The process of constructing an emotionally reactive stuffed animal has proven itself to be helpful in furthering our understanding of the development of social communication in children. The use of a computer controlled toy was a natural and logical means for beginning to look at `child-proofing' computer interfaces while maintaining the sense of `child orientedness' inherent in a toy.
The Affective Tigger toy, overall, was well received. A few people said they didn't like the toy after they'd seen it, or didn't think their children would want to play with it, but the overwhelming response was positive. Of all the people who have seen and played with the Affective Tigger, well over two-thirds said, either it was a good toy, or at least the idea of an emotionally reactive toy was promising.
The interface issues were addressed individually and, in the end, provided the combined effect of a passive stuffed animal, which was the initial goal. Such issues as size, character, expressive features, number of moods, and the design of the sensory system, were resolved to the point that they did not intrude on the child-computer interaction. Other issues like weight, and degree of expressivity, still need improvement.
The biggest factor in determining if the Affective Tigger was successfully constructed was the question: does the child recognize the expressions of the Affective Tigger? The answer to this question lies in the behaviors expressed by the child. Is she mimicking his expressions, or is she able to vocalize her recognition of his expressions?
As discussed previously, the perception, and even vocalization of an emotion can emerge long before the ability to internalize this information to guide one's actions [Piaget, 1981]. This ability of young children to say what they don't mean, and to vocalize as well as imitate what they don't understand is the largest obstacle to this study. The emergence of mimicry as early as 4 months (see section on mimicry, Emotional recognition in the very young infant and the following section on Emotional expression in the infant), suggests that it is not in itself a reliable measure of a child's recognition of an expression. However, combining vocalization with mimicry, we begin to be able to put more faith in the child's ability to recognize the emotion being displayed. The criterion for the independent assessment of the child's ability to recognize the Affective Tigger's expressions included: demonstrations of empathy and vocalizations in addition to mimicry.
There was a significant correlation of age with perceived recognition, and an interesting correlation between the children who have siblings and increased performance with emotional recognition. The correlation of age with recognition was expected, and it verifies that the Affective Tigger was able to project an emotion which was recognizable to the child. Additionally, the parents often commented on the expressions of the Affective Tigger in ways that demonstrated that they could identify the emotion being conveyed in both his posture and voice.
The second question regarding the success of the Affective Tigger was: does the Affective Tigger make clear the causal relationship between the actions of the child and his expressions? This was the Affective Tigger's greatest failing. Many of the parents, 45%, commented that his expressions were not dramatic enough. Additionally, 72% of the parents, and most of the other people who saw and played with the Affective Tigger, said that his voice needed to be louder. These two factors were a huge detriment to the Affective Tigger's ability to create the illusion of reactivity.
Overall the Affective Tigger, as a toy, was a partial success. With more time and resources, the obstacles of weight, voice loudness, and the subtlety of his expressions could all be overcome. The fact that 58% of the children spontaneously recognized his expressions, - happy vs. sad - despite these failings, is very encouraging.
