The four children in the first experiment were observed playing with the unaltered stuffed Tigger (see 2.1). From these observations, a prototype Affective Tigger was constructed. This prototype had animated ears that make him `look' happy when he was bounced, and `sad' when the child bend his neck forward.
The prototype Affective Tigger was then given to three teenaged children, and they were asked to identify his expressions, both with and without voice. The teenagers were eager to offer suggestions and comments on how to improve upon the prototype. They made it clear that more features should be included in the next version of the Affective Tigger. Going back to the observations from the first experiment, three more behaviors were selected (tilt, abuse, and tail pull) to be measured. The sensors were carefully chosen and positioned in the second Affective Tigger to measure these behaviors and to heighten the illusion that the Affective Tigger was responding to the child. Acquisition of this feedback was one of the goals of the pilot study.
After the redesign of the Affective Tigger, the evaluation of his success at achieving responsiveness was anecdotally corroborated when the group of 18 third grade children were presented with two Furby's and the second version of the Affective Tigger. The teachers and adults observing the interaction commented that even under these conditions, the Affective Tigger appeared to be reponsive. The children who figured out how to hurt or comfort the Affective Tigger seemed the most excited about playing with him.
In general, the pilot study provided the necessary feedback of which behaviors to measure for the redesign of the Affective Tigger. It also demonstrated that teenage children could recognize, more-so with the inclusion of voice than without, the two moods of the Affective Tigger - happy and sad.