Studies show that children with high levels of anxiety tend to pay more attention to threatening stimulus in the environment. They tend to attend to these stimuli and have difficulty to disengage from them. These attention biases enhance and maintain the level of anxiety. The aim of this study is to test a method of therapeutically intervention which focuses on shifting these attentional biases, with the use of a computer game which was designed to train the child to focus his/her attention away from the threatening stimuli and to focus on natural or positive stimuli. We will recruit 160 children with ongoing anxiety disorders who seek treatment. We will first assess threat-related attention bias and anxiety symptoms using structured psychiatric interviews and questionnaires. We will then randomly assign these children to one of four conditions: training to avoid threatening stimuli and attend to neutral stimuli; control placebo-training for threat-neutral stimuli; training to attend to positive stimuli and avoid neutral stimuli; and, control placebo-training for positive-neutral stimuli. Upon completion of training we will again assess attention bias and anxiety.

Two sets of predictions will be tested, one set concerns the effects of training on attention, and the other concerns the effects of training on anxiety. In terms of training effects on attention, we hypothesize that children with anxiety disorders can be trained to either avoid threat or attend to positive stimuli. In terms of training effects on anxiety symptoms, we hypothesize that the experimental training sessions will produce greater reduction in symptoms of anxiety than the placebo-control training sessions.