I received the full report, and I'm perplexed. It's been a long time since I studied statistics, and I know little to nothing about intelligence tests.

I have T-Scores, Percentiles, and Age-Equivalency for several subtests.

First, both kids had some T-Scores of 90. DAS-II's proprietor states that T-Scores range to 90 on the School Age version. Does a T-Score of 90 actually mean that they hit the ceiling on the subtest?

Second, I'm thoroughly perplexed about the results on my son's Matrices subtest: Age-Equivalence = 15 years, 9 months; T-Score= 75; Percentile=99. My confusion results because other subtests had even younger age equivalences, like 13 years+ or 14 years+, but each of these had much higher T-Scores and much higher percentiles (well, 99.9+ as opposed to 99.)

How can his age equivalence be appreciably larger on the Matrices than on other subtests, yet produce a markedly lower T-Score and percentile?

Thank you for any insight!

T-Score is just their alternate way of showing standard deviations (avg 50, 10 per SD; vs. IQ that is ~avg100 with 15 per SD.) Since they only report out to 4 standard deviations, then I think you'd call that hititng the ceiling with a 90. That just may be the max reliability they have.

For the age equivalencies... some abilities change less as kids get older or have a mean that plateaus after a certain age (e.g. average of 24 for 8 yr, 26 for 9 yr, 28 for 10 yr, 28 for 11 yr, 28 for 12 yr; would be a plateau at a score of 28 and would report as equiv age 10+) Matrices could be something that some kids mature into quicker giving a less bell-like curve at younger ages.

Hope that helps some; I'm not specifically familiar with that test, but with testing/stats in general.

Thank you. I appreciate you sharing your testing/stats expertise. It's interesting to think of the difference between a ceiling v. reliability.

I'm a little relieved I didn't have 50 responses, saying "of course, T-scores..."