Windsor Watch

Contributed by: Ronald Ch. Eleveld
Edited by: Bruce J. Twambly

I am writing about what is going on in the markets to shed some light and to foment some thought on the part of the readers.  You may not agree with all aspects, however I believe that the basic facts are correct.  I will know I am correct, if while reading this you are upset because of I am hitting chords that resonate.

On an individual basis we all know that having too much debt, can lead to less flexibility, and increased personal anguish.  If we do not pay our own bills we get calls and the ability to do things from replacing the tires on our cars, to fixing the bathroom, to buying a new or used car or even a house becomes very difficult.  The perils of too much debt should be understood by most if not all.

Our country now is a major debtor to the world!  The “bailout of Main Street, and Wall Street” is adding to the total.  This debt is growing and it means that more of the budget has to go to servicing or paying only the interest on the debt.  We are not even talking about the principal or paying it off!  This tremendously limits our ability to provide for our citizens, now and in the future.  Even more importantly it places us at the risk of being told what to do by those that lent us the money.  This has occurred in history, and history will repeat itself.

This problem was not caused by “Wall Street,” it was caused by “Main Street”!  Just like the last housing bubble that in the end took down 100’s of banks in the late 1980’s and 1990’s, and saw real estate prices drop by 50% or more, I will call this the “first bust”.  Think about it, where else can you buy something with almost no money down AND pay it off over 30 years.  The purchaser has little of their money in the transaction and thus limited risk.  Speculators or “investors” also have no risk, and if the value moves up they have a big pay day.  However, as we are seeing, as the values decline many just leave.  Interestingly we call this by the technical term the “greater fool theory.”  It says that I will pay a foolish price for something believing there will be a greater fool that will pay a higher price.  The risk is that YOU are the greater fool.

Today we see a rehash of the “first bust” again involving real estate.  We can call this the “second bust.”  This time the difference is that instead of the banks holding the loans they sold them to “investors.”  This means the pension funds, retirement plans, 401(k)’s, mutual funds, and other capital sources now carry the burden and the risk of capital loss.

If we want real change we need to require those that want to buy a house MUST put up 20%. In the stock market of the 1920’s you could control a $100 of stock for a few dollars. You carried little risk, pennies on the dollar you could lose.  Today you must put up $0.50 for every $1.00 and if values decline to a certain point you have to add money or sell the position.  We all know what happened in 1929, when the stock market dropped and the whole thing came unraveled, causing a major credit crunch that lead us into a depression.  At that time the US was NOT a debtor nation, unlike today.

Today its the no money down real estate loan.  This four wheeled wagon with purchasers on board is watching the wheels come off as the wagon is starting to careen off the road. The real estate agent gets paid a commission on the value of the house, the higher the value the more money they make.  Does the agent want to see prices decline?  The mortgage broker gets paid a commission based upon the value of the loan and the difficulty of getting a loan for a particular buyer, or person refinancing the loan.  Do they want to see prices decline?  The appraiser gets paid for producing the “best”, highest value appraisals.  They have less culpability, because they are making a guess. However truthfully, they are swayed to provide a desired value for the real estate agent and mortgage broker.  The final wheel is the attorneys that closed these loans and got paid for the closing, proclaiming they had no requirement to “know their client”.  They were there to make sure the paperwork was correct.  Does the attorney disclose the 50% or more of the title insurance premium they receive as a commission from every home buyer or person refinancing a loan.  No! They also profited nicely from this cart.

What did Wall Street do?  They provided the cash or grease so that the wagon wheels would roll.  The bottom line is there is plenty of blame to go around. That grease came from various sources as mentioned earlier.

So we have lots of people to blame, and every American will pay for the greed in one way or another.  The person trying to sell will be limited because of foreclosures in there neighborhood.  The properties not being liquidated a rapidly or as fortuitously as possible. Those left with lower values in their respective asset values.  All of us will have to pay higher taxes to pay for this bailout.  Those benefiting are the people living in properties for “free” because they are not making their payments or contacting their mortgage holders about helping them solve their problems, thus necessitating the foreclosure of the house; a clear minority of Americans. We hear a lot about change.  We do need change, but the change we NEED is change none of
us want!  All parties must take responsibility.  The attorneys must take some responsibility for advising the client about the risks they are entering into. They are college educated professionals that should understand what the client is signing and with a rudimentary examination of the client’s knowledge, should know if the client is being taken advantage of.  The “I am here to make sure the paperwork is correct” is an excuse no one should accept.  The Real Estate agents have culpability for not knowing their clients and they should be required to know their clients better.  The mortgage brokers and bankers need to also know their clients better as well and to disclose the risks of the various loans they offer.  Remember all parties will blame the other, but they are ALL a party to the profits, so the are a party to the additional work.

Remember we hear a lot about change.  Real change is not what we are hearing from any of our elected officials or prospective officials.  Real change involves pain and sacrifice. Getting a job is change and sometimes the sacrifice of a comfortable job for one that is unknown can be beneficial.  Leaving the home as young adult is change: sacrificing dependence for independence.  Getting married is change: sacrificing independence for companionship.  Having a child is also a very major change with sacrifice for a wonderful benefit.

Today we need to sacrifice for our benefit and our children and grandchildren’s benefit.  If we do not, then “bust” number three may well be the end of this great nation as we know it.

Contributed by: Bruce J. Twambly
Edited by:

Recently a colleague of mine pointed out an omission in my recent analysis of students achieving goal or higher in Windsor which also led to the recognition of a comparison error between the eighth and tenth grade.  The post “2008 CMT and CAPT: Performance Slides in Six of Seven Grades” contains two analyses: a state ranking of districts by averaging cumulative scaled scores and an annual change in the percentage of students achieving goal or higher on the CMT and CAPT.

