Sunday, November 16, 2014

Seattle Schools Enrollment Numbers: Hard to Reconcile with Reality

A guest post from Kellie LaRue and Meg Diaz in two parts.  My early comment is that these are two parents who know how to analyze both data and lack of data.  They have done their homework.  But like Charlie and I have said, you are only as good as the data that you can get from the district.  But we try.

Part One: Enrollment Data and Public Information

What seems like a reasonable way to gauge who might have erred is to look at the rest of the available enrollment data, sparse though it may be. If the rest of the data is pretty consistent and lacking in anomalies, then it seems reasonable to conclude that the district got it right, and Garfield had an unusual counting flub. 

But first, a second to rant about the incredible scarcity of enrollment data. 

Enrollment data directly drives hundreds of millions of dollars of public funding. It is important that it be both accurate and public. Right now, it is neither. 

  • Enrollment numbers are directly tied to staffing throughout the district (well, everywhere but JSCEE). 
  • Analysis of enrollment numbers is essential for capacity management and capital projects. 
  • Bad enrollment data will obviously result in erroneous conclusions, and misspending of public funds.
  • Detailed enrollment data, which used to be routinely put up on the district website every month, and directly influences how hundreds of millions of dollars of public dollars are spent, is no longer posted on the district website (because of a software change, according to the Public Records officer), and data has been removed from the district website. 
Okay, rant over. 

So, does the available data tell a pretty consistent story, without any major contradictions or anomalies? - Nope. 

The data is weird. Very, very weird. There are two major sources: Friday memos, and the enrollment data SPS files with OSPI. 

Let’s take a look at the OSPI data first. 

The above chart shows the enrollment change between September and October for all high school grades. Almost invariably, high school enrollment in every grade rises between September and October. Only twice since 1997-98 has any high school grade dropped between September and October. In 2006-07, the 12th grade lost 3 students between September and October (it’s a teeeeeny red bar, but it’s there, I promise). I mean: WOW. MIND THE 3 KID GAP.

And then this fall, between September and October, 566 9th graders vanished. 566 kids gone. Poof. What happened? Did they drop out? Join a cult en masse? Get abducted by aliens? Was there some kind of SPS 9th grader rapture event?

So. The numbers SPS filed with the state are extremely peculiar. But surely the Friday memos, reporting to the board and Superintendent, show typical enrollment patterns? Again: NOPE.

In September, district administrators reported that there were 14,000 HS students, and 52,068 students in the district as a whole. In October, district administrators reported that there were 13,786 high school students, and 51,988 students in the district as a whole. So, minus 80 students district-wide, and minus 214 high school students.

High school enrollment has never, ever dropped between September and October. Not even once. The state funds districts using the October count because October is almost always the highest enrollment for the entire school year.

How… odd. As mentioned earlier, the person most experienced with enrollment counting for SPS retired in June (we miss you, Tracy!).

My assessment? 
The best case is that district administration made a mistake with Garfield, and probably with other high school counts. 
But there could be a very reasonable case to be made for district administration lowering high school numbers intentionally.
Given the severely limited data available, it’s hard to estimate the scope of the mistake, but it very likely includes a miscount at Garfield High School.

Is it worth asking if someone at JSCEE intentionally altered the numbers? Yes, it is.
While there is some seriously dubious SPS history that could make my Magic 8 Ball come up with an answer more nefarious than “ERROR,” I doubt that this is a case of intentional alteration. Because if this is intentional alteration, it is really, really sloppy work. It’s terrible. My kids could do better.

But a stubbornly uncorrected mistake has the same result as an intentional alteration.

So will the district review their conflicting, strange numbers? Will they consider when there are multiple instances of anomalies, that they might have been the ones to make a mistake?

Or will they continue to stonewall, conceal and/or remove enrollment information, cut a teacher from Garfield, and insist that the disruption to student schedules will be lessened because the teacher will be removed at the end of the semester… even though the overwhelming majority of core classes are full-year classes?

Based on current behavior, I’d say they’d rather do anything rather than admit a mistake – even one that screws over students they are responsible for educating.
 One teacher matters at every school. At the high school level, removing a teacher immediately makes it more difficult for students in that building to acquire credits for graduation, for college admissions and for college scholarships. And given the anomalies in the data available, I have serious concerns that Garfield is not the only school the district flubbed the count for.

