Thursday, October 1, 2015

KPREP Results for Student Subgroups (Facing Some Brutal Facts)

First, a snapshot of results for different student subgroups using the newly released 2015 K-PREP elementary school results, and then a little explanation.
Explanations for three terms used above:
  • Students are counted in the gap group if they are African American, Hispanic or American Indian/Native Alaska, or have  identified disabilities or limited English proficiency or eligibility for free or reduced-price meals. The gap group is a way the Kentucky Department of Education shares combined data on multiple groups we know have been historically under-served by Kentucky's schools. Naturally, some students belong to more than one of those groups, but those students are only counted once in the group score.  In the tested elementary grades, they're about 67% of all students.
  • The students not counted in the gap group are the the other 33%. They don't have disabilities or limited English proficiency or low family incomes that qualify for free or reduced-price meals. They're also white, Asian, Hawaii native or Pacific islander, or of two or more races.
  • The weighted average combines all the KPREP scores at the proficient and distinguished levels. Reading, math, and social studies results each count 25%, with writing counting 20% and language mechanics 5%. That's the formula the Department used to calculate the gap group component of this year's overall scores.  Using that formula generates one number for thinking about what's happened to each group of students.
The weighted average lets some patterns pop out, like these:
  • Students in the "gap group" are 30 points behind those not in that "group."
  • Students with limited English proficiency are another 20 points behind the "gap group."
  • Students with disabilities, African American students and Hispanic students also score below the gap group, though Hispanic students are quite close.
  • Students with free/reduced meal eligibility have results essentially identical to the gap group result, for the obvious reason that those students with low family incomes hugely outnumber all the other subgroups.
A similar pattern –but with even worse gaps– appears at the middle school level.

The pattern worsens again at the high school level. At this level, reading, math, social studies and science each count 20%, writing 16%, and language mechanics 4%: the formula used for high school Gap Group reporting. For 2015, only high schools have reported KPREP science results.
Look hard at that last graph. 

Using this weighted average approach, we delivered proficiency for the high school students we serve best, the ones not counted in the gap group:
  • at more than twice the rate we delivered for African American students
  • at more than four times the rate for students with disabilities
  • and at more than six times the rate we delivered for students with limited English proficiency.

Overall Scores Rise for High Schools, Not for Lower Levels

In Kentucky's Unbridled Learning system, overall scores are the quickest summary of results for a public school, district, or the entire state. An overall score combines multiple measures to calculate a single number on a 0 to 100 scale that sums up student and program performance.

For our state as a whole, the high school overall score rose 1.5 points from 2014 to 2015, but the elementary and middle overall scores declined.

Below, you can see the main components that went into the changes at each level.

At the elementary level:
  • Achievement indicates KPREP scores for all students in all tested subjects.
  • Gap Group shows KPREP results for students in historically under-served subgroups: students who are eligible for free or reduced-price meals, have limited English proficiency, or have identified disabilities, along with African American, Hispanic, and American Indian/Native Alaskan students.
  • Growth reflects how students' reading and math scores this year compare to scores for students with similar results the previous year.
  • Program reviews reflect schools' analysis of the quality of their programs for primary students, arts and humanities, practical living/career studies, and writing.

At the middle school level, those four indicators are combined with:
  • Readiness results  from the grade 8 Explore assessment.  

At the high school level, the four elementary indicators are combined with:
  • Readiness results from indicators that include ACT and a battery of other tests to show that students are ready for college and/or career, with bonus points if a student is ready for both.
  • Graduation rates (using a four-year rate for 2013 and five-year rates for 2014 and 2015).

Graduation and Readiness Continue to Rise

Kentucky's statewide graduation rate showed continuing improvement, and the readiness of those graduates moved up substantially during the 2014-15 school year, as shown in results released at midnight by the Kentucky Department of Education.  This is important good news for Kentucky's Unbridled Learning commitment to college/career readiness for all!

Statement on 2015 Accountabilty Results

Here's the full text of the Prichard Committee statement released today:

The accountability results released this morning show strong growth in the college and career-readiness of Kentucky’s high school graduates, moving from 62.5% to 66.8% of graduates reaching those benchmarks. This good news is coupled with the fact that scores for high school students in groups who have historically struggled to meet state standards have improved at a quicker pace than the achievement results for all students. These indicators are positive news for Kentucky’s students and our shared future.

However, the decline in overall scores for elementary and middle school is cause for immediate concern and focused attention. Student outcomes in the early grades must continue to improve as they lay the essential foundation for later success. As a combined group, African-American, low-income, Hispanic, English-language learners, and students with disabilities also lost ground at the elementary and middle school levels, showing that we need to deepen our focus on providing richer opportunities for each and every child.

