2006 - NYSED English Language Arts Test Press Release and Data
2006 Grades 3-8 ELA Raw Score to Scale Score Conversion Charts
2006 Grades 3-8 ELA Scale Score to Performance Level Conversion Chart
New York Reading Skills Drop After 5th Grade
It's deja vu for some test questions

2006 - NYSED Math Test Press Release and Data
2006 Grades 3-8 Math Raw Score to Scale Score Conversion Charts
2006 Grades 3-8 Math Score to Performance Level Conversion Chart
NY Math Results
A Step Backward
New York Mills: Test can dominate study

TABLES AND CHARTS

Summary Data
Percentage of NY Schools Having Higher Scores Than S-G
A Note on Making Comparisons Across Years
Statewide ELA-4 Scores 1999 to Present
Statewide Math-4 Scores 1999 to Present
Statewide ELA-8 Scores 1999 to Present
Statewide Math-8 Scores 1999 to Present
S-G Math & English 2006 Scores
S-G Ranking Among 41 Capital District Schools
The Range of Passing Scores on All Exams -- 2006, Capital District

4th & 8th Grade ELA Results
Capital District Performance on the 4th-Grade ELA Exam
2006 4th-Grade ELA School District Rankings
Area ELA-4 Grade Point Average with High, Low & S-G
Capital District Performance on the 8th-Grade ELA Exam
2006 8th-Grade ELA School District Rankings
Area ELA-8 Grade Point Average with High, Low & S-G

4th & 8th Grade Math Results
Capital District Performance on the 4th-Grade Math Exam
2006 4th-Grade Math School District Rankings
Area Math-4 Grade Point Average with High, Low & S-G
Capital District Performance on the 8th-Grade Math Exam
2006 8th-Grade Math School District Rankings
Area Math-8 Grade Point Average with High, Low & S-G

3rd, 5th & 7th Grade ELA & Math Charts
Percentage Passing All ELA Exams in the Capital District
2006 3rd-Grade ELA School District Rankings
2006 5th-Grade ELA School District Rankings
2006 7th-Grade ELA School District Rankings

Percentage Passing All Math Exams in the Capital District
2006 3rd-Grade Math School District Rankings
2006 5th-Grade Math School District Rankings
2006 7th-Grade Math School District Rankings

Two Final Graphics
Area Average Pass Rate on All Exams by Grade
Area GPA for All Exams and All Years

* Note: S-G eighth graders who take 8th grade math in 7th grade--about 10% of the class-- do not take the 8th grade math exam until the 8th grade.

As the table pertains to Scotia-Glenville, its performance on the 4th grade exams is essentially unchanged from last year relative to other schools statewide.  The improvement of its ranking on the Math-4 mean appears to be driven by an average 46% increase in the number of students from other schools scoring in Level 1 and 2.  It does not appear to be the result of an improvement in the performance of our students.  Its rankings on the 8th-grade math exam are within the range established over the past 4 years.

The 8th-grade ELA exam, however, is a different story.  S-G's ranking based on its mean score fell to a new low.  This is likely related to an increase in the percentage of S-G students scoring in the lowest level of the exam.  This year, 7% of students scored in the lowest level, a number unmatched since 2000 and an increase of 5 percentage points from last year.  Yet, its ranking based on the percentage of students scoring in the top level improved despite a one percentage point decline in Scotia's score.  This is likely due to a statewide drop in the percentage of students performing in the top level.

Note on comparisons across years:  The table above, comparing the ranking of S-G's performance relative to other school districts, is the best means for making comparisons to earlier years with data from 2006 and thereafter.  From the ELA press release we read:

From 1999-2005, New York had tests given only in Grade 4 and 8. The standards are divided into Grade bands (Grade 2-4, for example). Therefore, the Grade 4 test in those years had some content that was somewhat simpler, more similar to lower grades. Some of that simpler content now appears on the Grade 3 test. Therefore, the 2006 Grade 4 test is not equated to previous years’ tests in difficulty. Statistical analysis indicates the 2006 Grade 4 test is therefore slightly more difficult than the previous year’s test and the tests given in prior years. However, the 2006 Grade 8 test showed no difference in difficulty from the 2005 test.

