School test scores results can be highly misleading
Aug 28, 2017 at 3:00 PM
Use of misleading statistics is distressingly common in the field of education, and it is doing a lot of damage. For example, The Baltimore Sun article, "Less than half of Maryland students pass English, math assessments" (Aug. 22), states that "Anne Arundel County spent last year focused on improving middle schools, middle school math instruction, in particular. The effort appeared to pay off. At Arundel Middle School, the pass rate in math rose 8 percent among sixth graders and 16 percent among eighth graders in one year. Of those students who took Algebra 1, 83 percent passed. Principal George Lindley said the county gave teachers specialized training that allowed them to make math more concrete for students. Anne Arundel's elementary and middle school math passage rates went down overall to 37 percent."
Those sound like contradictory statements to me. Whereas my statistics are limited to those from The Sun and the Maryland State Department of Education, though Anne Arundel County's 8th grade math scores may be higher than in 2016, they appear to be lower than in 2015. Also, in order for the claimed 16 percent improvement to be accurate and not inflated, the 2016 passing rate would have had to be 0 percent, though the data simply stated than it was less than or equal to 5 percent. Regarding the Algebra score, it should be noted that geometry was not a course offering at Arundel Middle which struck me as unusual for a middle school of 1,014 students. Was that in order to provide extra instruction to raise algebra scores and thus look good? Was that a priority over offering capable students the option of taking geometry (and thus more advanced courses later)? It seems like the "specialized training" and lack of opportunity that took Arundel Middle's algebra scores from 70.6 to 82.6 were less effective than what occurred at Lake Elkhorn Middle (46.5 to 75.4) or Wilde Lake Middle (29.5 to 68.8) in Howard County (it is difficult to compare counties overall, as differences don't appear or are compressed for schools like the several in Howard County with scores above 95 percent). Both of these schools have much smaller enrollments but offer geometry, nonetheless.
I guess that "specialized training" took a back seat to having the word "Lake" in the name of the school, statistically speaking. Another similarly misleading use of data is when administrators take credit for improvements that occur due to socio-economic changes brought about by boundary changes and/or new development. Another example of data abuse involves the teacher evaluation system used in most of the state which involves teachers being judged by student growth, measured by improvement in student grades often subjectively assigned by the teachers themselves (among many other disgraces in the system). It's not that data can't be useful, but there is far too much misleading data masquerading as useful data in the field of education and it has done much harm to curricular offerings, instructional time, morale, finances and other aspects of education that result in negative outcomes for students.
Let's put data in perspective and use it honestly and appropriately. Let's stop judging schools based on test scores which essentially just measure socio-economics. Let's stop assuming that a student will be more successful at a school with higher test scores. Let's stop using data as excuses for implementing job-justifying unproven fads. Let's realize that great teaching is an art and cannot just be measured by data. Let's give teachers more time to teach and communicate with families by reducing the burden of excessive standardized testing and inordinate data collection, maintenance and evaluation. Let's treat students and teachers as people and not statistics and school as a place of learning and human development and not as a business. Let's be skeptical when we deal with data, as teachers, parents, journalists, administrators or politicians. When we use data, let's not mislead, inappropriately manipulate others, or be mislead. Data is a tool; it should not be the ultimate objective. The most important things we learn in school can't be measured anyway.
Robert W. Miller, Columbia
The writer is a retired Howard County Public Schools teacher.