The omission in the second analysis was that I did not compensate for annual changes at the state level and then use this to offset Windsor’s changes.  I missed this largely because the Superintendent’s 2007 report did not normalize the data either.  However, this adjustment compensates for inherent difference in the difficulty of the tests from year to year.  So for example, if Windsor showed a 6.0% increase in one year, but the state averaged a 4.5% increase, Windsor’s true improvement would be 1.5%.

The error in the analysis was comparing the 8th grade CMT percentages against the 10th grade CAPT percentages.   Without normalizing to the state, Windsor’s performance showed double digit declines.  But this was strictly because the typical levels of those achieving goal between the two tests were very different.  However, once the results were adjusted for the state averages, the results were far more reasonable.

OK – the good news and the bad news.  The good news is that the difference between 8th grade and 10th grade performance was not nearly as bad as it appeared earlier.  The bad news is that the large majority of “class of” CMT comparisons shifted in a negative direction.  The year to year change in percentage of students achieving goal or higher “tightened” in that values moved slightly toward zero on both the positive and negative side.  However, all seven tested grades now show net losses in the year to year comparison and the “class of” tracking seems to have shifted more uniformly negative.  The following tables summarize the adjusted data.

This table contains a history of State averages of the percentage of students achieving goal or hisgher.  The disciplines are mathematics (M), reading (R), writing (W), science (S) and the average (A).

The following table differences the percentages between consecutive years, compensated for by changes in the state averages. The Administration failed to include this adjustment in 2007.

And as with the state rankings, if we focus on a “class of” as it moves from year to year. Again, the data is adjusted for state changes.

Contributed by: Bruce J. Twambly
Edited by:

This post will present my analysis of the 2008 Connecticut Mastery Test (CMT) for grades 3 through 8 and the Connecticut Academic Performance Test (CAPT) for tenth graders. Two different analyses will be offered: a state ranking of averaged performance and the change in the percentage of students achieving “goal” or higher at each grade level and academic discipline.

Why two different analyses? The presentation of data is a tricky business and the means of reducing the data to bite-sized chunks can either highlight or obscure important information. For purposes of public disclosure, the Windsor Public Schools Administration has chosen to present a breakdown of the percentage of students reaching the top two levels (goal and advanced) of a five tiered system. The five levels are: Advanced, Goal, Proficient, Basic and Below Basic.  Although the Federal No Child Left Behind (NCLB) system has accepted the “proficient” level as a guideline, Connecticut has chosen the higher “goal” level as its measure. Both the CMT and CAPT contain questions in mathematics, reading and writing. The CAPT also includes science and this year, science was added to the CMT for fifth and eighth grades.

The first analysis measures overall performance by averaging the scaled scores in each discipline at each grade level and then ranks the performance relative to the 120+ other districts in Connecticut. This helps to average out variations and captures the performance of all students.  Although the Administration’s presentation of data corresponds directly to the goal threshold as set by the state, the binning of scores obscures the movement within a performance level focusing attention to the students at the boundaries and ignores trends within the groups or overall.

Because social and economic factors greatly influence absolute performance, direct comparison between districts is difficult. However, a district’s performance relative to its economic peers is more clearly revealed using the state ranking. One might say that it is a good measure of how high or low the “tide” is for all students in a district.

Analysis 1: State Ranking

This section will present the combined, state ranking of Windsor student performance for the 2008 CMT and CAPT. First, the trend charts:

The following table presents Windsor’s rank in Connecticut as a percentile for the past five school years. The shading indicates a change in a positive (green) or negative (pink) direction.

It is clear that Windsor student’s averaged performance relative to the state has suffered setbacks almost across the board with dramatic drops in the third, fifth and sixth grades. Of course, each year a different group of students will be tested at each grade level. A so-called “longitudinal” analysis that follows a single “class of” students throughout the years can help, but even that perspective is somewhat erratic. The following table shows the state ranking for each “class of” students over the past five years.

To illustrate how difficult it is to make comparisons: in 2008 only the eighth grade showed improvement as compared with the same grade the previous year. Yet for the “Class of 2012″ this was their worst performance to date (25.5 in 2008, but 32.4, 29.1 and 28.3 in previous years).  The longitudinal view also indicates a drop in performance in five of six tested grades where a comparison could be made.

Although instructors and their methods are intertwined with student’s abilities and performance, you could view each of these tables as “stressing” one over the other. The first table stresses the staff and instructional methods at each grade level, but ranking is still subject to the performance level of the students themselves. The second table stresses the student’s combined abilities but may be hampered or helped at any one grade depending upon the curriculum and quality of instruction.

Analysis 2: Percentage change in students achieving goal or higher

Last year, the Administration presented an analysis of the CMT results by differencing the percentage of students achieving goal or higher at each grade level and in each of the testing disciplines. The Administration made the case that [possible] gains made in the 2005/06 school year were maintained. My analysis showed a split decision using the same methods.  Our disagreement may be semantics.  I was looking for sustained improvement which requires continual, positive movement whereas the Superintendent spoke of sustained gains which would only require that scores do not go down.  That is, they could remain flat, which was the result of my 2007 analysis.

As I have not published a full history of this approach, I will do so now for reference. The following table presents the percentage of students achieving goal or higher in each discipline and grade level over the past eight school years.  The disciplines are mathematics (M), reading (R), writing (W), science (S) and the average (A).

The following table differences the percentages between consecutive years.  This is the method presented by the Administration in 2007.

