High School Corruption and State Malfeasance

In 2016, VDOE conducted a course schedule audit at Armstrong as part of the work following denial of accreditation there.  VDOE reported discrepancies in course requirements and transcript accuracy, inter alia.  The 2017 follow-up audit “concluded that there was not sufficient evidence to determine that problems identified . . . had been resolved.”

In 2018, at the request of our Superintendent, VDOE expanded the audit to include all five mainstream high schools.  They found:

  • Bell schedules did not meet the number of hours required by the Standards of Accreditation.
  • Verified credits did not populate in the transcripts.
  • Attendance data was incorrect throughout the transcripts.
  • Some students received one credit for classes that should not have carried credit.
  • Some students received two credits for classes that should have carried one credit, such as Career and Technical Education (CTE)classes.
  • Credit was incorrectly given for what appear to be locally developed elective courses without evidence of approval by Richmond Public Schools Board.
  • Credit was incorrectly given for middle school courses ineligible for high school credit.
  • Course sequencing issues were identified.
  • Academic and Career Plans lacked meaningful content.

Translating that from the careful bureaucratese of VDOE: The schools were cheating wholesale to boost the graduation rates.  Indeed, the RT-D later reported the schools were:

Rubber-stamping student work. Choosing to use an alternative test instead of giving students the common state test. Putting students on individualized education programs to circumvent state graduation requirements.

As Scott Adams is fond of saying,

Wherever you have large stakes, an opportunity for wrong-doing, and a
small risk of getting caught, wrong-doing happens. That’s a universal law of
human awfulness. When humans CAN cheat, they do. Or at least
enough of them do.

RPS seems to be determined to prove Adams right.  This latest outrage, comes in the wake of the wholesale cheating at Carver.

To be clear: We’re talking here about cheating by the school staff, not by the kids.

In the past, this kind of thing has been manifest in the data.  An earlier post looked at the RPS pass rate averages and suggested cheating.  So I thought I’d take a (further) look at the graduation numbers.

Notes on the data:

  • Economically disadvantaged students (here, “ED”) underperform their more affluent peers (“Not ED”) on the SOL pass rate averages by somewhere around twenty points.  Where there are large ED populations, the overall SOL averages can be misleading, so we’ll look at the ED and Not ED pass rates separately.
  • The End of Course (“EOC”) tests are the gateway to obtaining “verified credits” toward the graduation requirements.  To graduate with a standard diploma, the student must pass two EOC tests in English, one in math, one in history & social sciences, and one in a lab science.
  • On the 2018 EOC tests, about 60% of the Richmond students tested were reported to be ED.


  • That’s an average.  The tested ED populations of our mainstream high schools vary considerably (the selective high schools all do very well, thank you, and the boot camp Richmond Alternative is a mess so there’s nothing to be learned here from their data).



  • The graduation rates here are the boosted rates the Board of “Education” uses to make their constituent schools look better than they really are.

With that out of the way, let’s turn to the cohort graduation rates of those five schools and the pass rates there on the EOC tests. 

First, on the reading tests:


The orange diamonds are the 2018 cohort graduation rates/2018 pass rates of the ED students at the five high schools.  The blue circles, the same data for the Not ED students.

The easy way to read the graph: Look at the slope from orange ED to blue Not ED: 

  • Up and to the right is the expected condition with the ED students both scoring below and graduating at lower rates than the Not ED. 
  • As the line approaches horizontal, the ED pass rate is lower but the graduation rate approaches the Not ED rate.  We can wonder what is going on.
  • When the line slopes down, the lower scoring ED students are graduating at a higher rate than the Not ED.  Think cheating.
  • When the line approaches or passes vertical, ED students are passing at about the same rate as the Not ED, and graduating at a higher rate.  Something is doubly rotten.

Here, the Armstrong data look reasonable: The Not ED students outscored the ED by almost fifteen points and enjoyed an almost ten point better graduation rate.  We can wonder whether this is the result of the earlier audit there (see below).

In contrast, the ED students at both John Marshall and TJ graduated at higher rates than their Not ED peers while passing the EOC tests at 20+ and ~15 percent lower rates.  Something boosted those ED graduation rates.  Cheating is the obvious first candidate.

Wythe shows a more striking, anomalous pattern, with a (smaller than expected) 10% pass rate deficit and a twelve point higher graduation rate for the ED students.

Huguenot is anomalous in two respects: The pass rates of the two groups are almost identical and the ED students graduated at a rate almost 30% higher than the Not ED.  We can wonder whether the large ESL population there bears on this.

The math data tell the same stories (at dishearteningly lower pass rates).


Armstrong’s data again look reasonable.  Wythe and, especially, Marshall and TJ show higher ED graduation rates than Not ED, with lower pass rates.  Huguenot shows a much higher ED graduation rate at about the same pass rate.

To gain some context, let’s look at some history. 

First, as background, the state averages.



There is the norm: Slope up from orange to blue.  In terms of numbers, ED pass rates about twenty points below the Not ED and cohort graduation rates about seven points below.

Turning to Richmond, here is Armstrong on the reading EOC tests:


2014 and 2015 clearly suggest cheating.  2016 suggests it; 2017 says it again.  But 2018 looks reasonable. 

We might infer that the presence of VDOE in 2016 had a salutary effect at Armstrong.  A caveat: Because of the very small numbers of Not ED students at Armstrong, the Not ED pass and graduation rates can be expected to fluctuate considerably from year to year (as they do here).  These data fit the inference of cheating before ‘18 with a course correction then but there might well be a more benign explanation.

The math data at Armstrong also fit the inference of cheating, abated upon the appearance of VDOE.


The Marshall data are consistent with unabated cheating, albeit with higher pass and graduation rates than at Armstrong.



The Wythe data tell a similar story, with a suggestion of robust cheating in 2018.



Huguenot data suggest ongoing, rampant cheating.



On both tests the 2015 Huguenot numbers show ED and Not ED pass rates that are nearly identical, so the positive slope that year does not offer any solace.

Finally we have TJ, with anomalous but OK patterns in 2017 (albeit with unacceptably low ED pass rates) but otherwise showing data consistent with cheating.



————————– Rant Begins ————————-

Our Board of “Education” has created a haven for cheating high schools:

  • They don’t look at their mountain of graduation data until after they review a school that has been denied accreditation (if they even look then), and
  • They now have rigged the system so it is almost impossible for a school to be denied accreditation.

The Board has institutionalized Adams’ “small risk of getting caught.”

Why did VDOE ignore its mountain of data and do nothing in Richmond until (1) Armstrong was denied accreditation (under the old, now replaced, system), and (2) our Superintendent invited them in to the other high schools?  Why are they not looking for this pattern at high schools in other divisions? 

I think the answer is at least nonfeasance and probably malfeasance.  It’s past time to fire the Board and appoint people who will actually do the job.

————————– Rant Ends ————————-

It will be interesting to see how all this affects this year’s graduation rates.  Those data should be available in mid-August.  Stay tuned.

Oops: Corrected the link and list of required EOC courses.