Polk’s “Best” and “Worst” Schools: Not Who You Think


What if I told you that the following are the top 10 of Polk’s 80 elementary schools:

1. Sandhill
2. Hartridge Academy
3. Frank E. Brigham Academy
4. Janie Howard Wilson
5. Alturas
6. John Snively
7. Eastside
8. Jesse Keen
9. Lake Shipp
10. Ben Hill Griffin Jr.

And then what if told you these were the worst 10:

71. Medulla
72. Elbert
73. Valleyview
74. McKeel Elementary Academy
75. Highlands Grove
76. Lincoln Academy
77. Sikes
78. South McKeel Academy
79. Berkley Elementary Academy
80. Ridgeview Global Studies Academy

Would you believe it? Well, I’m about to make that case to you, based on the results of the 2010 3rd grade FCAT.

(The 2011 version was some sort of weird experiment in recalibrating the test, and the results have to be converted to 2010 equivalents and all sorts of strangeness. So I distrust 2011 even more than I distrust the normal of years of this stupid test. Third grade was the first I saw on the state website, and educators seem to consider third grade especially important. That’s why I picked it, but I encourage other people to tackle other combinations.)

Anyway, this is going to get a little long and wonkish. Fair warning. But first, I want to revisit for a moment the origin of my ongoing online jihad against the McKeel model of “public” schooling.

It started back in September of 2010 with Principal/Czar Harold Maready mouthing off. The pro-charter movie “Waiting for Superman” and a handful of spinoff “news” specials, McKeel’s famous $70K beach trip, and revelations that McKeel dumps a lot of kids back on to traditional schools, had thrust McKeel into the news quite a bit.

Maready had this to say:

“In review of all the data, magnet, choice and charter schools are making a difference, which should be studied and implemented in other schools. Charter schools are not the total answer but are part of the solution.”

Maready said there needs to be drastic changes made in education, just as the MSNBC program pointed out.

“Analysis of the data would allow for an open forum to work together in solving the education issues in Polk County,” he said. “If we do not recognize there is a problem, then there cannot be a solution.”

He also said, in a different story about the same time: “We’ve got to be doing something right. If you look at the traditional public schools, that’s not us. We like to do things differently.”

Let me be clear. When you talk that kind of elitist public smack in and about the community I love–and the people who live and work in it–you make an enemy of me. (Ask “national economist” Brian Wesbury or Jay Dennis.)

I am not a significant or powerful enemy. But I am a persistent one. And I happen to be the type of enemy who might one day take up on an offhand challenge like this:

“In review of all the data, magnet, choice and charter schools are making a difference, which should be studied and implemented in other schools.”

And so I have. What you are looking at below is a scatter-plot chart of the mean scores produced by 3rd graders at the 80 Polk elementary schools in 2010. The scores are plotted against the free and reduced lunch population of the children who took the test.


Boy, take a look at America, right here in Polk County. So much that I think and write about here converges in this one chart. Notice the moat between the two-thirds of schools grouped on the right and the one-third grouped on the left. That’s the moat between 55 percent free and reduced lunch at Medulla and 48 percent at Scott Lake. The gap there between the top of the poor and the bottom of the wealthy is bigger than any other gap between two schools. In this county, schools are either overwhelmingly poor or overwhelmingly wealthy. There’s almost nothing in between. In 51 of the 80 schools, 70 percent of kids taking the test were free and reduced lunch. It’s 80 percent or more in 35 of the 80 schools–just short of half.

Now let’s look at what else this chart tells us about the correlation of test scores and wealth. The line you see links Lakeland Montessori (where my son goes to school) and Wahneta Elementary. LMS in 2010 was the richest 3rd grade and had the highest scores, by some margin. Wahneta scored second lowest and had the 79th richest population. (The poorest school, Haines City’s Eastside, actually overperformed quite a bit. I’ll get to that in a moment.) The remainder of the schools cluster like grapes along the line linking those perfectly aligned poles, right through the fancy magnet, charter, and choice schools. You’ll see them on the lefthand side of the Scott Lake moat.