As with the averaged, state rankings, the percentage of students reaching the goal level (average) slipped in six of seven grades this year.  And as with the state rankings, if we focus on a “class of” as it moves from year to year the results appear more promising:

However, although the classes of 2013 to 2015 appear to be making small, continual gains in the number of students transitioning from the proficient level to the goal level in some disciplines, the state rankings for the same classes dropped. In many cases the drop was significant.  This suggests that overall performance of the class is lower, despite the small gains in this one threshold.

It is important to remember that this entire section examines only a single threshold: the percentage of students in the top two tiers versus the bottom three.   Therefore, the results can be highly volatile and and subject to the amount of effort and attention that an Administration pays to only those students near this threshold.  For example, if an Administration puts most of its attention toward the basic or below basic levels then this analysis could easily show a negative trend at the goal level.  Or conversely, attention to those students just below the goal threshold could artificially boost the perceived performance of the class.  It is for this reason that I prefer the averaged performance and state ranking because it encompasses the totality of a district’s effort as compared with other Connecticut districts.

Performance Levels

The following three graphs illustrate the relative sizes of the five performance levels for each of the CMT grades broken down by discipline for Windsor, the State, Simsbury and Hartford. Click on a thumbnail for a full image.

The graphs were generated by the state test reporting site. I am presenting these graphs to once again try to broaden the view of Windsor’s performance to encompass all students not just one group.   The entirety of the second analysis presented here focuses on the transition between the green and light blue populations but says nothing about the growth, reduction or relative shifts of the other populations.

Concentrating on the left most image, mathematics, certain patterns emerge.  Using the state as a baseline is instructive as it averages over all students.  For 2008 you can see a slight bowing in the 5th and 6th grades math distributions suggesting that overall the tests were somewhat easier in those grades.   Moving down to Hartford, you can see that the pattern is similarly bowed, but shifted extremely to the right.  For Simsbury, a perennial over-performer, the distribution is shifted dramatically to the left.

Now look at Windsor’s distribution in math.  The pattern is not the same.  The relative sizes of the basic and below basic (yellow) distributions taper off relative to the state as the grades progress.  This indicates that the district is working to reduce these populations which is good.  But then turn your attention to the right.  As the basic levels shrink, so does the advanced level. This is not a shift to the left (as with Simsbury) and an indication that the gains in reducing low performance may come at the expense of advanced students. This is only a single snapshot, but an area for possible investigation.

The graphs for reading and writing show better adherence to the state pattern.  Unfortunately, what is common in all of these charts is that the population of advanced students are decidedly lower than those for the state.

Conclusion

For 2008, the overall performance of Windsor students dropped in almost all grades in both state ranking of averaged performance and the average percentage of students achieving goal or higher.  Declines were seen by comparing the same grade in consecutive years and in the longitudinal case where the performance of a “class of” students are tracked from year to year.

Contributed by: Bruce J. Twambly

For the past several years, I have maintained a summary of Windsor’s budget history dating back to the 1993-94 fiscal year. As we are now a couple of weeks from this year’s referendum on May 13th, it is important for this information to be available. So here it is:

Budget History 1993 to 2009 (proposed) (xls)

Budget History 1993 to 2009 (proposed) (pdf)

The Excel workbook contains five spreadsheets (don’t miss the tabs on the bottom). They include Expenditures, Revenue, School Staffing and Salaries, Charts and for the first time the town’s General Fund – otherwise known as the checking account.

This post will be limited to a few observations and explanations. Stay tuned for a more in-depth analysis.

A few quick points:

• The proposed 8.42% spending increase is the largest percentage increase since at least 1993 – the limit of my available data. (tab 1, cell D36)
• The current spending flurry is fueled by a spike in property tax revenue from business development (tab 2, cell F18)
• Student enrollment continues to decline yet the per capita cost for employee salaries and benefits has nearly doubled in a decade (tab 3, columns I and P)
• In the past five years, the town’s General Fund (cash reserve) has nearly doubled – from approx. $6.5M in 2003 to $12.5M in 2008. (tab 4, column E)

The “General Fund Balance” tab is new this year and deserves some explanation as it can be difficult to follow.

Each year, We, the taxpayers of Windsor, approve a budget containing expenditures that our collective government (Town Manager, Council, Superintendent and Board of Education) claim must be spent to maintain schools and services. Then based upon expected revenue, our mill rate is set to collect those funds. Generally, that rate goes up and more money is taken from taxpayer pockets.

In a perfect world, the town would spend every penny that they claimed they needed and would collect every penny that they predicted so that the General Fund would show no change at the end of the fiscal year. But the world isn’t perfect and that is to be expected. When the books are closed after June 30th a final accounting is made and a true end-of-year deficit or surplus is determined and is reflected in the General Fund balance.

It may interest readers to know that in recent years the town has never run a true deficit. There is always some cash left over – either from unexpected revenue or unspent expenditures. As a matter of fact, since 2003, the General Fund has accumulated approx. $11.6M in surplus. There are a variety of reasons, but routinely revenue estimates are too low and expenditure estimates too high.

Also in recent years, the Council has annually taken approx. $400K to $700K of the surplus and paid it forward into the next year’s budget and has appropriated approx. $1.3M in additional spending totaling $5.6M since 2003. So the General Fund balance has swelled by the difference, $6M, nearly doubling Windsor’s cash reserve.

So Windsor has and has had plenty of cash. Current policy requires that the General Fund balance be maintained between 8% and 11% of budget. This puts the current balance $1.75M above the maximum required by policy.

