Virtual Bullying Virtual Environments as an Educational Tool

Virtual Bullying: Virtual Environments as an Educational Tool Sarah Woods University of Hertfordshire

What does VICTEC stand for? V: Virtual ICT: Information Communication Technology E: Empathic C: Characters

Who is involved in VICTEC? • European Project with partners in the U. K, Portugal and Germany. n 3 universities, 1 private research institute, 1 small company • The teams have expertise in: n Psychology, 3 D interactive graphics, Socially intelligent agents, Artificial Intelligence and business enterprise.

New Research: VICTEC Previous bullying interventions strategies have had limited success. Based on the research findings so far, the team hopes to design a new intervention strategy.

Inspired by theatre Use of performance to dramatise the problem n Theatre in Education – Linked workshops But Theatre is not a cheap option n In every group some will be bullying others n Interaction is limited in conventional performance

The aims of VICTEC - 1 • Implement a Virtual Forum Theatre n 3 D interactive graphics and virtual actors synthetic characters – Individual interaction based on creating empathy with the characters – Computer game technology - but NOT a game! n Provide a safe & engaging environment for children to explore different perspectives on bullying behaviour.

The aims of VICTEC - 2 • Evaluate the system – Does it affect children’s views on bullying behaviour? – Do children actually feel empathy with the characters in the scenarios? • Integration – How this fits into schools’ social and educational curriculum for issues around bullying and aggression

Synthetic Character Design • VICTEC is designing synthetic characters which have a digital or physical entity with personality that can interact with people in real time. n Computer games characters are not very intelligent n Do not remember anything about interaction n Hard to believe in them as personalities

How realistic should the characters be? We want the characters to be believable n Is believability the same as realism? n Most of us believe in Mickey Mouse… Characters that are NEARLY realistic may be ‘creepy’ n We have high expectations n And then they clash with them

Realistic vs. cartoon characters

Realistic vs. cartoon characters

Characters and Cultural differences The characters need to be country specific if children are to believe in them and have empathic relations with them. In the UK children wear uniform……

Characters and cultural differences In Portugal children don’t wear uniform!

Designing the school: Realistic or Cartoon?

Cartoon characters & school Based on questionnaire responses from children, there is a preference for cartoon characters and cartoon looking schools.

School Library and John the Victim

How will we create empathy? The characters must be believable n No re-run - what happens counts n Different scenarios and characters – Emerging story as in improvisational drama, not pre-scripted Child acting as an ‘invisible friend’. n Give advice about dealing with the situation and see what happens

How will we create empathy? The characters must have a personality n Different personality traits will be modelled for the bullying roles (bully, victim, bully/victim, bystander etc) n E. g. Bully will have more confidence than a victim. n E. g. Victim will be more anxious and fearful than a bully.

Agent Architecture & Personality Model Agent architecture must be able to: • Capacity agents to recognise other agents’ states by emotional expressions • Agents must be able to communicate with other agents • Agents must have emotions • Be able to express emotions by voice, facial/body expressions, • Respond to emotional states through coping strategies

Components of the Agent Architecture • Appraisal Component: based on OCC cognitive model of emotions using active pursuit and interest goals. • Action Selection Component: After appraising emotion state, agent must decide on best action • A coping Mechanism: problem focused coping and emotion focused coping • Affective Expression Component.

Creating believable bullying scenarios • A software package called Kar 2 ouche (www. kar 2 ouche. com/imed/) assisted the psychology team in designing the bullying scenarios.

Believable Bullying Scenarios • Scenario Introduction n The scenario will begin with an introduction to the main character (victim). n A tour of the school and classroom will then be given highlighting some of the schooling history and the other characters. n The first victimisation event will then occur. The victim will ask the user what he/she should do?

Bullying Scenarios (3) • The user gives advice n This will influence the character • Repeated - 5 or 6 episodes n Similar-but-not-identical incidents but maybe in different locations with slight difference in cast. • Child has chance to try out different sorts of advice

Bullying Scenarios (4) • Ending the session n n Give the child a summary story (still shots of key moments) from the episodes Use this to promote reflection: Why did this happen? How did each character feel? • Educational messages n Also at the end: ‘Don’t suffer in silence, tell somebody you trust’. Internet addresses and contact no. s will also be supplied.

Fear. Not the finished software, but how we think it is going to look……………

Fear. Not: The evaluation • Event for 400 children in June 2004 at University of Hertfordshire: n Social role-play with robots and agents n Questionnaire and interaction with the software n Activities with the robots and ICT software used in research

Challenges - 1 There is no easy answer n No strategy always works n Tension between realism and a straightforward message n Do we have to present a happy ending? • Producing believable characters and varying scenarios n Pushing the technology here n Especially language for unscripted interaction

Challenges - 2 Cultural differences n There is no word exactly equivalent to ‘bullying’ in German or Portuguese n School uniform and other differences n Where bullying takes place – No school dinners in Germany – Children unaccompanied to school

Contact Details • Ruth Aylett (project coordinator) nr. s. Aylett@salford. ac. uk • Sarah Woods (Psychology) ns. n. woods@herts. ac. uk n 01707 281133 • The project website nwww. victec. org n. Carried out with the support of the European Community in the Framework V Programme