How to Accredit a Failed (and Failing) School
At the same time, the Board of Education proclaimed a statewide accreditation triumph.
- One school that was “Accredited Pending Review of Alternative Accreditation Plan” in ‘18 is omitted from the graph.
- The VDOE Web page says 92% Accredited. The data from their spreadsheet show 92.77%. Even without their curious stance on roundoff (see below), that rounds to 93%.
- See below for the demise in 2018 of the several “Partially Accredited” categories and the genesis that year of the “Accredited with Conditions” status.
We can gain some insight into the reason for this boom in accreditations atop a slump in pass rates by examining the Petersburg data.
The system was denied accreditation continuously from 2006 to 2017. On the 2017 data, Stuart Elementary and Johns Middle were denied accreditation; AP Hill Elementary accreditation was withheld because the staff there got caught cheating.
This year, notwithstanding all that “help” from the state, all the Petersburg reading pass rates, except perhaps for Peabody/Johns, decreased.
- Data here are from the SOL database for reading; they no longer test writing in the elementary grades. Numbers in the School Quality Profiles can be different (see below).
- Petersburg changed the names of three elementary schools this year:
- A.P. Hill became Cool Spring
- J.E.B. Stuart became Pleasants Lane
- R.E. Lee became Lakemont
- Peabody Middle School staff and students were moved to Vernon Johns at the end of the 2017 school year.
- There were no AP Hill (now, Cool Spring) data for 2017 because the school got caught cheating.
Except at the high school, the math rates dropped as well.
The science rates rose at Walnut Hill and Pleasants Lane; they fell at the high school and tanked at Lakemont.
The Petersburg schools’ home page announced the resulting change in accreditation:
All Petersburg schools are accredited in the 2018-19 school year.
(To their credit, they also said that, except for Cool Spring and Walnut Hill, those accreditations were “with conditions.”)
The accreditation changes were dramatic:
How shall we explain this situation?
This year, there are three performance levels.
- Level 1: 75% current or 3-year average pass rate (boosted for “recovery,” growth, and English learners) or Level 2 prior year and decrease failure rate by 10%.
- Level 2: 66% current or 3-year average pass rate (boosted for retakes, growth, and English learners) or 50% prior year (boosted) pass rate improved by 10%.
- Level 3: Everybody else, plus anybody at Level 2 for four years.
For math, the scheme is the same except the Level 1 benchmark is 70% and there is no boost for English learners. For science, the math scheme applies except there is no boost for growth or recovery.
Here, then, are the English accreditation scores for the Petersburg schools.
The colors tell where the numbers came from:
- Light green: The pass rate as reported by VDOE.
- Brown: The percentage of students who flunked reading last year, took the remediation program, and passed this year. They count twice(!).
- Dark Green: The students who failed but showed “growth.”
- Blue: English learners (who are counted only if they pass, but then count twice).
- Red: The Level 1 benchmark.
- Gray: The Level 2 benchmark.
There are some anomalies here:
- VDOE reported a 51.3% pass rate for Cool Spring but their accreditation page says 55% (they round everything off).
- Pleasants Lane numbers similarly improved from 60.38 to 61. (Formerly VDOE rounded up at 0.45; that does not explain this change).
- As well, the 71.49 at Walnut Hill became a 72 (perhaps because of the enhanced rounding up).
- The reading pass rate/accreditation numbers probably are not comparable for Johns and the high school because the writing scores get averaged in.
Even with all the Finagle factors, Vernon Johns did not come close, even to the new, diluted benchmark.
Presumably Cool Spring could not enjoy “Growth” or “Recovery” boosts because the cheating canceled the 2017 scores.
But these enhanced pass rates are not the end of the matter. There also is a potential boost from the three-year average:
Poor Vernon Johns still looks to be beyond help. (BTW: VDOE does not explain how they calculate the numbers to accommodate missing Johns data for ‘17 and the Peabody scores prior to the merger). We need not examine the 2015 rules or the 50%-with-10%-gain options here: VDOE tells us Johns is Level 3 (also in math and science).