The General Fund is difficult to track from documents published by the Town Manager’s office. If one were to simply add the surpluses reported in the annual proposed budgets, the total would come up short as compared with the actual growth in the General Fund. That is because in the time between when the budget is drafted in April and when the books close in June, more unexpected revenue may be deposited. Balances must be compared from year to year to determine the total amount of “late” transactions.

Finally, what effect has the use of this cash had on our perception of the budget? It skews spending as compared with tax increases. For example, last year (the election year) there was no tax increase, but the town still spent 4.87% more. The year before, a 2% tax increase but a 4% spending increase. This year (not an election year) the town wants to spend 6% more than the tax increase and the 8.42% spending increase is not even mentioned on the town website (as of this posting).

Of course, with all of the cash that Windsor has been accumulating I wonder why there is an increase at all. And why do we have to spend our short term business property tax windfall, especially on new employees whose salaries and benefits must be carried into future years? To be continued …

Contributed by: Bruce J. Twambly

As a public service and in an effort to make public data easily accessible, I have created a static page called “The Education Corner” where I will post publically available data from the Windsor Board of Education and School Administration that is not typically available electronically.

In particular, at each Board of Education meeting, the Administration provides a public data package to members of the Board, the Press and residents such as myself who request a copy. However, unlike the Town Council that publishes a scanned copy of the support material on the town website, this data is not electronically available.

These data packages can often be rather thick and frequently contain valuable detail that cannot be conveyed through presentations and discussions at Board of Education meetings.

I will try to make these packages available the day of each regular meeting.

Contributed by: Howard Jubrey, Jr.
Edited by: Bruce J. Twambly

[ Howard Jubrey, Jr. is the Republican Minority Leader on the Windsor Board of Education. The following is the content of a speech made by Mr. Jubrey at their regular meeting on January 15th, 2008 with minor edits.]

With Dr. Martin Luther King Day approaching I thought it important that I take this opportunity to pay a tribute to this great man. I would like to start with this quote:

Now, I say to you today my friends, even though we face the difficulties of today and tomorrow, I still have a dream. It is a dream deeply rooted in the American dream. I have a dream that one day this nation will rise up and live out the true meaning of its creed: “ We hold these truths to be self-evident, that all men are created equal.”

I am sure that many of you have heard this quote from one of Dr Martin Luther King’s famous speeches. I believe that Rev. King knew when he wrote this speech that his time on this earth was short. I also believe that when he read this speech he believed that his quest for equality for all men would soon be realized. He did accomplish the end of legal segregation that was being practiced in many areas of America. What I do not believe is that he put an end to segregation in the minds of all Americans both black and white.

Rev. King made many speeches. There were other speeches with messages equally powerful if we really listen. They were speeches for us to end segregation and discrimination not only legally but in our hearts and minds. I would like to share with you quotes from other speeches he has made that I believe have just as great an impact.

The first is:

“I want to be the white man’s brother, not his brother-in-law”

Think of what he meant by that. Was it that he wanted us, Blacks that is, to be referred to as Americans, not African-Americans, Negroes, Blacks, “People of color” or Minority? I wonder if when the Italians, Irish or Jews came to this country and were being discriminated against, if the word “minority” was used to define them. It seems that now that these groups have shed the chains of segregation and discrimination that they are considered Americans even though they are still a minority as individual ethnic groups. When you refer to us as African-American, Negro, Black etc., you set up a division in mind that still denotes us as being different. When you refer to a white person it is not as an Irish-American, French-American, Italian-American or even the Majority. It is usually just an American. When will blacks earn that right to be referred to, in all ways, as Americans? Africa is a continent with more than 40 countries in it. We may never know which country we came from. When will blacks refer to themselves as an American first? We should look to the day that when we say “my people”, “my sister”, or “my brother” we are talking about all our American brothers and sisters, not just our black brothers and sisters. How else are we to include ourselves in the Great American Society?

Another quote is

“In the end, we will remember not the words of our enemies, but the silence of our friends.”

I think this quote is very simple in its meaning. If you are an advocate for equality, then speak up about it. If you truly believe that we are all created equal, then treat all people as an equal in all matters of life. Allow us to fail as well as succeed. Allow us to accept responsibility for our success and failures as do all other citizens of the United States. Allow us to be Americans and not label us as a specific group.

“Our lives begin to end the day we become silent about things that matter.”

was a quote of the Reverend that needs no explanation.

And finally:

“Shallow understanding from people of good will is more frustrating than absolute misunderstanding from people of ill will.”

Of all of his famous quotes this one I believe is the most powerful. This quote completely supports one of the most widely recognized and repeated quotes of all:

“Look to the day when people will not be judged by the color of their skin but the content of their character.”

We are not a minority we are Americans. Isn’t it written: “We hold these truths to be self-evident, that all men are created equal”.

This awakening must begin in our schools. For all young people who we have the charge for educating, let us be sure that we do not steal their ability to succeed or fail because we have shallow understanding; or because we become silent about things that matter; or we only believe that minorities are equal because the law requires it. We need to find a place in our minds and our hearts to complete the dream. We need to drop the labels of division. We need to challenge ourselves and our students to be proud members of this diverse community; of this great country we call America. To complete the Dr. King’s dream, we need to unite in one goal, under one label, that we are Americans first and forever.

Contributed by: Bruce J. Twambly

Recently Windsor High School was awarded a “Silver” medal by the new U.S. News and World Report rating of the best high schools in the country thus placing it in the top 505 of 18000+ schools surveyed in 40 states. In Connecticut, sixteen schools/districts received a silver medal and five others received a bronze medal. So how did Windsor receive this award and how was it placed in a peer group with districts such as Avon, Simsbury and Glastonbury?

