What is considered a "good" (or "bad") standardized test score?

[Jenny in GA]

I just got back the ITBS scores for my two oldest kids this year, and I was ... well, a little disappointed.

 

I was also a little surprised. One child's vocabulary score went down 44 percentile points from last year! How can that even be?? Last year she had a good vocabulary, and now she doesn't?? There was nothing about the test-taking that day that seemed "off" or bad, as far as I could tell.

 

And one child really loves science, reads and talks about it all the time, and I was surprised at how low her science score was.

 

So I'm not sure what to make of these test scores.

 

But one thing that bothers me about these standardized tests is that they don't actually tell what you what grade level they are on in anything, and it's hard for me to know what a "good" score is, or what "passing" (so to speak) is.

 

Obviously everybody's opinion is different, but does anybody know, for example, what schools tend to strive for? For example, is there a certain percentile that causes one to be evaluated for special ed? Is there a certain score that a student is usually desired to have before they are considered to have "successfully completed" that grade?

 

Are there states that expect homeschoolers to achieve certain scores, and it so, what are the expectations?

 

Is anything above 50 percentile considered "good" because it's technically above average? Or is "average" pretty terrible?

 

Do you personally have any specific expectations for these tests? For example, if you're child scored below a certain percentile in math, would that indicate to you that what you're doing isn't working well? What score would alarm you?

 

Thanks!

[Chris in VA]

Do you have the page that has grade level equivalents listed? They don't mean what some think they do--that is, they don't describe the level your child is on, exactly, but they do help sort out how well your kiddo did.

[justasque]

There should be a grade equivalent number in there somewhere. For example, 5.0 would mean that the child did as well on the test as an average 5th grader at the beginning of 5th grade (or something like that). And the raw scores should tell you how many of the questions in a particular content area the student got wrong. On the ITBS, in some areas there are only two or three questions, so the percentile score doesn't tell you as much - you have to look at it in the context of the number of questions right/wrong. Obviously, how well a child answers three questions is only a very rough assessment of how well he knows the subject in general.

[golfcartmama]

50% means average. What other children at the same level scored. It is not a failing score. I look more at ones that are below 50 and see what we need to work on---punctuation, for example.

 

I'm mailing mine today. I know they're going to be horrible, but the kids got the stomach flu and then a bad case of strep back to back, but I have to send them back so we did it anyway, it will be a total loss---waste of time and money. Oh well!

[Princess Peach]

I test every year, per my state hs'ing law. The child must have a composite score of 40th percentile or above in my state.

 

What I look for is basically the same scores from year to year, which would show a year of growth. Of course, I never balk at an increase! But if there was a drastic decrease in scores for a particular subtest, I would evaluate that child's curriculum for that subject. For instance, did I change curric for that year? Perhaps I need to beef up that particular subject area.

 

In the case of your child's vocab score, if you didn't change curriculum from last year to this year, and if you feel that she was taught in relatively the same way for that subject area, I would honestly wonder if she had messed up filling in the dots somehow.

[Crimson Wife]

The test report didn't give stanine scores? Typically 7th-9th stanines are considered good standardized test scores, 4th-6th average, and 1st-3rd considered low.

 

I personally look for standardized test scores that reflect the child's ability. I would consider any score <90th percentile for a bright child to be a "red flag" but the 50th percentile for a child with an LD might be a really good score.

 

With a big drop in percentile from year to year on one particular section without any obvious struggles in day-to-day work, I would wonder if my child made an error in bubbling the answer sheet. It's easy to get off by a row and then have a misleadingly low score.

[Carrie12345]

*For the most part*, I don't consider any score to mean anything. The example you gave, with comparing one year to another, would give me pause, but I'd definitely suspect some sort of scoring/marking error.

 

Generally, perspective is everything. I always scored in the high percentiles, and so did ds. When dd hit a 70-something in one area, I freaked for a second, then realized that was much better than average, particularly for a kid who doesn't use traditional standards or have any test prep!

 

If I wanted my kids to master ps content, I'd probably send them to ps. ;-)

[OH_Homeschooler]

I just got back the ITBS scores for my two oldest kids this year, and I was ... well, a little disappointed.

 

I was also a little surprised. One child's vocabulary score went down 44 percentile points from last year! How can that even be?? Last year she had a good vocabulary, and now she doesn't?? There was nothing about the test-taking that day that seemed "off" or bad, as far as I could tell.

 

One possible interpretation of this is that last year her peers were not quite up to her level but this year they caught up. Your DD may be right on schedule (i.e., did not "lose" anything), and it just took her peers the extra year to catch up to her.