Data sources:

OSPI 1251 enrollment form for 2014-15, headcount (I also looked at the FTE forms)

OSPI 1251s relating to SPS, from 1997-87 to 2014-15


Pro-sleep Mom said...

Whoa. If we are missing 566 kids (or maybe more, since the trend is usually up rather than down), that represents around $4 million- seriously real money.

We have 13 high schools where attendance is taken everyday and transmitted to the District. How hard can it be to get real numbers? How could the staff not want to make this right??

Melissa Westbrook said...

Pro-Sleep, I'll have my own assessment thread on why this is happening this week.

Anonymous said...

All that matters for these counts are how many kids are ENROLLED, not necessarily in attendance, on October 1st.

This discrepancy is beyond bizarre, and certainly doesn't mesh with what I heard at Hale, that MORE 9th graders showed up than expected, not fewer. Classes are very full.

I was wondering how many 9th graders are usually in flux during the wait list moves? Maybe somehow those who left a school were subtracted out, but those who moved to a new school (before Oct 1st) were never added into the totals for those schools (perhaps due to a Central Office glitch?

Would it be possible to find wait list movement numbers from previous years to see if this adds up to 500-ish students?

- North-end Mom

Anonymous said...

If anything, it seems like the motivation would be to inflate the numbers in order to increase the amount of state funding to SPS. The problem with not sharing information is that it can turn rational people into conspiracy theorists.

I doubt this is the result of foul play; in fact, assuming that the numbers are actually wrong, my bet would be that SPS staff is trying to make the best decisions they can given the situation.

The conspiracy nut-case in me could possibly see motivation for not amending an honest mistake if these students trigger some other threshold that requires significant investment. For example, if these students require SPS to add more building capacity immediately it could be an expensive issue. Or a situation where SPS is required to bus kids to schools with existing capacity like Rainier Beach High School. This would upset a lot of families and increase transportation costs to the district.

Posting as Irrational Conspiracy Theorist today.

Lori said...

While that spike does look suspicious and is clearly unprecedented, I wonder how many of those 566 missing 9th graders are families that went private for high school but didn't officially un-enroll? Thus, they were on the roster in early September, but ultimately removed after not showing up all month.

I had a kid at Lincon for 5th grade last year, and I know a lot of families went private for middle school due to concerns about capacity, and these weren't all people who would have gone private no matter what. Lincoln families were asked repeatedly to please un-enroll by June if going private, but maybe 8th grade families weren't asked to do the same? Plus, if you're in a program like Spectrum or APP and unenrolling means you lose that designation, some people may have decided to remain enrolled just in case they changed their minds come September.

From my own limited vantage point, I think more people are exploring private school to avoid having their kid have to leave a school to go start up a new one and to avoid the possibility of split shifts/year-round high school and so on. I suspect there are also families who go private for middle and typically come back for high school, but maybe they too have capacity concerns so they're just staying private for high school now too? Or what about attrition to Shoreline, like northernmost families used to do for middle school? Anyone paying attention knows that the capacity crisis will be breaking the northend high schools very soon.

I completely understand how and why projections are based on past experience, but the protracted and largely unresolved capacity challenges could very well be impacting families' decisions, particularly at entry points like 6th and 9th grade. I doubt it explains 566 missing 9th graders might it might account for some of it.

Anonymous said...

I would be surprised if private school enrollment played a statistically significant roll in the huge enrollment drop between Sept and Oct. To my knowledge, there has not been a large increase in the number of private/parochial seats available, and almost all are running at full capacity already.

Though I do not doubt that applications from Spectrum/APP to private is on the increase due to instability in the program, the number of private seats are still fixed. If an APP kid is kicking out a formerly private kid, then that private kid needs to go somewhere. Plus these seats have been held onto in September in years past, so that should already be reflected int he numbers.

It does raise the question, though, if the surrounding districts such as Shoreline or Mercer Island have seen an uptick in enrollment, particularly of students with a Seattle address. I wonder if this data is available.

DataHound Too

Anonymous said...

Makes me wonder if someone who is hoping to be promoted to Tracy's position is unwilling to ask for help, or admit a mistake.

Using the equivalent of the 'VAX' excuse is not credible.


Anonymous said...

Kellie and/or Meg,

Thanks for your work on this. One of you mentioned previously that you had contacted the district about the odd numbers. Are they standing by them, or have they been unwilling to engage with you on the issue?

What, if any, are next steps other parents can take to help get some clarity?