To answer this question the reader should understand two things: the methodology behind the selection process and the motivations of the researchers who defined the process. Let’s begin with the latter.

The Motivation

Why a new rating system? In a piece on the U.S. News and World Report website, the author states that up until now the only high school rating system was offered by Newsweek and that system used AP placement and performance on the International Baccalaureate tests as the sole criteria for ranking schools. It is the contention of the authors that by using these strict measures of performance, that some districts can still score high while failing to provide what they consider to be a high-value education for most of the students. In addition, hard-working and innovative schools in urban environments would never be ranked.

So the U.S. News sponsored team devised a three step method in an attempt to capture some of the schools that cannot not compete with other districts in terms of raw scores, but provide other benefits to society. They cite a school in Boston where of the 200 students, 90% are minority and most (>50% presumably) live in poverty yet they had 100% college attendance among graduates for three consecutive years.

The authors rely heavily upon data that suggests that poverty is a critical factor in limiting the performance of students. They also claim that since their method uses more data than the Newsweek system that it paints a clearer picture.

The Methodology

So how does one go about creating a system that can somehow recognize both those districts that perform highly on an objective and absolute level while rewarding those schools that perform on a subjective and relative level?

The answer is that the U.S. News system fundamentally embeds two standards: the absolute standard of raw scores and performance and the relative standard of the performance of underprivileged students compared to the lower expectations of that group. In other words, as children in poverty are expected to score lower on objective tests then a school or district should not be penalized in a ratings systems because of the demographic makeup of the student body.

The vetting process involves three steps. Successfully passing steps 1 and 2 earns a school/district the “Bronze” medal. Passing step three earns a “Silver” and the top 100 schools in the silver group with the highest college readiness index are awarded “Gold” medals. The methodology is outlined here and will be summarized below.

The first step involves determining those schools that perform well on their state’s standardized tests. In a recent presentation to the Town Council, Board of Education member Adam Gutcheon described some of the statistical criteria for this step and he stated that only schools that were one standard deviation higher than the statistical mean qualified. For a Normal distribution, this equates to roughly the 84th percentile and is consistent with many of the silver school’s scores on the 2006 CAPT test. A “Poverty Adjusted Performance” index is calculated and any school above 1.0 passes the first step. The student population is presumably divided into at least two groups based upon their economic status. An unspecified weighting value is applied to economically disadvantaged students test scores and this can lift a school above the threshold. The specific formula for determining this index is not provided on the U.S. News website.

In 2006 Windsor tenth graders only reached the 31st percentile in averaged scaled scores for the CAPT test yet still qualified for the U.S. News award. However, Windsor’s student body does consist of a high percentage of economically disadvantaged students (38.2%) who do much better than the state average for that group. Applying the unknown weight pushed Windsor’s performance index to 1.23, higher than nearby Simsbury’s 1.12. Schools with high disadvantaged populations that performed comparatively well easily eclipsed the traditional performers. The most extreme case is Central High School in Bridgeport with a disadvantaged population of 99.8% and a performance index of 3.16 – nearly twice that of high performing, wealthier schools. Central High School’s CAPT performance was indeed the best of Bridgeport’s high schools in 2006, but its averaged score of 223.6 placed it in only the fifth percentile statewide. Bridgeport as a whole was the lowest scoring district in 2006.

The second step is a test as to whether the minority populations in schools (by ethnicity) perform better than state averages. It appears that this test would eliminate high-performing schools that somehow ignore or depress the performance of minority students. However, in Windsor’s case it was the very fact that disadvantaged students (strongly correlated with ethnicity) do well relative to state averages that allowed it to pass step one. Therefore the second test was passed automatically.

At this point a school qualifies for the “Bronze” medal. In Connecticut, five schools earned this distinction and they are:

It is clear that weighing in good performance by larger disadvantaged populations can compensate for the low, absolute performance on the CAPT scores. Stafford High School is and interesting case and appears very comparable to Windsor. Although the fifty-sixth percentile alone would typically eliminate the school from competition, Stafford’s 26.6% disadvantaged population was able to elevate the school over the step one cutoff. In a similar way, Windsor’s 38.2% disadvantaged population was able to elevate it’s thirty-first percentile ranking. It will be interesting to see if Windsor will make the cut next year as the 2006 CAPT scores used by U.S. News this year may have been artificially high and the 2007 CAPT results were markedly lower (22nd percentile).

So why did Windsor earn a silver medal and Stafford only a bronze?

The last step in the vetting process is the calculation of a “College Readiness Index.” The researchers used the percentage of the student population participating in Advanced Placement (AP) courses as the determining factor. However, they did acknowledge a distinction between taking the course/exam and passing the test. Therefore, the percentage taking the AP was weighted 25% and the percentage passing weighted 75%. The researchers claim that a participation index of 20 or higher is a “critical mass of students gaining access to college-level coursework.”

Windsor has a reasonably good AP participation rate of 36.2%. However, it also has a relatively poor participation rate for those who pass with a ‘3′ or above of 20.8%. Nevertheless, the composite College Readiness Index of 24.6 propelled Windsor out of the bronze group and into the silver.

How does Windsor compare with the other silver medal schools in Connecticut?

Sixteen schools/districts were awarded a silver medal and they are:

Clearly there are significant differences between Windsor and the other fifteen silver medal schools. The other fifteen silver medal schools passed the first step based upon their absolute performance on the 2006 CAPT test. Simsbury’s state percentile of 86.4% puts it just over one standard deviation above the mean. The disparity between Windsor’s CAPT performance and the rest of the silver group is clearly illustrated by the following diagram. The majority of Connecticut’s school districts lie between Windsor and the remaining silver schools, but were eliminated presumably due to a lower percentage of economically disadvantaged students and therefore did not benefit from the additional weighting.

