Skip to main content

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

Comment Preferences

  •  I will let an expert in ABA (0+ / 0-)

    or another data driven education model shred your paradigm of an education system with no quantitative accountability.

    http://en.wikipedia.org/...

    But In short if you cant quantitatively prove that your method of instruction is having a positive meaningful impact on your students.....you simply are not doing your job.

    •  well It hink the results we get in Texas (2+ / 0-)
      Recommended by:
      JerryNA, Sunspots

      Pretty much prove that this trend to data driven instruction doesn't  work.
      This year the passing score for 8th grade math on the State of Texas Assessment of Academic Skills is 35 percent.
      For Social Studies it varies from 46 percent to 49 percent, depending on grade level.
      Yeah, 35 percent on a multiple choice test, where chance gives you a 25 percent chance of getting any question correct. So if you are an average guesser, but you know the answer to 6 of the 60 questions,you pass.

      •  So (0+ / 0-)

        the Texas state education system is a complete failure and this is  somehow proof that a "data driven"  education system is is flawed? lol

        That sounds like logic which would come out of a state where 35% on a multiple choice test is considered passing.

    •  But that's the whole problem (1+ / 0-)
      Recommended by:
      Sunspots

      Trying to quantitatively prove you're having a positive impact and educating kids. Consider a couple of scenarios:

      Suppose Connecticut has harder tests than, say, Mississippi. In CT, eighth graders are taught the binomial theorem and tested on it. In MS, eighth graders are expected to add, subtract, multiply, divide. So on the standardized tests, Mississippi gets 90% but Connecticut gets 75%. Not because MS kids are smarter, but because the tests are easier.

      Or suppose a high school principal knows his school will be ranked in the bottom 10% of state schools and he'll be fired for having a failed school. He's got several kids who have problems (need counseling for problems at home, need drug treatment, are non-native speakers of English, or just need tutoring because they're a little slow). If he can convince these kids to drop out of high school -- instead of getting them the help they need -- his school's scores will go up.

      "Stupid just can't keep its mouth shut." -- SweetAuntFanny's grandmother.

      by Dbug on Sat Apr 27, 2013 at 08:25:33 PM PDT

      [ Parent ]

      •  Not sure what your points are (0+ / 0-)

        1) Are you trying to say that due to the fact that different states have different standards that a quantitative, data driven model for education would be unable to accurately be used to evaluate teach methods?  

        If so I would like to reassure you that the hundreds of thousands of students in even the smallest of states would be more than enough to come up a meaningful model.  

        Also even still teaching methods from one state are applicable in others. Human methods for learning do not change much according to state lines.

        2) The fact that you can make up a magic situation where someone can magically game said system is not a valid argument against a system. It is an argument for better more intelligent rules to be placed within said system.

        •  I guess if I had a single point it's this: (0+ / 0-)

          It's impossible to come up with a quantitative measure that is objectively measures learning.

          "Stupid just can't keep its mouth shut." -- SweetAuntFanny's grandmother.

          by Dbug on Sat Apr 27, 2013 at 09:18:56 PM PDT

          [ Parent ]

          •  Not even close to being true (0+ / 0-)

            I am not an expert in this field but nor do I need to be for this statement.

            I myself am not capable of coming up with an objective test (at least without years of training), nor apparently are you or the original author. But just because we do not have the skills does not make reasonable to proclaim its impossible.

            The proclamation is wholly egotistical and nonsensical.

            Experts within the field can and CURRENTLY do make reasonably accurate measurements of students academic performance and the impact of their educators.

            Educated professionals within the major camps of the educational and psychological professions  agree that understanding and applying methods which have been STATISTICALLY PROVEN to have positive impacts is a GOOD THING.

            Denying this is akin to being a climate change denier. It is the recourse of who are simply not professionals within the field, are scared of empirical facts, and perhaps those who themselves would not measure up.

            The scientific process works. Yelling "OMGZ what I do is so important it can not be measured. I am so special" Is a sign of incompetence no matter what profession.

            •  Strongly worded but true.... (0+ / 0-)

              still it misses the granularity of the constant testing discussed in the diary and it's value.  If the test isn't testing what is important to learn in a way that helps the student be effectively evaluated, then it isn't effective.

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

              by murrayewv on Sun Apr 28, 2013 at 04:38:04 AM PDT

              [ Parent ]

            •  0123458 - I've got to disagree with you (2+ / 0-)
              Recommended by:
              houyhnhnm, Sunspots

              The push for standardized tests does NOT come from experts in education, in fact education experts deplore it.
              The push for standardized tests to measure performance comes from politicians and blue-ribbon commissions of wealthy businessmen.
              I'm just now reading "Teaching the Best Practice Way" by Daniels and  Bizar. Here are a few of their thoughts:

              "To this day we screen, track, reward, and segregate people using tests rooted in bad science and redolent of bigotry."