 

Percentiles are just rankings. If your child scores at the 50th percentile, about 50 of the students scored below and 50 scored above. Typically this is based on normative data collected over several years and not the same exact sample your daughter is part of. That doesn't mean the scores are incorrect, usually testing organizations have copious amounts of data and these are considered valid.

 

Kids have strengths and weaknesses and good days and bad days. Tests are a good basic indicator of some skills but I would not put all my eggs in one basket. You can use test results to guide your instruction, however. If you feel perhaps your daughter could be falling behind in terms of vocab, then boost her lessons in this. It's a process.

[abbeyej]

25th-75th percentiles are "average" -- low-average to high-averge, but still within the range of "average". Above 75th and below 25th are considered a little outside of average. With a child scoring below the 25th, I would begin to ask questions... Is the child getting consistent instruction? Are there learning challenges or physical challenges (vision or hearing difficulties, for instance) that haven't been immediately obvious? Are there environmental concerns (child is malnourished or chronically sleep-deprived)? Were there testing day difficulties (illness, major distractions)? ... For some kids, the 10th percentile is a solid representation of abilities. For many though, a low score indicates that *something* else is going on.

 

On the other hand, even a "good" score could indicate one of the above problems if you believe that that score doesn't actually represent a particular child's capability. Another sign that something (of many different possible sorts) is wrong is a wide variation in sub scores. A kid who's getting really significantly different scores (I've known kids to get a 99th and a 4th percentile on different sections of the same test) might end up with a "good" overall score but still have some pretty hefty learning differences going on...

 

A "good" score will depend on the kid. I know kids for whom a 24th percentile is a *phenomenal* achievement! ... Generally speaking though, above 75th is "good", above 90th is "very good", and above 95th is excellent. (Consistently scoring in the 98th-99th probably means that the test isn't a very good measure for a particular kid. It's just not telling you anything about that student's strengths or weaknesses since s/he's clearly beyond the level the test can assess.)

[LucyStoner]

 

I personally look for standardized test scores that reflect the child's ability. I would consider any score <90th percentile for a bright child to be a "red flag" but the 50th percentile for a child with an LD might be a really good score.

 

With a big drop in percentile from year to year on one particular section without any obvious struggles in day-to-day work, I would wonder if my child made an error in bubbling the answer sheet. It's easy to get off by a row and then have a misleadingly low score.

 

:iagree:

 

If my son who generally tests in the 98th and 99th percentiles in most sub sets gets a 75th percentile on his same age and grade level, I might be concerned that we were not keeping him up to his pace or that he did something strange on the test form. If my sibling who dislikes school, does not read for pleasure, struggles with tests and school, and is not heading to college gets a 50th percentile, it would be time to celebrate. It is about what the child can do.

 

I agree that you want to see the test form or ask your child about the test. I know one mom of a gifted daughter who was called in for a special ed meeting because her child had totally bombed the standardized test. As it turns out, the girl had been told that as soon as you finished you could read and she was bored with the test and excited about her books so in her 7 year old logic, filling in all the bubbles without reading the questions was the best course of action. :tongue_smilie:

Edited June 1, 2012 by kijipt

[Jenny in GA]

I test every year, per my state hs'ing law. The child must have a composite score of 40th percentile or above in my state.

 

 

Just out of curiosity, what happens if they don't?? Are they legally ordered to go to public school?

[Princess Peach]

Just out of curiosity, what happens if they don't?? Are they legally ordered to go to public school?

 

The homeschooling program is put on probation and if sufficient progress is not made after one year, they may not be allowed to continue. There would probably be a hearing. I'm not really sure beyond that. Thankfully, I've never had to worry about it.

[ThisIsTheDay]

I just got back the ITBS scores for my two oldest kids this year, and I was ... well, a little disappointed.

 

I was also a little surprised. One child's vocabulary score went down 44 percentile points from last year! How can that even be?? Last year she had a good vocabulary, and now she doesn't?? There was nothing about the test-taking that day that seemed "off" or bad, as far as I could tell. If she was thrown off and did not know the first word this year, it could have undermined her confidence and blown the test for her. Or she could have answered in the wrong circles, so her answers would be totally off.

 

And one child really loves science, reads and talks about it all the time, and I was surprised at how low her science score was. Homeschoolers typically do not study science in the same sequence as taught in ps. For this reason, I would only focus on testing the majors, spelling, reading, writing, math.

 

So I'm not sure what to make of these test scores.

 

But one thing that bothers me about these standardized tests is that they don't actually tell what you what grade level they are on in anything, and it's hard for me to know what a "good" score is, or what "passing" (so to speak) is.