Po3 said...

I have worked w/ numbers for much of my career and whenever a number does not make sense you investigate and either find an error and correct or provide a explanation for the anomaly.

Obviously, nobody bothered to investigate before publishing these numbers.

So the questions is: what is the district doing now to provide an explanation for the historic drop in 9th graders this year?

cmj said...

Thank you for all of your hard work, Meg and Kellie!

I suppose that some parents are motivated by various scandals/crises in the district to put their kids in private school. How many? I don't know. If many parents were doing that, we could expect to see a flood of students leaving SPS before the 2003-2004 school year, after the 2002 $35 million budget shortfall. Instead, the number of high school students increased, as it usually does.

Anyone know what happened in 2005 or late 2006 so that the number of high school students across all grades dropped or barely increased, instead of increasing as usual?

Lori said...

Not to belabor the point, but private schools in and around Seattle draw from a very wide geographic area. So yes, the number of seats is relatively fixed, but the schools could increase their acceptance of SPS students and "displace" students from Bellevue, Mercer Island, Shoreline, Edmonds, etc. You might be surprised how far families are willing to drive for MS and HS!

I do know that one private school that historically accepts more students than it has seats for because they know that some percentage will decline the spot was suprised to see that that didn't happen for this year. A larger-than-expected number of families accepted 6th grade spots, and last I'd heard, they were trying to figure out how to accomodate them. Another school remarked on the surge in applications from APP students for this current school year across the grades.

I guess my larger point is that we've likely reached the point where capacity challenges are now a confounder to projecting enrollment. People's behavior may be changing and adding flux to the projections. If I were SPS, I'd be analyzing if/how attrition to private school changed this current year, although I guess when you are overcapacity system-wide, you don't care if/why people leave. In fact, you're probably happy about it.

Anonymous said...

The theory I would want to explore is that the method of counting has changed between September and October. There's a hint of this theory in the statement about Running Start students (though I don't think that should apply in 9th grade, right)?

The only rational explanation for a 100's of students disappearing in a month is that one of the numbers is substantially different from the other, i.e. counting oranges/apples (categories of students not being counted, students being classified in different grades, . . . .) or an error of some sort. The first doesn't have to be an error -- if, say a class of students were being erroneously counted twice and the numbers being correctly all at once.

Hope someone does come up with the right numbers, because otherwise, this kind of thing does spur conspiracy theories and lack of trust. I for one am imagining groups of kids disappearing in movie-plot worthy ways and wont' be able to shake those images without better numbers.


Anonymous said...

The theory I would want to explore is that the method of counting has changed between September and October. There's a hint of this theory in the statement about Running Start students (though I don't think that should apply in 9th grade, right)?

The only rational explanation for a 100's of students disappearing in a month is that one of the numbers is substantially different from the other, i.e. counting oranges/apples (categories of students not being counted, students being classified in different grades, . . . .) or an error of some sort. The first doesn't have to be an error -- if, say a class of students were being erroneously counted twice and the numbers being correctly all at once.

Hope someone does come up with the right numbers, because otherwise, this kind of thing does spur conspiracy theories and lack of trust. I for one am imagining groups of kids disappearing in movie-plot worthy ways and wont' be able to shake those images without better numbers.


Anonymous said...

Thanks, Kellie and Meg, for day-lighting this.

I took a look at the OSPA 1251 form. All grades, K-12, except for 9th grade showed GAINS in enrollment from Sept to Oct 2014.

With the trend showing gains at all but one grade level, you would think a "loss" of 566 9th graders between Sept and Oct would have gotten someone's attention!

- North-end Mom

Po3 said...

A loss of 566 9th graders is over 30 students per high school. Seems like a simple email asking if any school experienced this level of drop off?

Meg said...

I should be clear: I don't think SPS lost 500 9th graders - I think this year's numbers are simply wrong, especially at the high school level.

I think the 500 student loss is an example of an error - a big, dramatic error. Given that the rest of the (very limited) data is filled with problems, it's hard to believe that the district has high school counts right at any high school in the district.

Anonymous said...

The only logical explanation is an alien abduction.

-reality check

Maureen said...

Meg and Kellie, thank you so much for your work on this! I'm sorry if you have already addressed my question.

