I asked a little bit about this before, but I was wondering if anyone could give more in depth information. I have spoken with a few people and it seems if my dd's CSI score is lower than what it should be based on input I have been told. Can anyone weigh in or lead me to a site with more information on how to get an accurate score? I have been on the terra nova site and searched high and low, but nothing can give me what her CSI should be based on her scores. Her testing info is this:

Sequences :99*
Analogies: 81
Quantitative Reasoning: 92
Total Nonverbal Score: 97
Verbal Reasoning: 98
Verbal Reasoning Context: 85
Total Verbal Score: 97
Total Score: 98


Based on this information where can I find her CSI score? The school is giving me a number that seems too low after speaking with friends/colleagues from my former teaching days. They feel her score should be a minimum of 130 that typically a score of overall 95-96th percentile would fall at a 126 CSI (which is what they say my dd currently has) and that a 121 falls in the 91st percentile overall. I don't have her full report only one that lists the numbers at the top I gave. It has no other information. I just want to make sure they aren't shorting her b/c at a 130 all her other criteria she has met makes her STAR testing a nonissue on her acceptance into the program.

If the information I have been told is accurate I want to make sure they look into her test score and see if and error was made and where I can get concrete information on how the CSI is determined. I cannot find that anywhere that a certain percentile corresponds to a specific CSI score. I just don't see how she could have a 126 when the test cap is 141 and she is overall in the 98th percentile. Seems pretty low. TIA

InView has a mean of 100 and SD of 16. If you look up a bell curve/chart that has z scores corresponding to percentiles, you can easily find the range of z scores that could round to the relevant percentiles. Multiply the z score by the SD (16, in this case), and add (or subtract, if a negative z score) to the mean (100), and that generates your standard score. Once you top the 99th, of course, this won't work to spread the scores. In this case, I used the z scores for 96.5 to 97.4 %ile and 97.5 to 98.4 %ile:

97th %ile should be 129-131
98th %ile should be 131-134

However, schools are provided with multiple sets of norms. If your child attends a private school or high-performing public, your local norms may be stronger than the national norms, in which case it is quite possible that, though her national percentiles are as you've reported, her local or specialized (prep, private, Christian/ACSI) percentiles are lower (due to being compared to a generally higher-performing population).

ETA: and yes, I also get 126 for the 95th %ile, and 121 for the 91st %ile.

Thank you aeh. Is there anywhere online where I can pull this information as official to present prior to our appeal? I just want to make sure they aren't shorting her on the score and sadly the Principal didn't understand it enough to know if her CSI was accurate and was going to call the company. I expressed my concern about the accuracy and she told me if I could give her data showing why her score should be higher they would be place her in the program without an appeal.

She did call me back today to tell me the cap on this test is actually 138 not 141, but I'm not sure if that is even accurate with how unsure she is with this test.

sigep, the numbers eh is referring to are standard statistical calculutaions for a curve with a normal distribution, so the school can't really argue the calculations. You can google to find a picture of a curve with the distribution curve, percentiles and corresponding z scores etc. Here's an example:

http://www.gvglass.info/EDUC5716/Lesson-3/standard-scores.jpg

I took a quick look and guesstimated based on reading from this curve, and came up with a CSI of 127.4 for your dd - which is below 130 and close to the 126 the school report has, so I suspect the 126 is correct for this set of numbers (I was doing quite a bit of guesstimating when reading z-scores vs percentage so my # is likely off, but I don't think it's far off).

I suspect the reason your dd isn't scoring higher on CSI is due to her score on the Analogies, which is very different from her other score ranges. Has she had any other ability testing that might include a similar type of subtest that you could use to compare scores? It's possible that something was going on during this subtest that altered her score - she might have been distracted, it might have been the last test administered and she was tired, possibly the test is in error for some other reason such as a mistake in bubble sheet marks or timing... those possibilities are endless. It's also possible that analogies are a relative weakness for her - which doesn't mean she shouldn't be in a gifted program, but would be a possible reason that subtest score is lower than the others.

You mentioned on your other post that when you met with the gifted teacher she went through what you needed to do for the appeal - what type of suggestions did she offer?

polarbear

sigep, the numbers eh is referring to are standard statistical calculutaions for a curve with a normal distribution, so the school can't really argue the calculations. You can google to find a picture of a curve with the distribution curve, percentiles and corresponding z scores etc. Here's an example:

http://www.gvglass.info/EDUC5716/Lesson-3/standard-scores.jpg

I took a quick look and guesstimated based on reading from this curve, and came up with a CSI of 127.4 for your dd - which is below 130 and close to the 126 the school report has, so I suspect the 126 is correct for this set of numbers (I was doing quite a bit of guesstimating when reading z-scores vs percentage so my # is likely off, but I don't think it's far off).

