Employment, Social Affairs & Inclusion

FearNot!: a computer based anti-bulling programme designed to foster peer intervention

Evidence level:
Evidence of effectiveness:
? - 0 + ++
? - 0 + ++
Enduring impact:
? - 0 +

Rationale of practice

The FearNot! computer application uses virtual role play and autonomous agents to enable children aged 8 to 11 years to be part of a virtual school environment. The main aim of the program is to help victims escape victimization, reduce overall bullying by reducing the number of bullies and at the same time increase the number of new defenders. In this application, the child user views one physical bullying scenario and one relational scenario. At the beginning of the programme the child provides personal information (name, gender and age) and a unique personal code that matches their off-line questionnaires. After the introduction of the characters, school and situation, users view the first bullying episode, followed by the victimised character seeking rescue in the school library where it starts to communicate with the user. After each episode, the student engages in a conversation with the victimised character, acting as advisor and friend by suggesting coping strategies. Thus, while the learner is not directly part of the virtual role-play that is happening within the bullying episodes, he/she takes over the role of an off-stage “invisible friend” to the victimised character and can influence the storyline slightly.

Practice Category

  • Helping Vulnerable Children, Facilitating Positive Transitions to Adulthood,
  • Fostering Family-Friendly Workplaces

Recommendation pillar

  • Improve education systems’ impact on equal opportunities,
  • Put in place mechanisms that promote children’s participation in decision making that affect their lives

Countries that have implemented practice

  • Germany,
  • United Kingdom

Age groups

  • Middle Childhood (age 6 to 12)

Target groups

  • Children

Years in operation

  • 2007 - Unknown

Type of organisation implementing practice

  • National Government,
  • State/district or Other Sub-national Government,
  • City or Town Government,
  • Other Private Organization

The practice FearNot! (Fun with Empathic Agents to Reach Novel Outcomes in Teaching) has been developed over the past 7 years, initially within the VICTEC (Virtual ICT with Empathic Characters) project and then in the EU funded eCIRCUS (Education through Characters with emotional-Intelligence and Roleplaying Capabilities that Understand Social Interaction) project. The practice was implemented by schools, with support from implementers assigned by local authorities.

Evidence of effectiveness

  • Practice has been evaluated

FearNot! was implemented and evaluated in public primary schools in the UK and Germany in the school year of 2007-08. The schools were recruited by mailing letters, followed by phone calls or personal visits to present the study and the software to the school staff. Initially the design planned random allocation to the intervention and waiting control groups. However, the computer infrastructure was outdated in some schools and made it impossible to run the FearNot! software. The intervention and control groups were assigned based on the computer infrastructure in the schools. The students were allocated to the intervention group if FearNot! software could run on their school’s computers, and to the control group if not. A quasi-experimental, pre/post-tests control group design was used to evaluate program effects. The main individual-level outcomes studied were ‘escaped victimization’ and ‘overall rate of experienced victimization’. At the first follow-up, a significant reduction in the prevalence rate of experienced victimization was observed for the intervention group. However, based on the second follow-up data victims in the intervention group were not more likely to escape victimization. Further within-country analysis showed that UK students in the intervention group had overall reduced the number of victimization episodes compared to the control group at follow-up 1. The same results did not hold for the German sample, even though German students spent more time interacting with the software compared to the UK students. There are no adverse outcomes related to the intervention. The evidence suggests that bullying rates did not increase at follow-up 1 nor at follow-up 2.


  • Practice has not been evaluated in multiple populations

The practice was evaluated using two sub-samples – one sub-sample included students from Germany and the other one students from the UK. The two sub-samples were part of the same study. Overall the study has shown that there were significant program effects on escaped victimization. However, within-country analysis revealed that there is a positive significant effect for the UK sub-sample and insignificant program effects for the German sub-sample.

Within-country analyses revealed a significant short-term treatment effect on escaped victimization only among UK baseline victims (Table 2). This effect was not related to total time of interaction with the software or number of episodes watched.

Implementation costs

  • Implementation cost information is not available

Enduring impact

  • No follow-ups have been conducted

Program effects were measured one week and four weeks after the intervention ended. The four-week follow-up does not meet the enduring impact criteria which requires a follow-up of at least two-years.

Evaluation 1

Sapouna, M., Wolke, D., Vannini, N., Watson, S., Woods, S., Schneider, W., Enz, S., Hall, L., Paiva, A., André, E. and Dautenhahn, K., 2010. Virtual learning intervention to reduce bullying victimization in primary school: a controlled trial. Journal of Child Psychology and Psychiatry, 51(1), pp.104-112.