2006 Silver Medal Schools – Averaged Scaled Scores vs. Median Household Income

Windsor also has the worst AP pass rate of the group (with New Fairfield a close second) suggesting that either the student population who participates in AP courses are less prepared for the level of work or the instruction is below par or a combination of both.

Conclusion

The new U.S. News and World Report ranking of High Schools is a system that recognizes two distinctly different levels of performance. It does so by accepting the lower expectation of performance from economically disadvantaged students. In other words, this group is graded on a curve. The system has the effect of elevating some urban or nearby suburban schools who have large populations of disadvantaged students who perform better than state averages, but whose overall performance cannot compete directly with more affluent, suburban or rural towns.

While this effect may be what the researchers had in mind, there are many variables that are not considered or quantified. For example, does enrollment in a free or reduced lunch program (the statistic used to differentiate scores) correlate to the root causes of lower performance among economically disadvantaged students? Perhaps this is covered in secondary research. Nonetheless, the new ranking system is actually a combination of two awards: those schools that perform in the top 15% of the state and have at least 20% successful AP participation and those schools whose absolute performance is significantly lower, but have populations of higher achieving, disadvantaged students.

Windsor is a special case. Nineteen of the 21 Connecticut schools awarded medals clearly fall into one of two groups: urban schools (bronze) or affluent, suburban schools (silver). Windsor and Stafford fall between these categories due to higher, but not extremely high percentages of disadvantaged students. Presumably only Windsor was awarded a silver because of the recent push for more students to participate in AP courses. However, Windsor’s low AP pass rate is in no way enviable or desirable, but does not currently detract from the scoring. Windsor certainly cannot be compared with any of the other 15 silver schools or most of the bronze. It sits as a statistical outlier with enough high-performing disadvantaged students to negate its 31st percentile ranking and enough AP participation to reach “critical mass” despite the fact that nearly one in two tests taken are below passing.

In closing, it has long been argued (by myself for one) that comparing all schools based upon a single absolute standard is unfair as affluent districts with higher percentages of students competing for college acceptance and largely free of urban distractions will always win. However, if U.S. News and World Report wished to acknowledge urban schools that perform better than their peers, it could have done so with a separate award. By embedding an advantage for one group of students it creates the odd condition where a district such as Windsor is grouped with the highest performers in the state yet is not comparable to its peers.

Contributed by: Bruce J. Twambly

Now I can understand in the world of public perception that you must promote your strengths and downplay your limitations. But there are limits and with respect to the Windsor School System Administration’s portrayal of the 2007 Connecticut Mastery Test (CMT) scores, those limits are being stretched.

The Spin

Last year’s (2006) CMT results were lauded among educators and others as an indication that the recent downward trend in test scores had been reversed. Scores jumped in all three, tracked grades: fourth, sixth and eighth. And when the 2006 Connecticut Academic Performance Test (CAPT) scores for tenth graders also showed dramatic increases the press coverage was uniformly favorable.

When the 2007 CMT results were released, Superintendent Feser spread the word that the previous year’s improvements had been sustained.

Overall, we are pleased with the recent test scores, and that is because, if you recall, a year ago, our scores went up significantly, and a question arose if those results were just an anomaly or if those could be sustained year over year.”

- Superintendent Feser from an article in the Reminder News, Aug 14, 2007

Perhaps the Superintendent is alluding to an earlier piece on Windsor Watch documenting the CMT increase and questioning whether the dramatic, positive shift was indicative of systemic change and could be sustained over many years.

The Contradiction

On September 18th the Board of Education met for its first regular session of the school year. On the agenda was a review of the 2007 CMT scores. Assistant Superintendent Robin Sorensen presented a lengthy review of the results starting with a slide entitled “Guidleines for Proper CMT Data Analysis.”

Readers of this site are aware that I independently analyze the CMT and CAPT test scores and rank Windsor relative to other districts in Connecticut. So when Ms. Sorensen presented this first slide, my interest was piqued. “Was my analysis improper?”, I wondered.

From the fall of 2004 to the spring of 2006, the CMT underwent a generational change as the annual test was restructured. For the CMT, this is now the fourth generation. Throughout the presentation, Ms. Sorensen cautioned us that it is “inappropriate” to directly compare test scores across generations. For example fourth grade Math in the fall of 2004 (3rd generation) and fourth grade Math in the spring of 2006 (4th generation).

Fair enough, but wasn’t this warning coming a year too late? And more importantly, doesn’t this bring into question the “gains” that have been so widely publicized?

After Ms. Sorensen presented a slide highlighting that in grades 4, 6 and 8 gains were seen in seven of nine disciplines, Superintendent Feser interrupted the presentation and proceeded to state the following.

I want to interject something here. I think that the first bullet needs to be underscored because a year ago when we went up in grades 4, 6 and 8 across the board a question that was raised continuously to us by some people in the press, some people in the community and some people in the schools was “Was it an anomaly?” Was it just, you know, we went up, but it’s not going to sustain itself. And what this demonstrates loudly and clearly is that the work that’s going on in the system is good work and we are sustaining the gains. Yes, in two of the tests we went down. Quite frankly they were not huge declines. We know we have to address those. The fact that we went up, even though it was incremental, shows that we are really on the right path and I just want to underscore that.”

- Superintendent Feser, BOE Meeting, Sept 18, 2007.

