Beaufort County Schools needs a better accountability system | Eastern North Carolina Now

Hood Richardson, the perennial whistleblower in Beaufort County, recently penned an I Would Rather Be Right column entitled “Insanity by the Beaufort County School Board” in which he pointed out the insanity perpetrated against Beaufort County by its School Board in continuing to do things the way they have always been done and expecting different results.  He documented the poor student performance as measured by the State’s accountability program saying:

Beaufort County has 7 out of 12 schools that scored as low as 42 out of 100 points and got a “D” rating (the lowest passing grade) on the State’s school evaluation system. This is another indicator of the insanity of what is going on in Beaufort County. We have one school, Early College, that is rated “A” with a grade of 95. That is a spread of 43 points to between the lowest performing to the top performing school.  One must ask,   “What are the other 7 schools doing wrong?”  Chocowinity Primary School has a grade of “B,”  Northside  High School and Bath Elementary School have grades of  “C.”  No explanation was given about programs to improve ratings other than they are “working on it” and “Covid must have done it.”  Excuses rather like:  “the dog ate my homework.”  And remember, this low performance was going on for years before Covid.

Hood, as usual, was right.  Something must be done to turn around this dismal performance record by the Beaufort County School system.  Here is a idea whose time has come, we feel.

The School Board should adopt a formal accountability policy and exigent procedures to change the accountability from one focused on means (averages) to one focused on individual gain scores by each and every student.  The state will continue to report average performance data by school but what a new accountability system should focus on is “how much did each student improve their performance in one year.” 

The existing technology is present already to allow this to be done.

The way it would work is that each student would be assessed at the beginning of the school year (based on the End of Grade (EOG) or End of Course (EOC) scores from the previous year) to determine “where the student is” at the beginning of the year and then use the EOG and EOC at the end of the year to determine how much progress each student actually made in one year. The difference between the beginning score and the ending score would be the “Gain” score.  The teacher, school and system score would be simply the percent of students who met or exceeded the projected gain score.

For example, the results released to the public would look something like the percent of students in each grade and subject tested who exceeded the Expected Gain score.  And those data should include the standard deviation of the gain scores to measure how spread out or how bunched together the group scores are.

For the parents, the way it would work would be: At the beginning of the year a conference would be held with each student’s parent/guardian to explain the Prescore and the expected gain.  The Expected Gain score should be an adjusted gain score to include the standardized gain expected for one year of schooling plus whatever additional gain the student, parent and teacher agreed upon for that child that year.

One school that made the most effective use of this system actually developed three gain targets; one for minimum, a higher one for goal achievement and a third they called a “stretch” gain score.  Each grading period the student’s performance was reported as the usual A/B/C/D/F based on marking period teacher-based assessment but also a progress report of how close the actual performance for each marking period was to the expected gain.

That school actually built a school culture reflecting goal achievement using the individual student gain scores per marking period.  A list was published each marking period of the students in each category (minimum, “on target” and stretch) much as the honor roll has been done in the past.  A reward system was attached to each category that students in that category enjoyed.  During the year, at parent conferences, the progress of each student in meeting the gain targets was reviewed and the parent given ideas to help encourage high expectations being inculcated in the system.

This target setting system was used to turn around several schools in school districts that used the system.

Research done by the East Carolina University Rural Education Institute showed that while the system was extraordinarily successful in most schools, it was not uniform across all schools.  The research showed that the variable most closely associated with the highest performing schools was the effectiveness of the Principal in managing the system.  Leadership was deemed an essential ingredient in making the system successful, as one might expect.  Direct involvement by the Principal was a constant variable observed in the most successful schools. 

The same patterns were observed with teachers.  Those who became more committed to the system had the best results.  And surveys showed that teachers came to like the system better than the means based system as much as anything else because it was seen as “fairer” to teachers because success did not depend on the number of higher performing students a teach had.  In fact, it became obvious that lower performing students usually had the best gain scores, again as might be expected.  So a teacher with a higher percentage of lower performing students at the beginning of the year could look just as good as a teacher with a higher percentage of high performing students, as expressed by gain scores.

More importantly, the system of focusing on individual gain scores tended to be a more effective motivational system for instilling pride in lower performing students because they were compared to their “former self” as opposed to a mythical “average.”

It is noteworthy that the system was not sufficient however.  Several schools and districts added standardized measurement based on percentiles.  That requires some new technology for some schools but it was useful to assess reliability of the teacher/parent/student  derived gain scores.  In other words, if a student’s pre-score was at the 90th percentile, the gain score target would be appropriately lower than if the pre-score was in the bottom quartile.  There is a need to always compare local students to national norms for reference purposes.

The real advantage of the system is that it focuses on each individual student and the goal/reward system is tailored to each student.  It is more effective in focusing a parent on how well their child is doing rather than expect parents to get excited by mean scores. 

Research on the validity and reliability of the Gain Score model did show a significant “Hawthorne Effect” that must be taken into consideration in evaluating the system.  The Hawthorne Effect is simply the effect that being in an experiment has on the participants of the group as opposed to the participants’ behavior without regard to being in the experimental group.

But we would note that one of the perceived weaknesses of the way Beaufort County Schools monitors accountability is that it is seldom reported accurately.  For example, never have we seen composite data (such as mean scores) reported that included any measure of variance within the data results.  That is important.  Knowing the standard deviation of a data set shows you how spread out the results are compared to the mean.  Two mean scores, such as 80, have different meanings if the standard deviation of one set of 80-scores is 5 and the other is 25.  Another example is suppose I told you one basketball team has an average height of six feet while a second team also has an average height of six feet.   Which team would you pick to win?  But if I told you that the first team had a variance of three inches while the second team had a variance of twelve inches which would you then choose to win?  That is too much to explain here, but it is important that one should always be skeptical of group data that does not show the standard deviation score along with the mean.  Apparently, even the Beaufort County Superintendent does not understand the importance of measuring the amount of variance in group scores.  He has never used it in any public release we know about and did not include any analysis of variance in a statistical study he authored in his doctoral dissertation.

Be all that as it may, we return to Hood’s argument.  Beaufort County Schools is a low performing school system, it has been for many years, and it is not improving fast enough to save another generation of students from the impact of a low performing school system.  Hood argues that it is insanity to continue to do things the same way and expect better results.  That low performance is what our School Board has accepted and it should be stopped.

There is an election coming up next November and that election will give the voters a chance to weigh in on this situation.   We think the school board members who have been in the driver’s seat for several years should be replaced and that is probably the most important change that needs to be made to abolish this insanity our kids are suffering from.