              "Since most standardized test scores correlate highly with socioeconomic status, testing will reconfirm the unworthiness of the underclasses and comfort the privileged. The predictable results justify the blaming, shaming and stigmatizing of low scoring schools and communities. After all, why should we spend more money on those people, if they can't even pass the test."

              "Instead of reading real books, the kids read and fill out sample standardized tests. The tests  literally become the curriculum."

              They argue that ever since 1917, when standardized tests  were first used to rate the intelligence of immigrant groups to America (not surprisingly "proving" that Nordic people were at the top of the scale and Africans at the bottom), standardized tests have had 2 major flaws - they assume intelligence is a single, measurable trait and that it is permanently fixed.

              •  The details of that argument (0+ / 0-)

                are moot.

                Those points amounts to something along the lines of "people are using data incorrectly and not extrapolating correct results so lets not use them at all"

                This opinion is a "baby with the bathwater" opinion.

                Additional the argument follows from a logical fallacy.

                http://en.wikipedia.org/...

                "Hasty generalization, or converse fallacy of accident"

                Ive said it before and I will need to say it again. Creating magical situations or even finding real examples of a problem within a system is simply NOT a valid way to argue against an idea or system no matter what the context.  

                That is simply an illogical argument.

                It is an argument for better regulation and implementation.

                •  OK 012358, Let me spell out my (0+ / 0-)

                  Actual experiences in a real classroom as well as what actual education experts have found
                  * Curriculums have grown so large it  is impossible to test enough to gather a statistically significant amount of data on which to drive instruction. If there are 110 TEKS to teach, I would argue it would take about 10 questions on each TEK to have enough data to draw good conclusions. That would require students to answer 1,100 questions. It's not gonna happen.
                  * That means we are driving instruction with statistically meaningless data - one or two questions per TEK. And it means we are totally ignoring what the teacher actually observes. A teacher who spends hours every week observing students.
                  * In order to gather this meaningless, misleading data, we are testing students for more than 100 hours a year, a huge waste of time that could be better spent.
                  * In order to analyze all this data, teachers are spending five or more hours a week, again, time that could be better spent on a myriad of other tasks.
                  * All of this is done, not to improve student learning, but to improve performance on a single, high-stakes, standardized test,which has serious flaws, see my comment above about the dubious value of standardized testing.
                  * You are right in that properly analyzing meaningful data is a valuable exercise. As it is practiced in the real world, it doesn't work well in classrooms.

                  •  response point by point. (0+ / 0-)
                    1)"* Curriculums have grown so large it  is impossible to test enough to gather a statistically significant amount of data on which to drive instruction. If there are 110 TEKS to teach, I would argue it would take about 10 questions on each TEK to have enough data to draw good conclusions. That would require students to answer 1,100 questions. It's not gonna happen."
                    There are many ways to extrapolation meaningful, statistically significant conclusions from reduced bits of data per student.

                    google "sparse data analysis techniques"

                    http://www.ncbi.nlm.nih.gov/...

                    http://searchbusinessanalytics.techtarget.com/...

                    Sparse data analysis techniques are commonly used among numerous professions. For some reason SOME teachers think that using complicated crazy things like "science" is a bad idea.

                    "* That means we are driving instruction with statistically meaningless data - one or two questions per TEK. And it means we are totally ignoring what the teacher actually observes. A teacher who spends hours every week observing students."
                    It may be meaningless to you but it has meaning to people who are properly trained to use it.
                    * In order to gather this meaningless, misleading data, we are testing students for more than 100 hours a year, a huge waste of time that could be better spent.
                    * In order to analyze all this data, teachers are spending five or more hours a week, again, time that could be better spent on a myriad of other tasks.
                    Quality control is a VERY VERY important aspect of any profession.

                    http://en.wikipedia.org/...

                    "their attributes include a high degree of systematic knowledge; strong community orientation and loyalty; self-regulation; and a system of rewards defined and administered by the community of workers"

                    * All of this is done, not to improve student learning, but to improve performance on a single, high-stakes, standardized test,which has serious flaws, see my comment above about the dubious value of standardized testing.
                    This is all done because the education system in its current state has no accountability and thus is currently a complete failure in this nation.  

                    In order to bring accountability to the education system reformers are pushing to add more self regulation and bring back teaching to its proper place as a true profession. Some teachers will not able to make the cutoff  which is part of the goal.

                    If the current system had adequate self regulation we would not have situations where an entire state had a passing rate of "35%" of a state wide test.

                    Over the years teaching more or less  lost its status as a "profession"  as part of this degradation it became acceptable to use methods which were no empirically proven to work.

                    It will take time, effort, resources,  and properly monitoring the quality of our teachers to bring the educational quality back up to where it should be.

                    Anytime you have a situation where there is such a complete lack of oversight as there had been the last many decades, quality is bound to degrade. I am glad members within the teaching community are catching on to this and doing something about it by adapting to data driven educational models.

                     

Subscribe or Donate to support Daily Kos.

Click here for the mobile view of the site