benchmark scores different between dibels/aims

Discussion in 'Special Education' started by waterfall, Mar 30, 2012.

  1. waterfall

    waterfall Phenom

    Joined:
    Feb 5, 2011
    Messages:
    4,974
    Likes Received:
    98

    Mar 30, 2012

    I am wondering if someone can clear this up for me. Our district uses aimsweb for progress monitoring of sped/RtI students, as well as benchmarking for gen ed students. They consider the benchmark score that everyone should get to be the 50th percentile. We typically don't use dibels, but we have an intervention teacher this year who started using it instead of the r-cbm for fluency. The problem we're running into is that the dibels benchmarks are different than the aims benchmarks- aims are higher. So although what she's using as the "benchmark" for dibels is correct, in some cases this is significantly lower than the aims benchmark and what we as a school/district consider to be "grade level." Are the dibels words harder? Why is there a difference? Can we say that a student who met the dibels benchmark is on grade level, even if this is 10-20 words off the grade level aims benchmark? For example, in first grade the aims benchmark is 67 wpm, while dibels is only 54. The discrepancy only grows as students get older- by 5th grade, the dibels benchmark is more than 20 wpm lower than aims.

    As we're looking at trend lines for RtI students, this becomes an issue because it looks like these kids aren't really too far behind grade level, when in fact they may be 20-30 words off our "real" benchmark level. So, would you change your dibels goal to match the aims benchmark? Would that invalidate the test? Or just know that dibels is a different assessment, and as long as they're on grade level for that it's okay?
     
  2.  
  3. mopar

    mopar Multitudinous

    Joined:
    Aug 15, 2010
    Messages:
    10,953
    Likes Received:
    0

    Mar 30, 2012

    Dibels has done their norming on their passages. So I would just look at it as being on grade level is okay. However, if you notice that the students are testing at grade level on dibels but not on aims, then I would look further into the issue. Yet, if students are testing in about the same percentile on each test, don't worry about it.
     
  4. bros

    bros Phenom

    Joined:
    Aug 5, 2009
    Messages:
    4,046
    Likes Received:
    36

    Mar 31, 2012

    I was basically going to post what mopar did, also perhaps your school should make sure that all teachers use the same system, so the IEPs/benchmark data are uniform
     
  5. mopar

    mopar Multitudinous

    Joined:
    Aug 15, 2010
    Messages:
    10,953
    Likes Received:
    0

    Mar 31, 2012

    bros---I think waterfall has been trying to get everyone on the same page with progress monitoring...but this one teacher wants to do things differently.
     
  6. 1st-yr-teacher

    1st-yr-teacher Comrade

    Joined:
    Jun 3, 2006
    Messages:
    471
    Likes Received:
    0

    Mar 31, 2012

    We use DIBELs. From what I have seen, the passages do vary in difficulty. I have students who go up and down and then up and down again. One student read 39 wcpm one week andthe next week, she read 70wcpm! This student stays around 37 to 43 wpm in her weekly fluency checks.

    That is progress monitoring. I need to look more closely at the benchmarks. Now, I noticed that those who do well and those who don't do well usually line up to what kind of performance I am seeing in the classroom.(There may be a surprise or two.)
     
  7. waterfall

    waterfall Phenom

    Joined:
    Feb 5, 2011
    Messages:
    4,974
    Likes Received:
    98

    Mar 31, 2012

    Hmm...we only benchmark 3 times a year, so those students wouldn't have very many aims scores for me to tell- and the last one was about 2 months ago. I guess I could give a few of my students (who I normally use aims with) a dibels passage or two and see if they test in the same percentile. We have an RtI meeting for a 1st grade student on Monday who there has been a lot of concern about (I won't go into it). Anyway, I was looking at her data and saw that her trend line looks pretty good- she's very close to where she could be. However, when you look at the actual scores she's reading about 45-50 wpm. While that is very close on dibels, that's more than 20 wpm what she should be for aims. I don't want her exited out of outside interventions if she's really that behind.
     
  8. mopar

    mopar Multitudinous

    Joined:
    Aug 15, 2010
    Messages:
    10,953
    Likes Received:
    0

    Mar 31, 2012

    What was her last aims benchmark? She would have just taken it a few months ago.
     
  9. waterfall

    waterfall Phenom

    Joined:
    Feb 5, 2011
    Messages:
    4,974
    Likes Received:
    98

    Mar 31, 2012

    Her last aims was 17 wpm taken in late december. Obviously, that's very low. This dibels data wasn't taken until starting at the beginning of february, so I don't have two scores that were taken at a similar time to compare.
     
  10. mopar

    mopar Multitudinous

    Joined:
    Aug 15, 2010
    Messages:
    10,953
    Likes Received:
    0

    Mar 31, 2012

    I would suggest continuing intervention until at least you benchmark again. Then you would have the data to compare. This is unless the classroom teacher thinks they don't need intervention anymore.
     
  11. EdEd

    EdEd Aficionado

    Joined:
    Jan 12, 2011
    Messages:
    3,557
    Likes Received:
    66

    Apr 2, 2012

    I would definitely recommend against trying to interchange different assessment tools. Part of the answer is in the norms that were created, and probably part of the answer is some small variability in passage difficulty based on the formulas used to level passages in each system.

    So, even if the teacher wants to use DIBELS, I'd say she'd also need to use the AIMSweb passages for purposes of comparing with other AIMSweb scores. If she feels like doubling up on the work because she feels the DIBELS is hugely different from AIMSweb I guess she can, but it's not going to work to compare AIMSweb to DIBELs because they are different systems, even if purporting to measure the same construct.

    Also, ideally districts would create local norms, and set their own cuts points based on the amount of intervention resources available. So, if funds are available to serve the bottom 20%, then the cutoff would be the 20th percentile, although screening benchmarks typically need to be set higher so the type II error is minimized, with further weeding occurring at later steps within a multiple gating assessment system.
     

Share This Page

test