But the 2006 jump in CMT scores was anomalous by the Administration’s own definition and as the data shows, Windsor’s overall performance in 2007 did not go up.

The No-Spin Picture

In today’s day and age, it is almost cliche to describe ‘bias’ as not only what is said, but what is not said. The full picture of the 2007 CMT scores lies in the data not presented.

• In grades 3, 5 and 8 there were declines in two of three disciplines. (One of the Administration’s slides was in error as there was a 0.7% decline in fifth grade writing and not an increase as indicated.)
• Although there were increases in seven of nine disciplines in grades 4, 6 and 8, the two declines were in the eighth grade.
• The declines in the eighth grade were characterized by the Superintendent as “not huge declines” (they were -1.5% in math and -4.3% in reading of those reaching goal or higher). By comparison the gains in the fourth grade were insignificant (+0.4% in math, +0.2% in reading and +2.7% in writing).
• Only the sixth grade posted significant progress (+4.1%, +1.4% and +5.4%).

Conclusions

At the very least the claims of across the board gains in the 2006 CMT are in question since cross-generational comparisons have been deemed “inappropriate” by the Administration itself.

Taken as a whole, the 2007 CMT showed neither gains nor declines from that of 2006 based upon the data publicized in the press and presented at the Board of Education meeting. If we concentrate solely on comparing the 2006 and 2007 CMT scores (the “appropriate” analysis) we find that there were approximately as many declines as there were gains, mostly in grades not generally reported in CMT coverage. Grades 3, 5 and 8 dropped in two of three disciplines offsetting the gains in grades 4, 6 and 7. In my statewide rankings, three grades improved and three declined.

In my opinion, it would have been fine for the Administration to simply say that the 2006 jump in scores were obscured by the generational change in the test so we would have to wait and see. It would have been fine to say that the 2007 scores show approximately no change on the whole, but the sixth grade shows marked improvement.

Instead the public was presented with only half of the data and led to believe that overall performance was improving. Perhaps this is just putting your best foot forward, but I believe that all government organizations have a responsibility to present fair and balanced information when communicating with the public. Administrations should present the facts and leave it to the elected officials to explain them to the parents and taxpayers.

Contributed by: Bruce J. Twambly
Edited by:

After last year’s across-the-board jump (link) in CMT and CAPT scores, I was curious to see if the 2007 scores would add credence to the claim that there has been a reversal (link) of past downward trends due to a refocusing of resources. With only a few data points spread over a handful of years, it is difficult to definitively link cause and effect. However, common sense would suggest that a trend reversal involving the cumulative abilities of hundreds of students spread over four different grade levels would manifest itself as a gradual upturn in performance rather than the across-the-board leap that we witnessed last year. Nevertheless, a trend reversal could not be ruled out.

However, the 2007 rankings do not yet support the proposition that a systemic, upward trend has been established. Six grade levels now participate in the CMTs. Grades 4, 6 and 8 are the levels typically published and discussed.

In a recent article in the Reminder (arguably headlined “Schools continue to show gains in Connecticut Mastery Test scores”), Superintendent Feser described the performance of each class. Of the 18 individual opportunities for students to reach “goal” (six grades times three disciplines), 11 showed gains. In grades 4 and 6 gains were seen in all three areas. There were also gains in math in grades 5 and 7, reading in grade 8, and writing in grades 3 and 7. However, the “negative space” of this data indicates grades 3, 5 and 8 showed a decline in two of three disciplines. My aggregate state ranking also showed three grades improving (5,6 and 7) and three level or declining (3, 4 and 8). So I believe that it is fair to characterize the 2007 CMT as a stalemate. Considering the relatively dismal performance of classes earlier in the decade, the sixth grade progress is the bright spot in the data.

The 2007 CAPT result for tenth graders is not encouraging, but not unexpected. Last year’s leap from the ninth percentile to a recent high of 31 was due in part to policy changes relating to which students were allowed to take the test (link) . In 2007, Windsor’s percentile registered 22.8 placing on a par with the 2002 level. Both average scaled scores and percentage of students achieving goal dropped in all four disciplines.

As readers of this site are aware, I have chosen to measure Windsor’s performance relative to other districts in Connecticut by taking an average of scaled scores in each test area . Districts are then ranked and Windsor’s percentile determined. Windsor has remained in the lowest third of the state over the years that I have been processing CMT and CAPT data. However, this is not a fair measure of district performance as many factors contribute to relative performance, many of which are out of the control of any school district — the percentage of students with the expectation of attending college, the percentage of students from single parent households, etc.

In an effort to establish some method of gauging Windsor’s performance, I correlate the average scaled scores with educational indicators such as spending per student and class size. Surprisingly, these traditional indicators have virtually no correlation with average scaled scores in Connecticut:

The strongest correlation for which I have data is against the Median Household Income from the 2000 census. This makes intuitive sense as with higher income comes a greater expectation of college attendance, for example. As a matter of fact, the distribution is so well formed that it can serve as a measure of performance. The scatter graph below compares 1999 median income with average scaled scores for the 2007 CAPT.

The dashed line is a sixth-order fit to the data (not adjusted for outliers) to illustrate my personal measure of a district’s performance. Districts to the left of the line are under-performers. Districts to the right are over-performers. There are five districts that are significantly under-performing compared to all others. They are shaded in orange and are: Bloomfield, Danbury, Stamford, Norwalk and Windsor (in red). Windsor’s average CAPT scores would need to increase by ten points to reach the lowest performer of school districts in towns with comparable incomes and another ten beyond that to hit the middle (roughly the 60th percentile).

