next up previous contents
Next: Evaluation Up: Summary of Results and Previous: Summary of Results and

Error Analysis

As with any behavioral study, the risk of error introduction is high. Some potential sources of error could come from: the experimenter or accompanying adult's `coaching' the subject, the failure of the Affective Tigger's hardware, the inattention of the child to the Affective Tigger, and the inaccurate assessment of the child by the independent observer.

In fact, many of the trials were plagued with problems. As mentioned previously, the Affective Tigger's voice was too soft for the child to make out what he was saying. In two of the trials. the experimenter had to tell the child what he was saying before they could understand it. Additionally, his ears moved in a very subtle manner which was often hard to see amidst the bouncing and thrashing about. Four of the twelve children failed to notice that his ears were moving at all. In one case the hardware failed altogether and the Affective Tigger had no responses, however this didn't seem to bother the child and she was quite happy to demonstrate what he should look like when he is happy or sad. The two boys liked the `distracter' toy so much they refused to play with the Affective Tigger for more than a few moments each.

In general, future work should provide a more standardized setting for the experiments to occur, and the toy should function exactly the same for each child, rather than the approach here of modifying it slightly after each trial.

Dana L Kirsch
Tue May 25 08:59:22 EDT 1999