Thanks, I think that Carolyn (Hoagies) was referring to the one study from the 80s on the OLSAT as well as anecdotal evidence such as we've found here when she wrote that. I mentioned that one OLSAT study in another recent thread here where this same issues with group tests was coming up:

Quote
Originally Posted by polarbear
I would also *love* to have whoever has mentioned here in the past a link re CogAT not being reliable as an indicator of true intellectual ability for HG/HG+ kids ... please please please share a link or cite the source. We were once in the very same situation - we're long past that at this point in my ds' school years, but I would so love to send that article back to his elementary gifted program teachers because I truly suspect there are quite a few kids missed for gifted programming in our school district every year due to CogAT.

We, too, are past the point where it matters for our kiddos, but I would also like to see those studies if they exist. I've never seen anything like that, but there sure have been a lot of people posting here over the years (and more as of late it seems) who have kids whose CogAT scores and WISC scores don't line up at all.

The only link I've found about group tests for gifted kids is really old and relates to the OLSAT, not the CogAT:

Quote:
WISC-R Full Scale IQ scores, OLSAT scores, and OLSAT School Ability Index were available for 431 of the students referred for placement in 1985-86. Subjects were ages 6-16; 283 were male, 148 female; 273 were white, 158 black. Correlations between the two tests were statistically significant for all but the gifted group.
[emphasis mine]
I'm not sure if this link will work, but this is a scanned copy of the entire original nearly 30 yr old study on the OLSAT vs. the WISC: http://www.eric.ed.gov/PDFS/ED286883.pdf

I don't know how familiar you all are with correlation coefficients, but essentially an "r" (correlation coefficient) of 0-.3 is considered a weak positive correlation between variables (as one goes up, so does the other, but the relationship isn't very strong or compelling), .3-.7 is a moderate correlation and .7-1 is a strong correlation, with virtually nothing having an r of 1.

For the average population (kids with average OLSAT and WISC scores), the r was strong. For the learning disabled and kids with low IQs, it was moderate, and for the gifted, it was weak positive or weak negative in the case of African-Americans gifted kids.

However, I am still looking for a study that looks at more recent norming versions of the WISC (this one used the WISC-R) or another individual IQ test vs. the CogAT and/or OLSAT and one that is not 30 yrs old.

Additionally, and this is just my own personal suspicion based on small amounts of data from other kids, I'd really like to see if high group test scores always correlate positively with high individual scores and not just the reverse. I.e. - I'd like to see if we are getting false positives off of tests like the CogAT as well as the types of false negatives that we hear of here.

Study Strategies for Accelerated Learners