baseline
Intervention 31 1.71 .41
Control 38 1.63 .40
Total 69 1.66 .41
Post-Test
Intervention 31 1.80 .51
Control 38 1.64 .47
Total 69 1.71 .49
Follow-up Test
Intervention 31 1.79 .51
Control 38 1.67 .39
Total 69 1.72 .45
– No between groups main effect of ‘Group’, F(1, 67)=.1.699, p=.197
– No sig. within subjects main effect of ‘Time’, F(2, 67)=0.691, p=.503.
– No interaction effect of Group vs Time, F(2, 67)=..0.313, p=.732.
Baseline: MD MEAN & Bullying roles (peer nomination)
Baseline Roles
N MDMean
s.d.
Bullies 123 1.46 .31
Victim 99 1.39 .27
Defender 64 1.38 .25
Non-Involved 630 1.37 .28
Total 916 1.38 .28
• Sig. main effect for the Bullying Roles (peer nomination) – (F [3,915]=3.281, p=.020); Post hoc analyses (Bonferroni comparisons):
• Bullies significantly higher level of MD than non-involved children (p=.05)
MD MEAN & Intervention: Results Bully (peer nomination) Sub-Sample
Group N MD Mean s.d.
Baseline
Intervention 52 1.49 .32
Control 71 1.43 .31
Total 123 1.46 .31
Post-Test
Intervention 52 1.52 .41
Control 71 1.44 .38
Total 123 1.47 .39
Follow-up Test
Intervention 52 1.50 .41
Control 71 1.40 .32
Total 123 1.44 36
– No between groups main effect of ‘Group’, F(1, 121)=.2.225, p=.138
– No Sig. within subjects main effect of ‘Time’, F(2,121 )=0,447, p=.640.
– No interaction effect of Group vs Time, F(2, 121)=..0.219, p=.804.
Baseline: BV sub scale & Bullying roles (peer nomination)
Baseline Roles
N BV sub scaleMean
s.d.
Bullies 123 1.66 .39
Victim 99 1.63 .41
Defender 64 1.61 .35
Non-Involved 630 1.59 .37
Total 916 1.60 .38
• No Sig. main effect for the Bullying Roles (peer nomination) – (F [3,915]=1,571, p=.195)
BV sub scale & Intervention: Results Bully (peer nomination) Sub-Sample
Group N BV subscale Mean
s.d.
Baseline
Intervention 52 1.66 .39
Control 71 1.66 .40
Total 123 1.66 .39
Post-Test
Intervention 52 1.75 .46
Control 71 1.66 .35
Total 123 1.70 .45
Follow-up Test
Intervention 52 1.69 .50
Control 71 1.62 .38
Total 123 1.65 43
– No between groups main effect of ‘Group’, F(1, 121)=.726, p=.396
– No Sig. within subjects main effect of ‘Time’, F(2,121 )=1,134 p=.323
– No interaction effect of Group vs Time, F(2, 121)=..0.944, p=.390.
Peer nominated intervention bullies & Sub samples UK and Germany
• Even though it is not significant– As the former slides shows descriptively the mean score of MD as well
as of the “blaming the victim” (BV) sub scale of the peer nominated intervention bullies (either pure [N=69] or not pure [N=123]) increases (extremely) at post and (somewhat) at follow-up
– Analysing both bully sub sample (UK [N=82] and Germany [N=41]) separately regarding “Blaming the victim” (BV) subscale & intervention, also no sig. effects of intervention neither within the UK nor the German sub sample (see next two slides)
– But as the mean of BV subscale will be compared between UK and Germany intervention bullies there is
• no sig difference at baseline [t(50)=1,619, p=.112] • but at post test the UK intervention bullies (M=1,86; SD.40; N=34) scored sig
higher than the German intervention bullies (M=1,55; SD=.51; N=18) [t(50)=2,437, p=.018]
• and at follow-up UK interv. Bullies (M=1,80; SD= 47, N=34) score sig. Higher than German interv. Bullies (M=1,49, SD=49;N=18); [t(509=2,288, p=.026]
– As the mean of MD Scale will be compared between UK and German intervention bullies there is no sig differences at any of the assessment time points; only descriptive the UK bullies score higher at post and follow-up than the German intervention bullies
German sub sample
– No between groups main effect of ‘Group’, F(1, 39)=0,984, p=.327
– No Sig. within subjects main effect of ‘Time’, F(2,39 )=0,011 p=.989
– No interaction effect of Group vs Time, F(2, 39)=..0.553, p=.577.
Group N BV MEAN s.d.
Baseline
Intervention 18 1.54 .43
Control 23 1.39 .33
Total 41 1.46 .38
Post-Test
Intervention 18 1.55 .51
Control 23 1.41 .36
Total 41 1.47 .43
Follow-up Test
Intervention 18 1.49 .49
Control 23 1.46 .40
Total 41 1.47 44
UK sub sample
– No between groups main effect of ‘Group’, F(1, 80)=0,316, p=.575
– No Sig. within subjects main effect of ‘Time’, F(2,80 )=1,775 p=.173
– No interaction effect of Group vs Time, F(2, 80)=..2,572, p=.080.
Group N BV MEAN s.d.
Baseline
Intervention 34 1.72 .36
Control 48 1.79 .37
Total 82 1.76 .36
Post-Test
Intervention 34 1.86 .40
Control 48 1.78 .44
Total 82 1.81 .42
Follow-up Test
Intervention 34 1.80 .47
Control 48 1.69 .46
Total 82 1.74 41