And from the math press release we read:

It is not possible to make direct comparisons between the 1999-2005 math results in grades 4 and 8 and the 2006 math results in grades 4 and 8 because New York State’s math standards have been raised.

Several comments are in order.  First, only the 8th-grade ELA exam retains direct comparability to prior years, according to the state.  Yet, there's no reason why the 4th grade ELA exam can't be made directly comparable to last year's exam.  Supposedly, an important  rationale for converting raw scores to scale scores is to make exams of varying difficulty comparable.  See How the Scale Scores Are Calculated for New York State Regents Examinations.

Second, the allegation that the math standards have been "raised" is highly questionable.  It's more accurate to say that they have changed because it's not clear that the new requirements are any more difficult than the old ones.  Learning some aspects of algebra earlier can be easier than learning some of the elements of ratios, proportions, percentages and graphing, which used to be taught.  To some extent, the new standards substitute one kind of knowledge for another, rather than add to the amount of knowledge to be mastered.  Nothing is gained simply by changing the timing of when concepts are introduced.  Moreover, the new standards simply come closer to where they should have been in the first place (assuming no substitution of content).  The texts I used to teach my daughter math during summers, to do the work the school should have been doing during the academic year, began subtly introducing algebra and geometry in third grade.  See this letter to the school board.

Third, even though the scores on many of these tests cannot be "directly" compared to earlier years, they must be compared.  NCLB requires school districts to make Adequate Yearly Progress.  The only way you can do that is to be able to compare this year's data to last year's.  The state does this by setting an Annual Measurable Objective (AMO).  See Using School and District Reports to Determine Accountability Status [PowerPoint].  To my knowledge or S-G's, the state hasn't set the AMOs for 2006, but when it does, an indirect comparison of this year's test results to last year's will be possible.  For more information on this topic, search Google.

Fourth, while retaining an air of objectivity with endless numbers, the entire enterprise, from setting standards, writing tests, setting cutoffs, and converting raw scores to scale scores, to setting the AMOs to determine Adequate Yearly Progress, oozes with subjectivity.  See, e.g., Nine Commentaries on NY K-12 Exams: Testing to the Results.  In my opinion, it's impossible to know whether public schools are making progress or not, which means we are spending millions upon millions of dollars to produce worthless information.  Hopefully, with this year's data, the state has decided to recalibrate the system and we'll have a better basis for drawing conclusions in the future.  But I'm not optimistic.  The public-relations driven tendency is to demonstrate that public schools endlessly improve.  But when other data, like NAEP and international tests, suggest otherwise, or when they suggest that the progress is not nearly as large as advertised (see, e.g., 2005 NAEP and NY Scores Compared), then who should be surprised that the state would recalibrate its scoring and tell you that you can't "directly" compare current and past scores?  Then it can restart inflating the results until the gaps with other data grow unacceptably wide again, at which point recalibration begins anew to facilitate several more years of positive press releases touting improvement.

Fifth, for unexplained reasons, the 4th and 8th grade ELA raw scores were weighted to create the scale scores.  This seems to be the same as curving the raw scores.  I suspect the rationale for doing this is to adjust for an alleged change in test difficulty or a change in the number of questions on the exam..  If that's the case, then both the 4th- and 8th-grade ELA exam data should be comparable to last year's data.