Despite an appalling performance in 2018 and no data in 2017 (cheating), Hill/Cool Spring makes the diluted 66% level (and nearly makes the 75% benchmark).
A little arithmetic (nearly) shows where the Cool Spring 73 came from: Here are the accreditation scores for Hill/Cool Spring for the last three years:
The actual three-year average for Reading is 48. BUT if we just average the 2018 value with the (obviously cheating-enhanced) 2016 value, we get 72. Close enough to 73 for VDOE, it seems.
Math is even more dramatic:
But the three-year bonus takes care of Pleasants Lane and elevates Cool Spring to Level 1.
Again, the only way VDOE can get the Hill/Cool Spring numbers is by ignoring 2017. No telling why they report 70 when the actual average was 69.
By any measure, Johns and Lakemont strike out and, indeed, VDOE reports them as Level 3.
The science scoring does not include the recovery et al. adjustments. On the pass rates, only Walnut Hill makes either benchmark.
The 3-year average does not save anybody.
On these numbers, Cool Spring’s enhanced average is 65, one point short of the Level 2 benchmark.
Another way to make Level 2 is to take a score of 50 or better and raise it by ten points. That plainly does not work for Cool Spring, which started in ‘17 at zero (or in ‘16 at 83 in science and dropped to just over half that in ‘18).
Nonetheless VDOE reports Cool Spring as Level 2 for science.
These “levels” feed into the accreditation ratings: A school is “Accredited” if all its school quality indicators are Level 1 or 2. Any school with an indicator at Level 3 is “Accredited with Conditions.” A school accredited with conditions “may” be denied accreditation if it fails to adopt and implement relevant corrective action plans “with fidelity.” The regulation does not tell us what level of “fidelity” is sufficient.
Aside: The question whether a school has implemented “with fidelity” does not “rest entirely upon . . . [an] examination” or constitute “approval by the Board of Education of a school division corrective action plan” so the due process requirements of the Administrative Process Act should apply. If the Board of Education should ever try to deny accreditation somewhere, it will be interesting to see whether they comply with those requirements.
Another way to get accreditation this year is to make the grade under the previous regulation. VDOE tells us no Petersburg school is in that category this year.
A last way to make full accreditation is to enjoy the running three-year accreditation exception from the statute. On that subject, here are the Cool Spring data:
Cool Spring was accredited in 2015 and 2016 (certainly because they were cheating) but not in 2014 (and of course not in 2017, what with getting caught at the cheating). So there’s no three-year run and the statute does not apply.
(But the data do make one wonder why VDOE did not bother to examine those obviously bogus 2015 and 2016 numbers and the resulting accreditation ratings. Indeed, if not rewarded for that obvious cheating, Cool Spring would not enjoy being “accredited” this year.)
Thus, we have the curious situation where VDOE blesses the cheating at AP Hill/Cool Spring with bogus English and Math ratings and with a science rating that looks to be invented from whole cloth.
There are further “achievement gap” measures that provide useful information but let’s pass over those and look at the 2018 results for Petersburg.
As the Petersburg Web page said, everybody is accredited!
None of the “conditions” schools, even Johns, which flunked on all the academic measures, can be denied accreditation unless it “fails to adopt and implement . . . corrective action plans with fidelity.”
So there you have it: The process rewards past cheating. The old benchmarks at 75 and 70% are reduced to 66% (at least for four consecutive years) and apply to manipulated pass rates that can be nearly double the actual pass rates. A school that misses those relaxed benchmarks, and does not benefit from one of the helpful exceptions, is accredited, albeit “with conditions.” That school can be denied accreditation only if it does not, in effect, tell the Board of Education to go jump in the James. Or, in the case of Petersburg, the Appomattox.
So, Petersburg gets to feel good about being “accredited.” Never mind the increased numbers of kids who are not being educated.
98.2 million of your tax dollars at “work.”