I'll order photocopies of those reports from Jefferson Medical College and check them out. Even the abstracts accompanying the references seem to bear out our contention that methylphenidate increases the ability of a child to maintain interest-excitement, while anxious children would not be helped. The latter would be expected simply because while i-e is a response to an optimal gradient of neural firing, fear-terror (the affect underlying anxiety) is triggered by a higher-than-optimal gradient of neural firing. Methyphenidate would then be expected to push these latter chldren to a gradient that would not help them to concentrate on novel material.
A problem of variable significance in these papers is that most researchers use emotion labels with no regard to the underlying affect. Most of the time, what researchers call "anxiety" turns out to be distress-anguish, an affective response to steady-state overmuch. It would simplify things so much were scientists to use the Ekman and Friesen manual of facial affect coding, or at least to photograph the subjects so we knew what they looked like when the researchers said they were anxious or whatever.
Again, thanks for the references.