In addition, this year’s results fell short of some of the goals Kentucky set for educational improvement. Our statewide elementary and middle school results are below the goals set by the Kentucky Board of Education. While the high school outcomes met these goals and college and career readiness continues to increase, it is important that we recognize the weaknesses in other areas and actively build more consistent year-over-year improvement going forward.

The Prichard Committee plans further study of these results and urges all Kentuckians to renew our focus on making sure all students learn deeply, thrive, achieve, and contribute to our communities. The Committee’s Achievement Gap Study Group, representing participants from across the state, is working this fall to identify the most important next steps to support each and every Kentucky child’s academic growth and achievement. We applaud the Kentucky Department of Education for recognizing the moral imperative of ensuring all students achieve at high levels. It is clear that we have urgent work ahead.

The Prichard Committee for Academic Excellence is an independent, non-partisan citizens’ advocacy group. Since 1983, the Committee, made up of volunteer parents and citizens from across Kentucky, has worked tirelessly to improve education for Kentuckians of all ages.

Wednesday, August 26, 2015

Prichard Committee Statement on 2015 ACT Results

Here's the full text of the Prichard Committee statement released today:

The ACT results released today by Kentucky Department of Education show positive trends for the 2015 graduating class overall, but raise concerns about whether African American, Hispanic, and American Indian students are receiving the full support they need and deserve.

The good news is that members of the 2015 class as a whole looked stronger than their 2014 peers. In English, mathematics and reading, more of them met Kentucky’s benchmarks for college readiness. Progress in mathematics was especially strong, with an increase of more than 4% in students who are ready for credit-bearing college work.

For some student groups, however, the results were clearly not strong enough. Students from different racial backgrounds still have quite different results. Looking at ACT composite scores, 2015 African American graduates did no better than their 2014 peers, and Hispanic graduates did slightly worse. American Indian students improved just slightly faster than white students and not at a pace to close the big gap between the two groups. All three groups have results showing them less ready for college success than their white and Asian classmates.

The Kentucky Board of Education has already added important new efforts to reduce novice performance in all subgroups, signaling that it is time for deeper, more sustained work to end these gaps. To support renewed and expanding work on increasing achievement and closing gaps, a Prichard Committee study group will be working through the fall to analyze data, policy and practice, with the aim to share recommendations at the beginning of 2016.

For Kentucky to flourish, we need for Kentucky students of every background to reach their potential and join in building a strong, shared future. While progress is evident overall, important student subgroups still lag behind, demonstrating the need for more concerted efforts to close these gaps rapidly. Our goal must be for all Kentucky students to graduate from high school truly ready for adult success.

ACT Readiness Moves Forward, But With Important Gaps

The Kentucky Department of Education released ACT results for 2015 public school graduates this afternoon, showing a sturdy trend for students overall but raising concerns about how well we're serving our African-American, Hispanic, and American Indian students.

First, here's the overall good news, with upward movement on the percent of students reaching all three of the college readiness benchmarks set by the Council on Postsecondary Education.

Second, here's a look at the trends for subgroups, this time showing composite scores (combining English, math, reading, and science) rather than percent meeting benchmarks. Highlighted, you can see that:
  • African American students made no gains compared to last year
  • Hispanic students lost one-tenth of a point
  • American Indian students gained three-tenths of a point, growing a bit faster than students overall, with results still far behind their white classmates

Results for students with disabilities and students who receive free or reduced-price meals were not included in the press release but will be included in the 2014-15 school report cards now scheduled for release in early October.

Friday, August 14, 2015

Readiness Gaps: Important Work Ahead

While we can take pride in the work that has brought Kentucky to a 62.5% overall rate of demonstrated readiness for college and/or career for our 2014 high school graduates, we still have lots of work to do, both to raise the overall readiness rate and to close some major readiness gaps. Here comes a frank look at the basics of those gaps, in the form of two charts and some comments on each.
Here, the painfully low readiness rates for students with disabilities and with limited English proficiency are first to catch the eye, but the rates for students receiving free and reduced price meals are also bad news.

The gap group is pretty much identical to the free/reduced meal group, which is pretty much to be expected. Within the gap group, the free/reduced students hugely outnumber the other groups, including students with disabilities and limited English proficiency, along with African American, Hispanic, and American Indian or Native American students.

