Just for some perspective:
The WMI drop is only a difference of six standard score points, well within the standard error range (confidence interval). I wouldn't even give that one a second thought.
The PSI drop is 14 points, which is a little more notable.
KADmom is correct, both indices are more vulnerable to anything that affects attention, including time of day, what/when your last meal was, medication, having a good night's sleep. Another factor for PSI is the age of the child. If the first administration was at age 6 or 7, he had one version of each PSI subtest. If he was then 8+ for the second administration, he had different versions. The slight alteration in tasks may be enough to account for a little variation in performance (it's not supposed to, but GT kids, as we know, don't always respond the same way NT kids do).
Neither drop is dramatic. If you don't see any clinical or functional indicators of concern (that is, IRL), then I wouldn't be too concerned about the PSI sag.
