Here is a sorting exercise for anyone building a K-12 dashboard. Below are twenty measures that turn up again and again on district dashboards. Each one goes in one of three piles.
Keep: it belongs on the board’s public dashboard.
Move: it’s useful, but it belongs on the cabinet’s internal dashboard.
Cut: it looks like progress but counts activity, so it belongs on neither.
Three questions do the sorting. Is the measure about students, or about what adults did? Does it move during the year, while there’s still time to act? And would the board do anything differently if it went up or down? A measure that passes all three goes in the first pile. One the cabinet would act on but the board wouldn’t goes in the second. One that nobody would act on goes in the third, however good it looks on a slide.
The first pile, for the board:
- Chronic absenteeism, by school and student group. Attendance Works notes that it’s most often defined as missing 10 percent of the school year, for any reason, and it can be calculated at any point in the year from absences and days enrolled. That makes it one of the few measures that moves every month.
- K–3 students at or above benchmark on the fall, winter, and spring literacy screener.
- Ninth graders on track to graduate, measured by credits earned and core courses passed each marking period.
- Course failure rates in grades 6–12, by marking period.
- Graduation rate. It’s annual, so pair it with the on-track measure that predicts it.
- State assessment proficiency and growth. Keep them, but don’t let them be the only academic measure, since the results arrive after the year is over.
- Suspension rates by student group.
- One survey measure tied to a plan goal, such as the share of students who say there’s an adult at school they can go to, asked the same way every year.
The second pile, for the cabinet:
- Staff attendance and substitute fill rates.
- Vacancies filled by the first day of school.
- Bus on-time rates.
- Work orders closed and facilities projects on schedule.
- Technology tickets resolved.
- Tutoring or intervention sessions delivered.
- Percent of the plan’s action steps complete.
Every one of these matters, and a superintendent who ignores them will hear about it. But they’re management measures. They tell the cabinet whether the work is getting done. The first pile tells the board whether the work is working. The action-step count is the one boards most often ask to see. If it goes on the public page, it belongs under the outcome it’s supposed to move, never in place of it. That’s the line between the two dashboards a district needs.
The third pile, cut:
- Professional development sessions held, or teachers trained.
- Family events held.
- Newsletters sent and social media followers.
- Website visits and app downloads.
- Any status written as a word instead of a number: “ongoing,” “on track,” “robust.”
These aren’t useless to the people doing the work. But on a dashboard they make the plan look busy rather than effective, and a board can’t act on any of them. A training session counts the same whether or not anything changed in a classroom. And a status written as an adjective usually means the evidence lives somewhere else.
Run the sort honestly and the board’s pile usually ends up with eight to twelve measures across the plan’s goals. That’s the right size: few enough that a trustee can read them in the five minutes before a meeting, and enough that every goal has at least one measure that moves during the year.
Two things finish the job. Each measure in the first pile needs a baseline, a target, and a deadline, or the board can’t tell whether 87 percent is good. And each needs its breakouts decided before the data arrives, by school and by student group, because an average can rise while a group falls behind.
The sort takes about an hour with a strategic plan committee. Do it before the first progress update, not after. It decides what the board will look at for the next five years.
BoardK12 builds the board’s public dashboard from the first pile: each goal, its measures, and the targets set before the first update.