You have reported the change in enrollment between September and October. I'm sure you've examined the levels as well. Can you tell if there is anything odd about either the September or October level of 9th grade enrollment this fall compared to other years? In particular, does it look like more 8th graders showed up for 9th grade as compared to previous years? I'm trying to figure out if the issue is with the September count or with the October count or a combination of the two.

Ingraham had fewer 9th graders counted in October than had been projected in the spring, but I don't know what the September numbers were compared to either. Should I try to find out?

Anonymous said...

They did change how they count 9th graders/10th graders. Where they formerly counted them based on credits earned to "cohort" year (apparently because of the changes in testing from HSPE to SBAC). Could this adjustment be the source?
-10th grade teacher

Melissa Westbrook said...

So the questions is: what is the district doing now to provide an explanation for the historic drop in 9th graders this year?

Answer: they are not because no one realizes it has happened.

Send this to the Board and ASK them to ask. That's your best bet.

Anonymous said...

So perhaps the acceleration of APP kids in middle school has changed how they are counted based on credits they are coming into high school with? Is this true?


Anonymous said...

APP students don't come into HS with any credits. I I don't think APP is the cause behind the odd numbers. The only course an APP student can take that puts them in any different position come high school is Biology in 8th grade. They take the EOC and get credit in terms of meeting the prerequisite for other courses that require Bio first, but they don't get actual high school credits for that course. For other APP classes, "2 years ahead" is just a nice phrase, not really corresponding to reality or having any meaning.


Anonymous said...

Isn't it possible to receive HS credit for HS-level classes taken in middle school...as long as the instructor is HS certified?

Though, 566 incoming 9th graders with enough credits to be classified as 10th graders (or higher?) seems like a stretch.

- North-end Mom

Lynn said...

North-end Mom,

If you look at this High School Enrollment Report you'll see (on page two) that last year 3,917 students were enrolled who had earned less than 5 high school credits. Page four shows us that there were only 3,298 students enrolled in their first year of high school. After Policy 3240 was amended in July, only 3,298 of those students would be counted as freshman. It looks like enrollment services didn't take the policy change into account until after reporting September enrollment.

Anonymous said...

OK. The cohort -vs- credit piece makes sense, but I don't see anything in Board Policy 3240 (discipline policy) that addresses how students are classified, but maybe I missed something?

- North-end Mom

Anonymous said...

@ North-End Mom,

I think it's actually policy 2420 that's relevant.

It includes this:"This revision will promote students to the next grade level based both on credit earning or the number of years they have been enrolled in high school.... By modifying our high school promotion criteria, students will advance to the next grade level each year, which provides scheduling predictability, increasing access to additional assessment pathways, and improving the ability to provide students most in need with support services."

So perhaps the original projections for the number of 9th graders were based on the old definition (# of credits), but using the new definition some of those could be moved to 10th grade with their cohort? Same would happen at higher grades, too.

Half Full

Josh Hayes said...

If it IS this classification issue, then shouldn't there be a concomitant bump in the number of 10th-graders? Shuffling kids from one category to another could produce a drop in one with an increase in another, but a drop with no associated increase suggests those kids just fell off the map somehow (unless the 10th grade count already included them in September, but if that's the case, shouldn't the September 10th-grade count be dramatically higher than in previous years, and it isn't?).

Anonymous said...

If a 9th grader were then classified as a 10th grader, wouldn't we see this number go up for 10th grade, like a mushroom at that grade level? Also, they would all still be considered enrolled as students so I am not clear on how this would negatively impact the overall enrollment numbers of a school. They would still be in high school, just another grade, right?


Anonymous said...

We would only see a bump in the next grade level if only 10th graders were designated as ninth graders. How many of the 566 were in 11th or 12th?

That is IF the numbers are exclusively the result of that particular policy shift. There maybe a combination of factors that contributed to the data change.

-never assume

Anonymous said...

Add the running start count change to the cohort not credit change and you might have an answer.

-perfect storm

Lynn said...

Sorry about the policy number error - aging eyes and a tiny screen aren't a good combination.

Total high school enrollment increased by 203 students between September and October of this year.

I agree Fedmomof2. I don't think this affected the total enrollment numbers for Garfield. I'd love to see the school by school head counts that make up the September totals reported to OSPI.

Po3 said...

The problem is the math. The math is wrong. The math needs to be corrected. Once the math is corrected they can use the numbers - until then - they can't.

Why aren't they correcting the math?

Anonymous said...


I second that! It would be so nice to be able to see what is going on at the school level (as was possible in previous years).

