Gifted Issues Discussion homepage Forums Identification, Testing & Assessment Obscure SCAT scoring question

I finally got my very own SCAT score interpretation guide to have and to hold, and am I ever disappointed! There is almost no data in there!

So, here are my questions for Dottie, and other test experts:

1) How do they derive the scaled scores from the raw scores?
2) Would a given raw score translate into the same scaled score for a student regardless of whether they are in 4th grade or 5th grade?
3) Why do the top scaled scores vary from year to year?

It appears to me that this year's top Verbal scaled scores for 4th/5th are several points lower than the top 2008-2009 scores listed in the guide (I'm figuring a 6 point difference, which seems pretty large). How am I getting this? When I look at the bar charts, presumably, the top score for which there is actually a bar is likely the scaled score for a raw score of 50/50 (although I suppose that t could be 48 or 49/50). The next score down, would presumably be 49/50 and so on. Conversely, if your kid scores, say, 47/50, there should only be three possible scores that are higher. Right? The way I'm figuring it, the top scaled score for 4th/5th (if raw scores translate the same for both grades) would be 481 this year (in contrast to a 487 last year, at least for grade 5).

Why do I care? Well if you're supposed to be able to use the bar charts to compare your child with the talent search pool, it doesn't really work that well if the scaled scores aren't exactly comparable.

Not that is really matters, but like many of you, I like my data.

Hi NJMom,
I'm not an expert, but I think I may have your answer, since I was also comparing the last two years SCAT interpretation. They note that the SCAT interpretation guide sent out this year is based on test taken between Sept 2008-March 2009. The interpretation guide from last year is based on actual test data between Sept 2007-March 2008 etc.

I liked that numerical table that they had in last year's SCAT guide, which allowed me to calculate how my DDs did percentile wise vs. other gifted grade peers who took the test. Now we are left to interpret the bar graph....

I only have the interpretation guide for this year, so I don't have a comparison to see if the ceilings have moved before, but it seems to me that the ceiling score for verbal MUST be different this year. Extrapolating from DD's raw and scaled scores, I don't see how the scaled scores for 4th/5th could possibly go up to 487.

I guess we have to wait until NEXT year, to see the actual data for how our kids did THIS year.

mominsac and kcab, what is the numerical table that you were referring to? Is it like the one this year that they give for the STB test? It certainly would be nice to have a table giving the percentile within the talent search pool.

Honestly, I'm not reading a whole lot into these scores. It's a VERY limited test. It would just be nice to get an accurate percentile. And I would have liked to be able to see what scaled score and percentile data DD would have gotten if she were grade accelerated. But I don't see how I can figure it out from the data in the guide.

Hi again, I have chance to now sit down and compare the two year's data guide

Yes, last year, they provided us the numerical table (like the STB table) for all the SCAT scores, divided by grade and gender. I don't know how the raw scores translate to scaled score, so its not clear whether that translation formula is different for 4 vs 5th grader, but as Dottie, says, I can suppose that the translation formula can differ from one version to other (but I would think the 4th and 5th will take the same version the same year.)

REgarding the top scaled score, the hypothetical top score seems to be the same year to year, comparing last year vs this year (for 4th/5th, it is 487V/506Q.), but the number of students hitting it or any of the scores will depend year by year.

And yes, I took my daughters' score from LAST year, and put it against THIS years frequency data to see how she really fared amongst her peer group that took the same grade test the same year. I actually saw some difference from last year's data in terms of calculating their "percentiles" this way. I suppose if you did this many times over though, it will even itself out, such as how its suppose to work statistically.

Another data-driven mom, lol.