GUEST COLUMN, David Dorsey, Kansas Policy Institute
The state board of education is in the throes of adopting cut scores for state assessment results; the test was taken earlier this spring by Kansas public school students. These cut scores divide individual student achievement into the four levels created by the state board, with Level 1 the lowest and Level 4 the highest.
These levels have unique descriptors that express how each student (third through eighth and tenth grades) performed on the latest state assessment. The state board in May changed the operative words in each of the Levels in terms of student performance on those standards (the replaced words in parentheses):
Level One: limited (limited)
Level Two: basic (basic)
Level Three: proficient (effective)
Level Four: advanced (excellent)
These descriptors apply to both math and ELA (reading).
The board is scheduled to adopt the new cut scores that correspond to these levels later this summer. A group of approximately 140 Kansas educators convened this month to analyze the results of the new 2025 state assessment and recommend those cut scores to the board.
If this process sounds complex, well, that’s because it is. And, it was seemingly designed not just to sound complex but to actually be complex. Perhaps the most puzzling is the score itself that a student receives after taking the test. The student’s score is somewhere in the range of 220-380, which is not necessarily an indicator of the percentage of questions the student answered correctly. Below are the cut scores that have been in effect since 2015.
Notice the cut scores above are similar, but vary from grade to grade and from math to ELA. Also notice that cut scores dividing Levels 2 and 3 are always at 300.
Conspicuously absent from the entire process of scoring, leveling and describing the results are two simple, yet very significant words: grade level.
At the last board meeting, the board was given a presentation on the process by Dr. Karla Egan, a psychometrician (careful not to divide that into two words!) and founder of EdMetric. Dr. Egan led the board through the process of creating the cut scores without describing how a student receives their individual scores.
Of course, it didn’t take long for the discussion to turn to what constitutes a student being on ‘grade level.’
At that point, commissioner Randy Watson interjected. Addressing the board, Watson stated, “The question that will be asked of you forever is: Is my kid at, below, or above grade level? The answer is: The assessment (is) at that grade level and the descriptors telling you how well they do at that grade level…And that is a question you are going to be asked.”
Everybody got that?
Watson implied in this statement that because students are tested only on the standards adopted for their grade level, it is impossible to say whether a student is performing at “grade level” because they are only tested on their grade level. And he tried to convince the state board members of that. Using his logic, a student could get all or nearly all questions on the assessment incorrect, but because there are no questions from a grade lower, you can’t say a student is below grade level. I don’t know about you, but if a student answers 90 percent of the questions incorrectly, I believe that student is NOT performing at grade level.
However, on the other end of the scale, it’s not possible to use the state assessment to declare a student at ‘above grade level’ because they don’t get questions from the next grade level.
Since there will be over 100 educators analyzing these results, they should be able to come up with a reasonable range for a student to be described as performing at ‘grade level.’ That way, parents will have some real meaning to how their child or children do on state assessments. After all, Watson himself told the board that the question they (board members) will always get is: “Is my child performing at ‘grade level’?
But no, that would give some teeth to the state assessment results. You know, results that might actually have consequences because of accurate descriptors and avoid that nebulous descriptor of ‘basic’. But the Watson-controlled board is loathe to accountability and what better way to avoid that pesky accountability than to continue the practice of putting meaningless descriptors to state assessment results?
The lack of accountability via vague and confusing descriptors is a major reason KPI created the A-F Grading system. KPI has given some much-needed simplification to understanding state assessment results by translating state assessment scores into letter grades, making it much easier to understand how schools are performing on those assessments.
Here’s a solution (which will likely never happen due to its simplicity and lack of wriggle room in analyzing). First, get rid of the psychometric scoring. Does the state board really need the tutelage of a psychometrician to determine if a third-grader is performing at grade level? The results should be reported in terms of percentages of correct answers, not a psychometric model. You know, numbers the average Joe can make sense of. How exactly did my kid get a score of 297? And what does ‘basic’ mean? Is that good or bad?
Convene those same educators who are determining cut scores to establish a range of correct answers that can be described as ‘grade level’. Those scoring under that range would be ‘below grade level’ and those scoring above that percentage would be considered ‘advanced.’
The situation does get a little complicated because the feds require the states to have at least two levels that a state considers ‘proficient’, but that’s a discussion for another day.
Alas, the recent and coming activities of the state board regarding state assessment results will provide another decade of avoiding the $64 dollar question.