Scatter graphs for lower grades are similar but become wider and more dispersed. For the fourth and eighth grades, Windsor just nudges up against the left edge of the distribution. The gains in the sixth grade have allowed Windsor to pass five comparable districts, but it remains ten points below the middle. If we view tenth grade performance as the “product” of a decade of education, then it is worrisome that the aggregate performance is so far out of family with the rest of the state. However, there may be forces introduced into the high school that are independent of the quality of Windsor’s elementary education. If so, then these forces need to be identified and addressed.

Contributed by: Bruce J. Twambly
Edited by:

2007-2008 Superintendent’s Budget

On January 17th, 2007, I attended the Board of Education meeting at Town Hall for the purpose of hearing Superintendent Feser’s presentation of the 2007-2008 budget. It was high level and spent a significant amount of time lauding the 2006 rebound in standardized test scores. The Superintendent is proposing a 3.89% increase over last year’s $54.8 million expenditure. The proposal contains a $646K reduction in staff citing a decline in enrollment and new expenditure requests for software, maintenance and transportation totaling $354K.

However, there is a bit of a mystery remaining. In the presentation, the Superintendent highlights the fact that by maintaining the status quo and adjusting only for fixed costs, the necessary increase would be 4.91% rather than the proposed 3.89%. Or in terms of actual dollars, roughly $2,691,875 versus the stated $2,134,697: a $557,178 net reduction in growth. But if we factor in the advertised staff reductions and new requests, there is a $265K reduction that remains unspecified. I have called over to the Superintendent’s office inquiring into the availability of a more detailed budget to disclose the further reductions. Considering last year’s reduction of 10 seats to the MLC magnet school, I am concerned that some of these reductions may be in this area and will further reduce school choice for Windsor parents. This is only speculation spurred on by public comment at Wednesday’s meeting and I will reserve judgment until I have the opportunity to review a detailed budget.

As readers of this site are aware, I have been questioning the significant run-up education staffing from 1996 to 2002 (see “The Education Bubble” ) and highlighting the significant and continuing burden that this places on our town budget. Since the Superintendent’s arrival, the increase in staffing has slowed then leveled and I applaud the proposed and modest reduction due to student attrition. Nonetheless, I continue to question the cost-benefit of the 100-150 employees that have been added over the past fifteen years.

Block Scheduling

Item 07-08(b) to the Board of Education agenda was innocuously titled “Follow-up to Block Scheduling.” I had no idea what this topic was about and was floored to discover that this is a proposal championed by Windsor High School Principle Arcarese to dramatically alter how classes are scheduled.

I am largely unfamiliar with the debate over what is called ‘The Block,’ and am just now coming up to speed on the topic. In a nutshell (and I apologize if I over simplify this), block scheduling is the process of reducing the number of periods per day, roughly doubling the number of minutes per period and staggering instruction for one particular class every other day rather than every day.

You may ask as I did, whether there are studies as to the benefits or drawbacks of such a radical redesign of “traditional” instruction. Of course there is a plethora both for and against. However, with respect to the most tangible and measurable benefit, test scores, the faculty could not make a compelling case for those districts in Connecticut that have already adopted this type of scheduling. Rather, intangibles such as improved discipline and a more collegiate atmosphere were cited as well as a measurable reduction in the drop-out rate.

Principle Arcarese cites widespread teacher support for the move at 95 percent. However, Board member John Eichner stated that there is significant parental concern and the feeling that this change is being forced down their throats. Apparently this debate is only a couple of months old.

The most notable comment of the evening came from Principle Arcarese when, in an effort to be forthcoming, stated (and I paraphrase), that as the process moves forward and the system takes shape, more teachers will be needed. I believe this is due in part to new elective courses that can be accommodated with this approach. Considering that the Windsor School System has added 100-150 employees over the last 15 years for the same number of students, I have a feeling that once the taxpayers become aware of this new liability that many will join the parents with concerns.

Finally, and this is important: The Board of Education voted 6 to 2 (Panos and DiGiore in dissent) to “approve” the proposal from the Superintendent and to “direct” the Superintendent to provide further information at a later time. Once again, I paraphrase. It is important to understand that the Board of Education is not imposing this change, but rather tacitly approving the proposal. This may only be a subtle distinction, but I do believe that I hear the band warming up in the background for a little bit of the “Windsor Two-step.”

Although I cannot as of yet comment on the technical merit of the Block, my initial reaction is that a policy that provides few tangible and measurable benefits and offers primarily intangible and unmeasurable results is not a prudent move. I would urge readers to do their own research from the volumes available on the Internet. As there is considerable momentum on the ‘Pro’ side, I did find a fairly comprehensive site containing information, studies and opinions for the ‘Con’ side.

Update (1/24/07)

Today I received a more detailed budget proposal from the Superintendent’s office (thank you Jodie!). With respect to my concern for support for Magnet schools, the budget line item is $8,700 lower than last year, however, this difference could be attributed to a number of variables such as the reduced sixth grade seats from the previous year or attrition that is not back-filled at higher grade levels (both policy decisions from the previous year). I see that the 2008 proposal has a new Appendix C that specifies the number of students attending regional magnet schools. This will be very helpful in tracking Windsor’s continuing participation in school choice.

Hopefully this more detailed budget will resolve my other observation concerning a $265K of reductions that were not obvious from the Superintendent’s presentation. However, there does not appear to be much variance between the items discussed last Wednesday and the executive summary of the detailed budget. It looks as if it will take some more digging to understand how the budget increase would have otherwise been 4.91%. It may simply be lost in the shifting of individual line items.