 

Obviously everybody's opinion is different, but does anybody know, for example, what schools tend to strive for? I personally would want my public school to be testing in the 80th percentile or higher. So much is taught to the test, and I would hold them to high standards. We typically have lived in better school districts because of this expectation. For example, is there a certain percentile that causes one to be evaluated for special ed? Is there a certain score that a student is usually desired to have before they are considered to have "successfully completed" that grade?

 

Are there states that expect homeschoolers to achieve certain scores, and it so, what are the expectations? My state law requires my kids to test higher than the third stanine; if the overall score is third stanine or below, the homeschool family may be required to have interference by the school district.

 

Is anything above 50 percentile considered "good" because it's technically above average? Or is "average" pretty terrible? That totally depends on your child and your circumstances. I know where my kids' strengths are as well as their weaknesses. I know what results I can expect to see. There is no one answer to your question here.

 

Do you personally have any specific expectations for these tests? For example, if you're child scored below a certain percentile in math, would that indicate to you that what you're doing isn't working well? What score would alarm you? My first reaction would likely be to retest. I would consider either using the same test (ITBS) or switching. If you are overseeing their work during the year, and you know they have kept pretty much on schedule since their last testing, you have a much better idea of where they are than what results of any test will tell you. If previous testing scores were consistent, I would not be alarmed by a sudden drop. Unless I had a(nother) reason to suspect my curriculum, I'd consider many other possibilities for the test scores instead: Mixup in answer sheets either at time of testing or at grading; an off testing day; test day nerves or other unpreparedness; deliberate sabotage by student.

 

Thanks!

.

[kathkath]

The test report didn't give stanine scores?

 

I personally look for standardized test scores that reflect the child's ability. I would consider any score <90th percentile for a bright child to be a "red flag" but the 50th percentile for a child with an LD might be a really good score.

 

With a big drop in percentile from year to year on one particular section without any obvious struggles in day-to-day work, I would wonder if my child made an error in bubbling the answer sheet. It's easy to get off by a row and then have a misleadingly low score.

 

 

I agree.

I would expect my child's stanines to be consistent with iq over time. If not there is a problem needing addressing. http://www.mathnstuff.com/math/spoken/here/2class/90/stanine.htm

[EKS]

Please keep in mind that what *you* know about your children's achievement day to day is far more accurate and important than what any test might say.

 

That said...

 

This is just my opinion, which is based on years of testing and knowing where my kids are in terms of mastery of material:

 

Scoring at the 50th percentile does not indicate mastery of grade level material.

Scoring at the 75th percentile indicates partial mastery.

Scoring above the 90th percentile indicates mastery.

 

Did you give the tests yourself? If not, it's hard to know what went wrong. A drop of 44 percentile points is significant. This happened when my son (who has dyslexia and ADHD) took the ITBS under standard conditions in a classroom with other students for the first time (he had previously taken it with me under standard conditions, but he was by himself). The next year he got appropriate accommodations and his scores went up dramatically.

 

If you did give the tests yourself, or decide to in the future, it is really helpful to look through the tests before you send the back to see what specifically your children are having trouble with. It can be very reassuring to see for yourself that the science and social studies questions are really dumb, for example.

 

Also, there is a big change in format between the 2nd and 3rd grade tests--filling in bubble answer sheets instead of marking answers in the test booklet--and this can make a child's scores drop.

 

Did your children finish the test? If not, that can have a huge impact on scores.

[Jenny in GA]

This is just my opinion, which is based on years of testing and knowing where my kids are in terms of mastery of material:

 

Scoring at the 50th percentile does not indicate mastery of grade level material.

Scoring at the 75th percentile indicates partial mastery.

Scoring above the 90th percentile indicates mastery.

 

Ouch.

 

If this is true, then maybe homeschooling is not going as successfully for us as I'd hoped.

 

In answer to your other questions: Yes, I gave the tests myself, and did last year as well. There were several sections of the test that each child did not quite finish.

[EKS]

There were several sections of the test that each child did not quite finish.

 

If your children weren't finishing sections then the scores are not giving a true picture of what they know. You just know they are working more slowly than their peers. One thing you could do is give the Stanford (which is untimed) to see if they do better (the PASS test is also untimed). The other thing would be to start using a timer for certain tasks during your homeschooling day to get them used to being timed. A visual timer works particularly well.

[Laurie4b]

I just got back the ITBS scores for my two oldest kids this year, and I was ... well, a little disappointed.

 

I was also a little surprised. One child's vocabulary score went down 44 percentile points from last year! How can that even be?? Last year she had a good vocabulary, and now she doesn't?? There was nothing about the test-taking that day that seemed "off" or bad, as far as I could tell.

 

And one child really loves science, reads and talks about it all the time, and I was surprised at how low her science score was.

 

So I'm not sure what to make of these test scores.