Sixth, State Ed's blanket insistence that current scores cannot be compared to previous scores is specious.  In some regards they can't be compared, but in other ways they can.  What State Ed doesn't want you to conclude is that educators are doing a worse job because the outcomes on 3 out of 4 of the exams fell.  The state prefers to interpret the lower outcomes to changes in the math standards and increased ELA test difficulty.  But test difficulty is supposed to be compensated for in the raw-to-scale score conversion process.  And a change in standards doesn't mean the old and new scores can't be compared.  Both this and last year's results can be interpreted as a measure of how well teachers were able to teach the standards in effect at the time.  By this interpretation, teachers were better able to teach the math standards in effect last year.  They were less able to teach this year's standards, even though they knew in advance what they were.  Consequently, test scores declined.  State Ed's pronouncement of non-comparability is more about a sensitivity to teacher feelings and a desire to avoid dissing them in the press than it is about accurately reporting the results.  The state emphasized the new information from the 3rd, 5th and 7th grade tests and "reinterpreted" the declines on the 4th- and 8th-grade exams.  Keeping in mind that about half the improvement shown in these charts has been the result of changes in scoring, here are the steps backwards the state didn't directly reveal:

More students statewide failing the the 4th-grade ELA exam than in 2005:

More students statewide failing the 4th-grade math exam than in 2005:

Slightly more students statewide passing the 8th grade ELA exam than in 2005 but more students scoring in the lowest level, too:

More students statewide failing the 8th-grade math exam than in 2005:

In addition to showing S-G's performance on state exams, the table, above, permits a comparison of the performance of 8th graders in 2003-2006 with their 1999-2002 4th grade performance.  The attrition of students passing the exams by 8th grade is unacceptably high.  As Commissioner Mills reportedly said here,  the decline in scores from the early elementary grades to 8th grade is "not inevitable" and "many schools defy the trend."  Unfortunately, S-G is not one of those schools.  Our best performing students are being under-challenged and basically ignored because of an entrenched ideology to equalize students rather than provide every student with the opportunities needed for her/his greatest success.  Thankfully, that ideology is changing to a broad, yet shallow, extent.

The following graphic shows the percentage of students passing the ELA and math exams based on the average for each school district in the capital region. To calculate the average, I added the percentage of students passing each exam and divided the total by the number of exams--typically 12, except for Edinburg (8, because it doesn't have 7th or 8th grades) and Cornith (11, because the database provided by NYSED did not include its score on the 5th grade ELA exam). The graph also shows the range of the percentage of students passing all of the exams. Most typically the highest pass rate was on the 4th grade math exam and the lowest pass rate was on the 8th grade ELA exam. In general, the more narrow the range or variation of pass rates, the better or more consistent the quality of education across all grades.  The variation in S-G's passing scores is average for the area.

Niskayuna ranks at the top with an average of 87% of its students passing all the exams. Schenectady and Albany bring up the rear with a pathetic 42% average pass rate. Scotia-Glenville turned in a 73% pass rate, which puts it in the middle and slightly above the region's average 68% pass rate.

4th-GRADE ENGLISH LANGUAGE ARTS RESULTS

In the above chart, I answer the question, "How has the number of students performing in each level changed over time, assuming the same number of students took the exam each year?"  The number of students performing in each of the levels for 2006 are the actual numbers.  For prior years, the number of students performing in each level reflects the number of students who would have been placed in each level had the same number of students taken the test as in 2006.  I calculate these numbers using the percentages provided by the state for each performance level.  

Interpretation of Performance Levels:

Level 4 – exceeds the standards
Level 3 – meets all the standards
Level 2 – meets some of the standards or partially meets the standards
Level 1 – shows serious academic difficulties.

The above chart shows the percentage of students scoring in each level in each of the 42 NY capital district public schools with elementary schools.  The schools are arranged in order of the percentage of students passing the exam (scoring at Level 3 or 4) from most to least.  Scotia-Glenville is in 26th place--a substantial decline from last year--and its score is outlined in red.

The following graphic shows the grade point average for the capital district, which has been calculated by summing the scores for all pupils and finding the average by awarding 0.55 for Level 1, 1.7 for Level 2, 2.85 for Level 3 and 4.0 for Level 4.  The area GPA is then compared to the area low, the area high and Scotia-Glenville's GPA.  The graph clearly shows that S-G has moved away from the best performing schools with scores nearly equaling the average for our area.

