Skip to main content

View Diary: Education: Why data-driven instruction does not work (98 comments)

Comment Preferences

  •  it's also part of (12+ / 0-)

    the mentality that no one wants to spend a dime on doing something for someone without "proof" that they are "getting their money's worth."

    I know, because this data-driven instruction is nothing more or less than what therapists like me have always had to do.   When we initiate something like speech-language therapy with someone, we have to get a baseline of their performance on whatever it is we think needs fixing, then set a goal, then provide intervention all the while measuring whether the person is progressing toward that goal or not.  It's been this way ever since I was trained in the early 70's, and when the special ed law (now IDEA) came into effect, speech therapists had no problems adjusting because we were already doing it.  Special ed teachers were not so easy to make the switch.  And now regular ed teachers are having to try to fulfill some distorted version of this.

    And what happens is that daily data points for therapy sessions generate a lot of "noise."  If I am working on word retrieval with a stroke patient, I don't want to use the same set of items every session, that would be bad therapy.  But then my data will look screwy because I can't control for stimulus difficulty level (which is pretty idiosyncratic for each person so it's not like you can just use a graded set of stimuli).  Now Medicare seems to recognize this and you only have to give data every 10 sessions.  But our state Medicaid wants data every single session with detailed notes on progress toward goals and barriers to progress etc.  And what that does, is it makes therapists use cookie-cutter activities that are easy to get data from, rather than some of the rich social-interactive strategies that may be harder to quantify.  In the end, everything CAN be quantified but it is way too much of a pain in the ass to figure out how to do that.  Unless you had a grad student videotaping each session and doing it for you.  Like people who dream up this stuff must have.  But the constant drive to engage in activities that can be easily quantified and tracked leads to therapy that I think is much less than what it can be if I'm left to just do my job and use my brain and my knowledge of the therapy process.  Beginners in a field might need to do some of that but once you know what you are doing it just gets in the way.  

    But all of this is because they don't want to spend any money on these services, they want to find excuses not to pay for services, and everybody wants proof of effectiveness of services because god forbid we waste a nickel of money.  When in reality, as a therapist who has treated people long-term (kids with autism, Down syndrome stay in therapy a long time), people don't make steady consistent progress and lots of time not much happens in therapy--sometimes for months.  And then we get a breakthrough and all that "wasted money" turns out not to be wasted after all.

    I had one child with Down syndrome that could not read anything but the word CAT at age 8.  The guardian begged me to add reading to my therapy because "school was not teaching her to read."  I spent one whole year and got her to learn 30 words.  This cost quite a bit of Medicaid money, maybe 8-10 thousand dollars.  So, doesn't look so good, right?  But then she read a sentence, then next thing you know we were reading adapted novels like Black Beauty.  And now she reads 5-th grade level books if you help her with a few words here and there.  And her speech got incredibly more intelligible because of the slow reading (she talked too fast).  And recently her case manager commented on what a huge vocabulary she has in conversation.  She can recite the Pledge of Allegiance from memory because she could learn it by reading it, whereas she can't remember anything she hears past a 4-5 word sentence at a time.  

    •  Nice story.... (1+ / 0-)
      Recommended by:

      I teach and the key questions are about common misunderstandings of important concepts.  Those data help you retest students.  And some of the data driven teaching is based on the concept of flipping teaching.  Having students do homework to learn concepts on their own, with something like khan academy, and then testing or evaluating the homework electronically to check what things folks are confused by and reteaching these concepts.  This type of teaching can be very effective in some fields which are concept heavy.  But history or reading English stories and evaluating their meaning?  Perhaps not good.  Dates in history can be taught in some fun ways too.  I cannot completely reject databased evaluations, but folks need to spend quite a bit of time figuring this stuff out to make it more effective and not a random noise generator.

      You shall know the truth, and the truth shall make you mad. Aldous Huxley

      by murrayewv on Sun Apr 28, 2013 at 04:03:50 AM PDT

      [ Parent ]

Subscribe or Donate to support Daily Kos.

Click here for the mobile view of the site