Yes, that’s them, hanging out well below the line of where they ought to be, judging by the relative performance of LMS, Wahneta, and the schools between. A bit ominous, isn’t it? I’ll also get to that in a second. But Stevie Wonder can see the almost rigid correlation between test score and free and reduced population. My mission for this project was to come up with a predictive numerical expression for that correlation.

This posed a challenge because, while I am not math illiterate, it is not my first language. Coming up with this factor or co-efficient or whatever you want to call it was sort of like trying to write poetry in my broken Spanish.

Thus, what I did was pretty simplistic. Briefly: I compared the spread between scores, which ran from 270 to 411 to the spread for free and reduced lunch test populations, which ran from 7 percent to 98 percent. That’s 141:91. I then calculated how many score points equated to each percentage point of free and reduced lunch population. The answer: 1.549.

From there, it was just a matter of devising, with lots and trial and error, the correct spreadsheet formula to come up with a way of adjusting scores for free and reduced lunch population. I decided to use McKeel Elementary Academy’s free and reduced population (23 percent, third lowest) as a baseline. I designed my formula to tell me how each school would perform if it had McKeel Academy’s free and reduced population.

Let’s just say the results, as you saw at the opening, were fascinating. And surprising, frankly. I hadn’t yet made any charts out of my spreadsheet when I ran this conversion. I expected the fancy schools to come in more or less in the middle; and I expected more of the traditional schools, even with adjusted scores, to find themselves at the adjusted bottom. I was very wrong.

In truth, my data shows that all but two traditional Polk schools in 2010 would have outperformed McKeel Academy if that traditional school had the benefit of McKeel’s free and reduced student body. And every single one would have scored better than South McKeel. Every. Single. One.

Only Berkley Elementary and Ridgeview Global studies, both charter or special schools of some flavor, performed worse than South McKeel when controlled for a consistent free and reduced lunch makeup. And as you can see from above, my bottom 10 included such reputational stalwarts as Lincoln and Valleyview.

I was not surprised, for the most part, by who showed up at the top. It wasn’t hard to see that Sandhill, with its 88 percent free and reduced lunch and fourth-ranked raw score, would blast any other school in the district. (That’s the gravitational pull of the late, great principal Sue Buckner at work again. More on that in days to come.) Only Lakeland Montessori, the smallest and richest of the fancy schools, seems to hold its own with the traditional schools in adjusted score. And it’s not among the best. (Again: My son is a third grader at LMS and will be taking this goofy test this year.)

So let’s review this proof:

1) There is an obvious, consistent, and predictable correlation between a Polk school’s free and reduced population and its 3rd grade FCAT mean (average) score.

2) I find that correlation to be 1.549 points of score for every percentage point of free and reduced population.

3) When adjusted for a common population, traditional schools — and a handful of poorer charters — performed significantly better than the McKeels, the Lincolns, and Valleyviews of the county.

4) Thus, the evidence suggests that faculty at traditional schools and charters aimed specifically at poorer kids are a doing a “better” job than their condo-hopping colleagues at the fancy schools.

If you are a part of the McKeel Empire, it seems you can take this information in one of three ways:

1) You can dispute it. That idiot union-loving leftist Townsend just hates us; and he’s a writer not a statistician. Why didn’t he run a true standard deviation? Etc, etc. I’d welcome this. I have long thought that scoreboard education can only remotely work if we use a school’s predictive score versus its actual score to evaluate teachers. That’s the only half-fair way to do it. The district and state should do this after every testing cycle. If you can come up with a better way to calculate the correlation, please have at it.

And I’ll give you a hint of where to look hard: how the state calculates the scores. I have not been able to determine the minimum and maximum score a school can receive. (I do have a full-time job. And at some point, as the saying goes, complexity is fraud. Testing in Florida long ago crossed that line, I think.) I suspect there may some compression of the correlation at the very bottom and very top of the scale. Start there and get back to me.