8th-GRADE ENGLISH LANGUAGE ARTS RESULTS

The above graphic shows that performance on the 8th-grade ELA exam has declined since 1999, with 4,097 students not meeting state standards in 1999 compared to 4,348 in 2006 (given equal sized cohorts).  Whenever the Commissioner talks about 8th-grade ELA results, he generally makes his comparisons to 2001, the year of worst performance.  What the Commissioner never talks about is how the cut-off and scale scores on this exam have changed since that time to make the exam easier to pass.  See, e.g., NY Makes Huge Scoring Change to 8th Grade ELA Exam.  When factoring in these changes, it appears that performance on this exam has remained steady or declined since 2001 despite the appearance of a small improvement.

The above chart shows the percentage of students scoring in each level in each of the 41 NY capital district public schools with 8th grades. The schools are arranged in order of the percentage of students passing the exam (scoring at Level 3 or 4) from most to least.  Scotia-Glenville is in 25th place--one place below last year--and its score is outlined in red. 

The following graphic shows the grade point average for the capital district on the 8th-grade ELA exam, which is compared to the area low, the area high and Scotia-Glenville's GPA. The graph clearly shows that S-G has average results on the 8th-grade ELA exam.  Any idea that we simply can't be doing better ought to be flatly rejected.

4th-GRADE MATH RESULTS

In the above chart, I answer the question, "How has the number of students performing in each level changed over time, assuming the same number of students took the exam each year?"  The number of students performing in each of the levels is based on the number of students taking the test in 2006.  I calculate these numbers using the percentages provided by the state for each performance level.

Interpretation of Performance Levels:

Level 4 – exceeds the standards
Level 3 – meets all the standards
Level 2 – meets some of the standards or partially meets the standards
Level 1 – shows serious academic difficulties.

The above chart shows the percentage of students scoring in each level in each of the 42 NY capital district public schools with elementary schools.  The schools are arranged in order of the percentage of students passing the exam (scoring at Level 3 or 4) from most to least.  Scotia-Glenville is in 14th place--up two places from last year.  S-G doesn't place nearly as many students in the top level as better performing schools.

The following graphic shows the grade point average for the capital district, which has been calculated by summing the scores for all pupils and finding the average by awarding 0.55 for Level 1, 1.7 for Level 2, 2.85 for Level 3 and 4.0 for Level 4.  The area GPA is then compared to the area low, the area high and Scotia-Glenville's GPA.  The graph shows that S-G is struggling to break away from the average performance trendline.  S-G's performance in 2005 was not as good as it was in 1999 relative to other area schools.

This is the worst performance on a 4th-grade math exam ever.  The state says this is because standards have been raised.  But it's still elementary math and the teachers knew about the changes way in advance.  Apparently, even the small changes the state made overwhelmed the system.  Either it takes years to realign the curriculum or the elementary teachers had difficulty teaching it.  In either case, this is a big problem in a fast-paced world with increasing global competition.  As an aside, I recently called Prudential and Merrill-Lynch about my mother's accounts.  The Prudential account assistant was in Ireland and the Merrill-Lynch assistant was in Puerto Rico!  What's up with that?  Better math skills and lower labor costs in other areas of the world?

Most of the "improvement" shown in 8th-grade math is very likely due to changes made by the state in things like exam content, cut-off scores and the standards that define proficiency.  One luxury middle school teachers don't have relative to elementary teachers is to drop items from the curriculum and to curtail recess so more instructional time can be devoted to math.  The middle school curriculum is packed with mandates.  The second article in this post suggests that elementary schools may have increased math instruction time by 60%.

The above chart shows the percentage of students scoring in each level in each of the 41 NY capital district public schools with 8th grades. The schools are arranged in order of the percentage of students passing the exam (scoring at Level 3 or 4) from most to least.  Scotia-Glenville is in 15th place--3 places below last year.  Its score is outlined in red. 

The following graphic shows the grade point average for the capital district on the 8th-grade math exam, which is compared to the area low, the area high and Scotia-Glenville's GPA. The progress that S-G had made in pulling away from the middle vanished this year.