In total 27 schools (13 in the intervention and 15 in the control group) and 1178 students (555 in the intervention and 623 in the control group) between the ages 7 and 11 participated in the study. The attrition rate through the study was low. The response rate was high. Baseline questionnaires were returned by 94, 7%, while the first and second follow-up was completed by 91.9% and 92.7% of the initial participants. On average children in the intervention group watched 20.2 episodes and had 10.5 interaction sessions, or in total children in the intervention group have spent 51.6 minutes using the software. The implementation results suggest that students from the UK sample watched the same number of episodes as students of the German sub-sample but on average they’ve spent significantly  less time interacting with the software (46.3 vs. 57.7; t = –4.80; p < .001). The implementation analysis show that German teachers followed the manual instructions more thoroughly than UK teachers (24.70 vs. 23.08, F(1,21) = 6.128, p = .02) but on average they had more hesitant attitude for the program. 

Multileveled regressions were used to investigate changes in victimization and bullying risk rates one and four week after the intervention between comparison and treatment groups after it has been established that both groups have similar levels of bullying victimisation rates at the baseline. The study provides results for the combined sample of UK and German students as well as country –specific program effects.  First, in the combined sample with UK and German students it is found that the probability to escape victimization was significantly higher for the intervention group when measured at follow-up. However, the effect was not present at the second follow up. These short effects can be closely related to the duration of the program. The within country analysis showed that the beneficial effect was observed in the UK sample at the first follow up but not in the results of the German sample. Significant decrease in the overall victimization rates in the intervention group compared to the control group was found for the UK sample.  There were no program effects for the German subsample. The observed country specific effect can be attributed to the way teachers accepted and dealt with FearNot!. Potentially, the subsample differences can be explained by the age differences between the two samples.  The UK sample was composed of older children and older pupils are more likely to have higher reading and writing skills.   

At the same time, the study finds that the application of the software did not generate any adverse effects in the schools. 

Summary of Results for Evaluation 1












Treatment Group %


Group %


Adjusted Risk Ration (95% CI)





Outcomes with effect








Escaped victimization at follow-up 1


















 1.41 (1.02-1.81)*



1.90 (1.23 -2.57)*





Victimization prevention at follow-up 1




20.5 (48/234)




.60 (.36 -.93) *





Outcomes with no effect








Escaped victimization at follow-up 1










.96 (.58-1.37)







Escaped victimization at follow-up 2










53.8 (56/104)









50.4 (62/123)








 1.06 (.76-1.36)



1.10 (.66 -1.56)



1.02 (.63-1.33)





Victimization at follow-up 1







20.8 (91/438)





27.4 (127/463)




 .74 (.52-1.02)



1.02 (.61 -1.6)





Victimization at follow-up 2










20.5 (91/438)









21.4 (101/471)








 .74 (.52-1.02)



1.02 (.61 -1.6)



1.02 (.62-1.59)





Note: Values represents percentages of students (N/ total N)


Issues to consider

Assignment to control and intervention groups: The assignment was not random but selection into one of the groups is determined by the computer infrastructure of the schools. Such non- random design can be argued that it creates selection bias. However, the baseline comparison of observable characteristics does not suggest that there are statistically significant differences between the two groups.

Implementation challenges and possible deviations in implementations Since this is a computer based intervention technical problems can affect the implementation and directly influence the estimations related to the effectiveness of the program. The program relies on the schools having good computers where the software can run without crashing.  Since the program developers do not have knowledge of the computer limitations in each schools, they did not design the software on the school computers so it is likely that there are problems related to the implementation of the program.

Evaluation 2

Vannini, N., Enz, S., Sapouna, M., Wolke, D., Watson, S., Woods, S., Dautenhahn, K., Hall, L., Paiva, A., André, E. and Aylett, R., 2011. “FearNot!”: a computer-based anti-bullying-programme designed to foster peer intervention. European journal of psychology of education, 26(1), pp.21-44.

This second study complements the evaluation presented above by looking at additional outcome effects of Fear Not!. However, the second study did not meet EPIC’s evidence of effectiveness criteria. The second study evaluates if “FearNot!” has the potential to foster peer intervention, by looking at non-involved children and evaluating if they have been nominated as New Defender within the German sub-sample. The study found that there were more New Defenders in the intervention than in the control group at both assessments (one and four weeks after the intervention ended). 

Available resources

Preliminary results on the software prototype:  

 Hall, Lynne, et al. "Achieving empathic engagement through affective interaction with synthetic characters." International Conference on Affective Computing and Intelligent Interaction. Springer, Berlin, Heidelberg, 2005.

Practice materials:


Information on the EU FearNot! Project:



Sapouna, Maria, et al. "Virtual learning intervention to reduce bullying victimization in primary school: a controlled trial." Journal of Child Psychology and Psychiatry 51.1 (2010): 104-112.

Vannini, Natalie, et al. "“FearNot!”: a computer-based anti-bullying-programme designed to foster peer intervention." European journal of psychology of education 26.1 (2011): 21-44.

Contact information



Last updated

May 2018

Find a practice (within the 'Evidence based practices'):

The practices can be also searched along the three policy pillars of the Recommendation for Investing in Children:

Share this page