But understand this: I didn’t invent Lakeland Montessori’s 411 or Wahneta’s 274. I didn’t force their population ranks and score ranks to match perfectly at opposite poles. I didn’t invent the McKeel schools’ 356 and 336. I didn’t invent the the 75-point gap between South McKeel and the homework-eschewing hippies at LMS. There’s a greater gap in raw score between South McKeel and LMS than between South McKeel and the lowest-scoring Dundee (270). I didn’t invent Sandhill’s 342 with 88 percent free and reduced lunch or Alturas’ 315 with 83 percent. Go back and look at that first chart and then do your own analysis if you don’t like mine.

2) You can embrace it and get better. I have a friendly acquaintance associated with McKeel who is fond of asking me, in the context of teacher evaluation, “Do you believe in accountability at all?” The implication is always that McKeel-type folks believe in “accountability” while the unwashed masses–and those who teach them–just want to sleep and eat Cheetos all day. I can’t wait for this person to ask me that question again.

If you can’t refute my findings, McKeelish types, I hope you’ll take them to heart. Maybe the McKeel board, with a state legislator at its helm, can start by asking, Why exactly have we made Harold Maready the highest-paid education official in the county? Don’t we believe in accountability?

I don’t think you do, but we’ll see. I think you’ve made Maready the highest-paid education official in the county because he does exactly what you want him to. He’s rush chair for the coolest frat in town; he’s the guy working the velvet ropes letting the hot kids in; he’s the guy in the neighborhood association who measures the length of your grass.

That’s your business if you do it with your money. But you’re doing it with my money and that of the 70 percent of the population on the right side of the Medulla-Scott Lake gulch. You should at least pay somebody who has enough sense to keep his mouth shut and enjoy his pretty good gig.

If you want a principal who earns his money through educational “accountability,” you might direct him to spend a week at Alturas Elementary or Janie Howard Wilson of the Lake Wales Charter system, learning what real educational accomplishment looks like.

3) You can ignore it. This is what I expect. Lalalalalala. The various glibertarians among you will look at that chart and say, What correlation? That’s just Randian supergenius on display on the left side of Galt’s, errrr, the Medulla-Scott Lake gulch.

And you’ll believe it so fervently that you’ll also, simultaneously, reject my call to cap non-free and reduced enrollments at special schools so we can get honest evaluation. You’ll say free and reduced lunch population doesn’t matter; and there is no correlation, which is why we must keep our schools rigidly wealthy and poor. You’ll say traditional schools really aren’t getting more out of what they’ve got than the fancy schools. Lalalalala. None of that fits your narrative.

No, my experience and observation is that a critical mass of people on the left of the Medulla-Scott Lake gulch have enormous portions of their self-images bound up in thinking of the people of the right as problems and/or victims. Bullshit. The people on the right are strivers, every bit as much people on the left–and, numbers suggest, even more than the people on the left. The teachers who help them strive are doing a good job. They can always improve. But by any objective reading, they’re doing better than anyone thinks. As a class, they are outperforming their peers on the left.

I’m going to write more about this in the coming days. It’s goes to the heart of much what’s happening in today’s America.

But for now, I’m telling you straight up, McKeelian, that the teachers and staff at Alturas are getting more out of their kids than you’re getting from your relentlessly-screened, precious little display-window children. Do the work, like I have, to refute it.

If you can’t, remember who said this: “If we do not recognize there is a problem, then there cannot be a solution.”

If you value at all what you say you value, these numbers will cause you to look in the mirror and ask yourself, Who’s really a problem?

Answer honestly.

Creative Commons License photo credit: Chuck Welch for Lakeland Local

One thought on “Polk’s “Best” and “Worst” Schools: Not Who You Think

  1. Another article that backs up this research. http://www.huffingtonpost.com/robert-niles/public-schools_b_1002466.html?ref=fb&src=sp&comm_ref=false

Comments are closed.