That is correct; GE and AE are not accurate on most tests, including the WJ. The RPIs are useful for indicating that she is well above mastery level compared to her age peers (or grade peers, if it was scored using grade norms). Finally, WJ achievement scores can easily be lower than face expectations based on the WISC, due to regression to the mean. Did the tester actually look up predicted achievement/discrepancy tables for the WISC/WJ, or did they use a ballpark/face validity approach? I don't have tables for the WISC-IV, but the WISC-III/WJIII predicted achievement tables have an FSIQ of 140 predicting WJIII ACH scores in the 120s and low 130s. (There's a statistical method for deriving these numbers for the WISC-IV, but in the interests of time, we'll just use the WISC-III predictions as an example.)
From a placement standpoint, RPI is the only score on the WJ that directly addresses placement, although it doesn't go very far above the mean.
