The field tests are upon us… Both consortiums are now actively trying to do field testing to get ready for their “for-real” assessments to be levied next year….

The PARCC field testing is stumbling next door in Maryland… as seen here….  Tests were given without headphones and the fact that headphones were required and not provided, was actually not realized until children began to raise their hands to ask how they were supposed to “listen” to a reading….  A sound vote of confidence in corporate ingenuity and corporate planning.

Already the Smart Balance Assessment has had its nationwide roll-out delayed  to solve similar bugs…

The problems lie with the adaptive nature of the Smart Balanced Tests… The Smart Balanced test is designed to figure out the next question to ask each student, based entirely upon what answer the student gave for the preceding questions…  This of course, requires a larger data base from which to draw, than would an old fashioned test based on asking true and false questions,…

So here is the dilemma.

For the Smarter Balanced Assessments to work, they need large pools of data. The way that works is its creator tests the common items across the tests in a variety of forms….  For example if you wanted to test cucumbers, you would fondle and eat cucumbers in a variety of forms… pickled, fresh, refrigerated, room temperature, sliced, shredded, diced, sautéed, stewed, as well as distinguish between seeded and seedless, baby and grown.  You would after eating 4 million cucumbers, have a good idea of what a good cucumber would and should taste like…

Same principle in standardized testing.

But… you are testing all foods, not just cucumbers… Therefore to get data on all foods, you have to limit the time you spend on analyzing cucumbers…  There are squash, beets, carrots, broccoli, kale, and corn to consider….

On the Smart Balanced Assessments, whenever the Common Core sample is large, there becomes less time for sampling the non common items… So in a test case, as is this spring testing session, the Common Core sample used in analysis, has been pared down to a smaller consistent core of 10 items….  What this means, is that the validity of this future test will be based on data not gleaned from an infinite selection, but from only ten items hand picked to be the consistent core…..

This is sort of a big issue.  In 2015, the Smarter Balanced test will have an adaptive design and adaptive tests require many items…

Inside the Smart Balanced organization, apprehension is palpable over the use of THIS data to hold anyone, or anything accountable… This run is to test questions.. not to test students or teachers.

It appears that until today, no one in Delaware’s Department of Education had the slightest clue over this distinction.  That should be alarming.  The Smart Balanced Assessment pilots cannot and should not be used to evaluate either children or teachers… In fact, under the bylaws of September 17th, the U.S. Department of Education in a significant concession, declared states would not be required to report scores for those students taking the field test. ….

Therefore  for our students, parents, and teachers, it is a day wasted.  But for the Smart Balanced Assessment consortium, it is their phishing for data….

As more in depth analysis unfolds, it become clear that like another historical event in our nation’s past, everyone is guessing and no one really yet knows exactly that the outcome will be.  I would be referencing  the Manhattan Project and the formulation of the first atomic bombs.  It would probably be wise to remember that even though those outcomes were deemed successful for the creators of that project, they did not bode well for those upon whom it was first tested…

This time of course, it will be our own children.

 

 

 

 

 

 

 

 

 

 

Advertisements