One more note: I've estimated the results for four advantaged groups: those without disabilities, without limited English proficiency, without free/reduced meal eligibility, and not included in the gap group. Most data for this post can be found in the state's school report card, but these better-served groups are not shown. Still, with a bit of multiplication and subtraction, it's possible to get quite close to what those results must be. They're marked with asterisks to show they did not come directly from the Department of Education.

Next, these are the 2014 rates by students' ethnic background.
White and Asian students are clearly being better prepared for adult success than the other student groups, showing another challenge we face if we want all students to reach the readiness they need.

A second, uncomfortable thing needs to be said out loud: the results for African-American students are lower than those for the free and reduced-price meals group, showing that something more than the challenge of family low-incomes has been going wrong for those students. There's something we're doing or leaving undone for those students, beyond the main economic challenge. That's a distinctive problem that we need to look at frankly and tackle with vigor.

More broadly, each of these gaps are about student possibilities not yet realized and adult contributions that may be lost as a result, weakening all our communities as well as their individual opportunities. Changing that pattern and ending those losses will be important work we must take on.

Monday, July 27, 2015

How People Learn: More on Synapse Development

Last week's post on synapse development drew a comment with questions:
I wonder what the research shows for different age groups, and if the results vary, regarding synapse development. I wonder how long synapse development continues!
From How People Learn, here's some background on those issues.

There are two different patterns to how we develop synaptic connections.

In one process, "synapses are overproduced and then selectively lost." That process is especially in early development, and the time it takes varies depending on the part of the brain, " from 2 to 3 years in the human visual cortex to 8 to 10 years in some parts of the frontal cortex."

The other process lasts all  the way through life, and involves adding new synapses as one adds experiences.

Noting the two different processes seems to fit the simultaneous ideas that the early years especially important and yet learning is a lifelong process.

Thursday, July 23, 2015

How People Learn: Synapse Building For Rats (And Humans)

Continuing in my summer book study...

In the book How People Learn, the chapter on "Mind and Brain" is heavy on neuroscience, starting with the essential process of synapse development.  Our synapses are connections between neurons, and substantial research shows that learning occurs through synapse development.  Much of the chapter is about studies of what does and does not produce rich synaptic development in lab animals.
One group of rats was taught to traverse an elevated obstacle course; these "acrobats" became very good at the task over a month or so of practice.  A second group of "mandatory exercisers" was put on a treadmill once a day, where they ran for 30 minutes, rested for 10 minutes, then ran another 30 minutes.  A third group of "voluntary exercisers" had free access to an activity wheel attached directly to their cage, which they used often.  A control group of "cage potato" rats had no exercise.
Researchers then examined the rats' brains, looking both for blood vessel development and for synapses per neuron, and found that both sets of exercisers had higher density of blood vessels than the acrobats and cage potatoes, but
But when the number of synapses per nerve sell was measured, the acrobats were the standout group.  Learning adds synapses; exercise does not.
What struck me in this was something I'm not sure the the authors meant me to notice: I heard the word "exercise" in its classroom context, as meaning an activity assigned by the teacher, often with repetitions and an emphasis on speed, like spelling lists and sets of arithmetic problems.

I wonder how many of our teaching traditions reflect the idea that the brain is like a muscle and will build through steady repetition that truly resembles physical exercise.

More than that,  I wonder how much we will need to change if we want learning that happens as developing sets of synaptic connections.  That understanding suggests that the some of the most important work comes in the opportunities to "put things together " and "see how it all connects."

It seems likely that learning of that kind will require fewer drills and more exploration, fewer lists and more reasoning about how different elements relate, fewer details and more depth on key organizing concepts than we have expected in the past.  That does not have to  mean no drills, no lists, and no details.  It does mean realizing that exploration, reasoning, and organizing concepts must be given a rich share of the time and energy students bring to their learning.  And it does mean that the number of drills, lists, and required details has to be restrained to allow the richer elements opportunity to occur.

Thursday, July 16, 2015

Kentucky School Staffing (National Comparisons)

In the fall of 2012, Kentucky enrolled 1.38 percent of all students enrolled in public schools nationwide in pre-kindergarten through grade 12.