I suspect the reason your dd isn't scoring higher on CSI is due to her score on the Analogies, which is very different from her other score ranges. Has she had any other ability testing that might include a similar type of subtest that you could use to compare scores? It's possible that something was going on during this subtest that altered her score - she might have been distracted, it might have been the last test administered and she was tired, possibly the test is in error for some other reason such as a mistake in bubble sheet marks or timing... those possibilities are endless. It's also possible that analogies are a relative weakness for her - which doesn't mean she shouldn't be in a gifted program, but would be a possible reason that subtest score is lower than the others. If you don't have any other tests that are similar to compare to, and you aren't successful with your first attempt at an appeal, I'd suggest requesting a similar test from the school district.

You mentioned on your other post that when you met with the gifted teacher she went through what you needed to do for the appeal - what type of suggestions did she offer?

polarbear

I just thought of another possibility, which is that the NPA (percentile by age) and NPG (percentile by grade) may be slightly different, depending on when in the year your child's birthdate falls. It is possible that the principal gave you the percentile for one, and the standard score for the other.

This ppt is one of the training documents used by the publishers (this one happens to be for ACSI):

https://safe.acsi.org/iweb/upload/PPT_InView-PTCS%205_17_Ppt.pdf

Unfortunately, I couldn't find a document directly from CTB that would provide the percentile-standard score correspondences. I calculated the standard scores based on my prior knowledge of statistics and easy access to percentile-z score tables.

One other thought. One tactic that parents in my school district have used when scores are close but not quite in range is to argue potential error in measurement. If you can find statistical error ranges for these scores it might help.

polarbear

pb, 126 would be correct if the SD were 15, like it is on most other measures, but in this case, it is 16. Although that does bring up the point that the principal may have been given the incorrect %ile-to-SS conversion table.

Thank you again for all the help. I just thought if I could show them that the CSI was off by a couple of points I could avoid the whole appeal process.

I think it is slightly alarming that who I was speaking with is unfamiliar with the testing and scoring. I wonder if they can show me other students scores who were in the same range as hers or with a CSI of 130 or higher(without their identifying information) to be sure hers falls within the same range. Is that something a school would do?

I also find it slightly frustrating that their only measure of acceptance is 2 tests/teacher recommendation and one of the test they will look at the last score, but if you look at how she does historically on the STAR test she would qualify hands down. Seems unfair, but they are already giving me the line if they bend the rules for one person than they have to for everyone, so not sure how likely she will get in. The good news is she is in for math and they will reevaluate in 2 years when she goes to 7th grade.

I do appreciate the information and help from everyone!

I am not an expert, but here are a couple more possibilities.

1. The school provided percentiles based on your district and your district is slightly lower than the national standards: For example, if 98 percentile in your district would only be 96 percentile nationally and thus 126 on the scale.

2. The school provided percentiles based on age (and your child is on the young side) but calculated the 126 based on grade.

Of course, it is entirely possible that the school doesn't know what they are doing at all. Like others have posted, the 16 SD means that 132 rather than 130 should be the magic number for this test.

Good luck an keep pushing.

Actually, #2 would be the other way around: the school provided percentiles based on grade, and a CSI based on age, and your child is old for grade, so the grade percentiles are higher.

sigep, I doubt you will be able to get information on other students, but there is a piece of information I think you should be able to get. I think the school most likely received a report on the Terra Nova with a bit more detailed info than what you've received. You should be able to request a full report from your school district. I can't remember the technical name to ask for (other than "full report"), but in our district, the name of that report is listed on our website where testing info is provided.

polarbear

My dd is older. I feel it is odd that they would penalize her for that. It isn't as if her grades in the class will be weighted based on her age. I would think they would be looking for 4th graders more than capable of doing the work of the class and that won't change because she is older than some in the class.

I have found several sample InView CSI reports online and the other odd thing is they all show a range of : Range 110-125 CSI 118
Range 77-109 CSI 93
Range 102-108 CSI 105


My dd's show Range 120-138 CSI 126

I would think her CSI would be 129, not 126. If her score is lower based on her age why do the example reports fall into the typical formula for finding a score based off of range. This test is very confusing!!

I asked yesterday and they said the only thing they had was what they referred to as the label. It had her scores for NPG and NPA and her CSI Range and score. Does that sound like it?