 

But one thing that bothers me about these standardized tests is that they don't actually tell what you what grade level they are on in anything, and it's hard for me to know what a "good" score is, or what "passing" (so to speak) is.

 

Obviously everybody's opinion is different, but does anybody know, for example, what schools tend to strive for? For example, is there a certain percentile that causes one to be evaluated for special ed? Is there a certain score that a student is usually desired to have before they are considered to have "successfully completed" that grade?

 

Are there states that expect homeschoolers to achieve certain scores, and it so, what are the expectations?

 

Is anything above 50 percentile considered "good" because it's technically above average? Or is "average" pretty terrible?

 

Do you personally have any specific expectations for these tests? For example, if you're child scored below a certain percentile in math, would that indicate to you that what you're doing isn't working well? What score would alarm you?

 

Thanks!

 

Average isn't just 50%; average is a wide range--something like 25% -75%. (don't recall exactly) Average is average, not bad.

 

The tests designed to be administered in groups like the ITBS are not designed to tell you any statistically significant information for a specific child. Surprising, right? What are they designed for then? To compare groups of children to other groups of children: School A to School B, or School A in 2011 to School A in 2012. For individual children, the only statistically valid use is as a screening mechanism. Very low scores warrant further individualized testing to determine if there are learning disabilities. Very high score would warrant further exploration to see if the child is gifted.

 

Grade equivalents are pretty meaningless if they come out above grade level because only grade level material is on the test. So a 3rd grader scoring grade 5.2 on a subtest means that that 3rd grader scored the same as the average kid in the 2nd month of 5th grade taking the same test of 3rd grade items. In other words, your 3rd grader really does well on 3rd grade items and makes few if any mistakes on them.

 

If you want a pretty good grade equivalent and statistically valid info on your individual child to compare from one year to the next, you would need to get an achievement test designed to get that data. Those are typically administered 1 on 1 by a professional with a child. The data is statistically valid to say something about that child, there a confidence range to tell you the probability that an apparent improvement or regression is due to chance, and you get the benefit of the test administrator's observations. (For instance, you might find that the math score was lower than expected because your child went through and did every problem as if it were addition, ignoring the subtraction signs. Or you might find that the administrator suspects an LD based on specific observations, etc.) What you don't know with the group administered bubble tests is why your student did better or worse. Did s/he skip one question, not notice, and keep filling in bubbles? Did she not understand instructions? Is is just a random occurrence? With a person sitting there interacting with your student, you will know.

 

The grade equivalent is more realistic because the test items administered don't end until the child has missed a certain number in a row and the material is progressively more difficult, usually going up to college level. ( I once tested a 1st grader who got a math question about the circumference of a circle based on the radius correct. He ended up with like an 11th grade score in math. It was kinda scary! But that was really about the level of work he was doing. ) Anyway, these typically cost $50-$75.

Edited June 2, 2012 by Laurie4b

[In the Rain]

I would assume that your child made an error during the bubbling process to end up with a 44%ile difference?

 

Did you look over the tests? I always do. It is very helpful to me to see what they missed. Was it carelessness? Something not covered? Covered but not understood? This is especially important to me on the math test.

 

One one test my dd missed all of the money problems. She knows how to use actual money during transactions at stores, so I would have been shocked to know she failed the money section. It turned out that all the money problems had 50 cent pieces. LOL. She had never seen one before.

 

Do you personally have any specific expectations for these tests? For example, if you're child scored below a certain percentile in math, would that indicate to you that what you're doing isn't working well? What score would alarm you?

 

Thanks!

 

I would expect the percentile to remain about the same. I have used the PASS test for two consecutive years. I expected to see at least one year of growth.

 

I really don't know that any particular score would alarm me. I assume that kids with really low percentiles are either performing at pretty low levels at home (so their parents are aware) or they are bad test takers. I would be much more concerned if I looked over a test and saw my child missing questions on material she should have known. This happened to one of my dds. She had lots of careless errors, and saw a 10% drop in her score. We discussed the impact of carelessness, and her next score was about 15% higher.

[magnificent_baby]

I have no idea if this will be helpful to you, but wanted to share.

 

My cousin is a elementary school principal, and according to her, do not be alarmed if you see ITBS test scores drop this year as compared to previous years. She says they have made the test harder, and therefore the bell curve is now wider and fewer kids falling into the 90th percentiles.

 

This got me really curious, as dd took ITBS for the first time this year. I called BJU press, and they did some checking for me. The state we live requires homeschoolers to get the same ITBS test that ps kids get (this differs from state to state). Also, according to BJU, the test has not changed since 2004 (or something like that, can't remember the exact year), and the bell curve is also the same.

 

So, I have no idea who is correct, but wanted to give you some food for thought, maybe this would explain the drop in scores for you.