3rd, 5th & 7th Grade ELA & Math Charts

Percentage Passing All ELA Exams in the Capital District

You'll notice that S-G's pass rate on the 5th grade ELA exam was very near the top and discordant with its other scores.  The district's explanation is that more students close to the cutoffs fell on the high side than normal.  That's plausible.  What really stands out is the percentage of S-G students scoring in the top level, which exceeded the percentage of Niskayuna students scoring in that level.  Typically, S-G has about half the percentage of top scorers as Niskayuna.  In my opinion, something more irregular than random luck affected S-G's 5th grade results.  It might have been the test, itself, or it might have been something else.  Unfortunately and unlike Texas, NY does no statistical analysis to detect irregularities in testing.  If the score is "real," then one would expect the 2007 6th-grade results to be similarly high.  We'll see.

2006 3rd-Grade ELA School District Rankings

2006 5th-Grade ELA School District Rankings

2006 7th-Grade ELA School District Rankings

Percentage Passing All Math Exams in the Capital District

Here we see another aberration for S-G.  The percentage of students passing the 7th-grade math test is way too low.  Unlike the high score on the 5th-grade ELA test, the district has a reasonable explanation for this.  Unfortunately, I'm not at liberty to tell you what it is.  In my opinion, the public deserves a straight-forward explanation, but it will have to come from the district.  I can only say that in my opinion, the score is an aberration that is not likely to be repeated in the future.

Better than any other chart, this one shows that if it weren't for S-G's math department, the district's local overall ranking would be far worse than it is. (That will become more clear in the last two charts).  Even so, there's still too much of a downward slope in passing scores for math.  Our passing rate for the 8th grade should be near 80%, not 70%.  Yet, if our students have the ability to outperform area schools to this degree in math, they have an equal ability to do the same on the ELA test.  Until now, I've made broad complaints against the district's academics.  But based on this data and my experience with the district, I believe the biggest problem can be more narrowly isolated to the English Department.

By the way,  take a look at the scores from the worst performing schools.  I'd like to know why people aren't rioting in the streets in these school districts, namely Albany, Schenectady, Troy, Watervliet, Green Island, Lansingburgh and a few more.  If these schools were charter schools, they'd be shut down permanently.

2006 3rd-Grade Math School District Rankings

2006 5th-Grade Math School District Rankings

Area Average Pass Rate on All Exams by Grade

Notice how S-G starts out well above the area average and ends up with average performance in the 8th grade.  I've repeatedly said that S-G should be doing better than it is and this chart comes the closest to any I've seen of proving it.  Notice, also, that an average of about 40% of S-Gs 7th- and 8th-graders are being left behind.  That's atrocious.  I say it's way past time to take down the National School of Excellence signs from the 1990s.  And it's time for middle school principal Sharyll Keller to eat her words and issue a public apology:

quote:

The switch to a middle school is not an experiment. It's a commitment. We will see immediate improvements within the year.
--Sharyll Keller, S-G Middle School Principal, from this September 2002 Gazette article. 

The middle school was not just an experiment, but it was a predictably failed experiment.  We spent millions to convert the junior high to a middle school and added another $700,000 or more to operations annually and what did we get for it except broken promises?

Area GPA For All Exams

The final graphic shows the grade point average for the capital district on the  4th- and 8th-grade math and ELA exams, which is compared to the area low, the area high and Scotia-Glenville's GPA. The graph shows S-G's GPA has fallen at about the same rate as for other area school districts.  Despite being above average, its GPA ranks 19th out of 42 area school districts.  More than anything, the chart proves the system produces the results it's designed to produce and only system change can dramatically improve academic outcomes.  8 years of state testing and this is all we have "gained." It's not worth a plugged nickel, let alone the billions that have been spent.  Whatever else might be said, this is nowhere near the amount of improvement American students need to be globally competitive.  See here.  We need schools for The 21st Century Student.