Our share of public school staff was at or below that 1.38 percent level in three categories, with Kentucky having:
  • 1.02 percent of student support staff nationwide
  • 1.21 percent of administrative support staff
  • 1.36 percent of officials and administrators
  • 1.38 percent of teachers
Our share of public school staff was above the nationwide level in the other categories, including:
  • 1.46 percent of guidance counselors
  • 1.48 percent of Instruction coordinators
  • 1.87 percent of Instructional aides
  • 1.94 percent of principals and assistant principals
  • 2.09 percent of school and library support staff
  • 2.10 percent of other support services staff
  • 2.33 percent of librarians
If instead, Kentucky schools and districts had consistently had 1.38 percent of each kind of staff, we would have had:
  • 1,021 additional student support staff members
  • 325 additional administrative support staff
  • 136 more teachers
  • 10 more officials and administrators
  • 7 fewer guidance counselors
  • 69 fewer instruction coordinators
  • 443 fewer librarians
  • 955 fewer principals and assistant principals
  • 2,028 fewer school and library support staff
  • 3,560 fewer instructional aides
  • 8,225 fewer other support services staff
Back in March 2009, I posted a similar analysis using Fall 2005 data. As I wrote then:
I’m not arguing that Kentucky should staff schools to those averages. There may be important benefits to what we do differently, and our students may have different needs. I do think, though, that this is an interesting mirror to look in, inviting us to think about how we currently staff public education.
Coming back to this analysis this time, I still see that issue, and I have these added thoughts:
  • We have 1,087 librarians spread over more than 1,200 schools. That may be the starting example of where our added commitment is a good idea, especially as we ask students to go deeper on research, designing their own investigations, and learning through major projects. 
  • We’re now asking our principals to do sustained observations and give thoughtful feedback for every teacher: for that big growth in responsibility, our added numbers may again be just right.
  • Other support staff seem likely to include food workers, custodial workers, and bus drivers. In other states, that work is often handled by contracting companies, and it's possible that Kentucky isn't so much engaging more workers as engaging them in  a way that shows up under staff rather than service fees.
  • I'd love to know what other states are doing (and Kentucky apparently isn't) in student support services!
Source note: the data for this analysis comes from the Digest of Education Statistics, using tables 203.40 and 213.20 The staff analysis is based on full-time equivalent positions.

Wednesday, July 15, 2015

How People Learn: Learning That Transfers To New Contexts

Continuing my summer book study...

"Transfer from school to everyday environments is the ultimate purpose of school-based learning."  That's the kind of statement that seems obvious and turns out to be important. In How People Learn, the transfer process gets close attention--and now it's got mine as well.

Here, transfer is about ability to use knowledge in multiple contexts. For example, veteran shoppers can be very good at figuring out cost per unit and identifying bargains, but struggle with related division when dropped into a formal classroom.  Conversely, most of us have watched kids part way through elementary school who aren't at all sure which bit of "school math" to use in stores. The studies in the book give examples like seeing if Latin or computer programming develops logical reasoning for other kinds of work and sorting out which kinds of simulations create lasting and useful understanding. If you've learned something that you can only used in the situations that are most like being in school, it isn't going to be a lot of help for other kinds of challenges.

With research citations for each claim, the chaper looks at what scientists know about when transfer is and isn't likely to succeed. Some major points:
  • Time spent on understanding how a process works and when it matters yields better transfer than memorization.
  • Teaching a set of knowledge in multiple contexts makes the learners more able to transfer it.  
  • Transfer is also improved when students are equipped to monitor their own understanding and evaluate their own progress (with "metacognition" as the power word for that process of learning about their own learning.)
All through this section, I was haunted by images students doing worksheets and computer drills to prepare for a math assessment. 

The research in this study gives support to parent concerns that a certain kind of "teaching to the test" creates knowledge that will be only useful on the test. That can be learning for a single context, focused on procedural accuracy, with little insight into how or why the same knowledge could be put to work elsewhere.

Plus, what happens if the school's response to early difficulty is more of the same kind of drill, and more, and more and more again? On this understanding of learning, students may succeed on this year's test, but not be able to transfer that knowledge to next year's work or future challenges. 

That "learning that doesn't transfer" may be central to what's going on when middle school teachers say kids come from elementary school lacking key skills, and high school teachers say that about middle school, and college teachers and employers say it about high school graduates. The folks at the lower level know they worked on that exact skill, but don't know why kids can't put it to use as they move on. The issue may be quality of learning, with students needing to move well past memorization into understanding why the knowledge matters, using it in multiple contexts, and joining in evaluating their understanding as the work goes on. Adding to the quantity of work a student turns in may not change the long-term results much at all.

One final connection: Kentucky has committed to standards that are "fewer, higher, and deeper." Learning that can transfer may take more intensive study, and that's part of why it matters to have a shorter list of expectations with deeper demands about putting understanding to active use.  This chapter adds to my sense that we're on the right track in that approach, and we'd be moving in the wrong direction if we added lots of detailed demands to our standards documents.