According to the OSPI form 1251, there were the following changes for total high school (per grade) between Sept and Oct:

9th -566
10th +232
11th +295
12th +242

(net increase of 203 from Sept to Oct)

The policy change didn't affect just 9th graders, so even if 566 kids who were previously classified as 9th graders (due to insufficient credits) are now classified as 10th graders with their cohort, the 10th grade number wouldn't necessarily increase by the full 566, because there would be a subset of kids who were previously classified as 10th graders (due to insufficient credits) who would now be classified as 11th graders, and so on.

It doesn't seem like this would dramatically change the budget situation, but it seems like changing the way kids are classified could be a big problem for enrollment planning.

I hope they go with cohort numbers, instead of what is reported to OSPI (which would be a mix of the old and new reporting systems) when planning for enrollment/capacity, especially if they use averages over a 5 year span...otherwise, that decrease of 566 9th grade students (on paper) is going to throw things off for a long time.

- North-end Mom

Anonymous said...

It doesn't have to create a mushroom at 10th grade, as the same thing would happen in older grades.

Say, for instance, the old "credits only" approach said there would be 4000 9th graders--b/c there are are 3500 incoming 9th graders and 500 current 9th graders who don't have enough credits to be considered 10th. Once you switch to the cohort approach, these get counted as 10th graders. But there's probably an similar number of 10th graders in the same boat, so when you shift them to 11th graders under the new system the 10th grade numbers are still fairly similar to the pre-adjusted number. Same would happen at the higher grades. The fact that the total number of 10th, 11th and 12th graders grew more than is typical during this same Sept-Oct data period seems to support this theory, doesn't it? A lot fewer 9th graders, and a smaller increased in EACH of the upper three grades? It's not unusual to see data anomalies like this when category definitions change.

If this is the explanation, it's disappointing to see that such a large number of kids each year don't pass the number of credits they need...

Half Full

Anonymous said...

Following up on my previous post.

If the Oct 1st 9th grade headcount (3521), as reported by OSPI represents the 9th grade COHORT size, then there was an INCREASE of 223 9th graders this year (district-wide). This is as compared to the 9th grade cohort size of 3298 given in the SPS enrollment report for 2013-14.

However, if you look strictly at the OSPI counts for 9th graders from the 2013-14 OSPI form 1215, which states there were 3916 9th graders on Oct 1 2013, and compare it to the OSPI 1215 form data for October 1st 2014, which gives 3521 9th graders in SPS, it would appear that SPS 9th grade enrollment DROPPED by 395 students this year.

All of us who have been lobbying SPS for more capacity since our now 9th graders were in kindergarten know that there could not possibly be a DROP in this year's 9th grade cohort size!

If the enrollment discrepancies are due to a change in how students are classified and reported to OSPI, this helps to explain why it is IMPOSSIBLE to get the attention of the Board and/or SPS staff when it comes to the need for more high school capacity.

Cohort sizes are GROWING, not decreasing! Really, they are!

- North-end Mom

Anonymous said...

This thread is so depressing. Has anyone directly asked a specific person in enrollment or finance - not sure the correct department, to explain the numbers? If so, who is that person - can anyone with knowledge post? If no one has asked directly can someone more knowledgeable than me do so? If someone has refused to answer, would we be able to get information through a public disclosure request?

I am not trying to stir up trouble at JSCEE but it does seem that the shortest distance to sanity here is for official information, with a name attached, on how enrollment was figured this year and clarification on if/how the formula did/will change.

Surely the district by polite inquiry or something more forceful if absolutely necessary owes the public this answer? We are the taxpayers after all, right?


Meg said...

I've asked. Several board members have responded to me. But since the start of November, the only staff member from JSCEE who has replied to me has been the public records officer.

kellie said...

Thank you everyone for your contributions to the conversation. Some of the ideas and links have really helped to inform my thinking on this.

It is very possible that there is a good reason for the enrollment counts. However, the only thing that is clear is that there is multiple conflicts across all the various data sets and nobody seems to have a compelling narrative that links all the data points.

It would be better and easier for everyone involved if the same level of historical information that has been published on the SPS website were just published so that folks can be discussing facts, rather than conjecture and the gaps in the fact set.

For example, in the past a simple spreadsheet that noted all of the teacher add's and subtracts's from the time of budget until the final October adjustment was just available. This helped everyone to have confidence that we were all at least discussing the same numbers.