>>10993177Again, cbf
Seventh, another reason for excluding data from one
source (Ohuche & Ohuche, 1973) was that in this study, the
5–6 year olds showed an average IQ of 69.5, while the 7–12
year olds showed an average IQ of 94.2. According to Lynn
and Meisenberg this discrepancy indicated “serious problems
with the data”. However, a study involving Sudanese children
(Fahmy, 1964) that was characterized by equally serious
problems was included by Lynn and Meisenberg. Specifically,
these Sudanese children lacked experience in using a pencil
and averaged an IQ of 50 on the Draw a Man test. These same
children averaged an IQ of 94 on a test that did not involve
the use of a pencil. Lynn and Meisenberg did not consider
the discrepancy between subtests' IQs problematic, and
took the DAM scores into account in their mean estimate
of this sample (average IQ of 69 according to Lynn and
Meisenberg).
Well to start with, I recall there being a reason to reject data if it compares early childhood results with later childhood or on results, I can't remember exactly why, so take this with a grain of salt.
Second, in regards to the "Draw the Man" test comments above, what was the other test used?
Third, of course they wouldn't care about a discrepancy in the subtests, there's always a discrepancy within the subtests (admitedly not that large for most) but still, you would need to show what the other scores for the other subtests were in order to claim it would be unreasonable to use that data.
