Tag: AzMERIT

Categories
Association Blog In the News

Arizona Parents Want School Choice and Support Charter Schools. Here’s Proof

By Jake Logan and Nina Rees

Arizona classes are back in session.

For parents of the more than 1.1 million Arizona students enrolled in a public school, it’s hard to imagine that less than a generation ago, the decision about where your child would attend school was made for you, not by you.

Indeed, many of us are old enough to remember a time when school assignment was dictated not by a student’s needs, but by the neighborhood in which his or her parents could afford to live.

In Arizona, that was the reality until 1994 when Arizona enacted charter school legislation. Together with the adoption of district open enrollment, the statutory changes affirmed a simple principle: Parents know best when it comes to picking the right school for their child.

Continue reading “Arizona Parents Want School Choice and Support Charter Schools. Here’s Proof”
Categories
Education Blog

2017 AzMERIT INTERACTIVE DASHBOARD – HOW DID YOUR SCHOOL DO?

If you are one of the many Arizonans wondering what to make of last month’s release of the 2017 AzMERIT results, then the Association’s school-by-school dashboard will help to provide meaningful context. The dashboard ranks each school in Arizona based on overall pass rates in Mathematics and English Language Arts, and results can be filtered by city and free and reduced lunch rate. Click the image to see the full interactive dashboard.

HOW TO USE THE DASHBOARD
  • Each icon on the scatterplot represents a school. Hover over an icon to see more information about the school, such as its district/charter holder, its overall pass rate on the ELA or Math exam, or its state rank based on those pass rates.
  • Use the sliding filter for FRL to examine schools by their reported Free and Reduced Lunch rates. Use the clickable “Sector” and “Alternative” filters to toggle charter schools/district schools, or alternative/traditional schools. The scatterplot icons as well as the lists at the bottom of the dashboard will update to include only the schools that are within the parameters of the filters.
  • Use the “City” filter to see only the schools that are within a certain city. The scatterplot icons as well as the lists at the bottom of the dashboard will update to include only the schools that are within the given city.
METHODOLOGY
  • This dashboard assigns each school in the 2017 AzMERIT file to a city, based on that school’s physical address.
  • Ties in pass rate resulted in the same ranking being assigned to each tied school, and subsequent schools being ranked based on how many schools were tied in the group ahead of them. Example: If the top five schools in the state were tied for first with a 90% pass rate, those five schools would all be assigned a ranking of “1st”, while the school or schools with the next-best pass rate in the state would be assigned a ranking of “6th”.
  • The pass rates of schools with fewer than 11 students tested were redacted from the state file. Such schools were not included in the rankings, and are not shown in the scatterplot, as they do not have reported pass rates for ELA or Math.
  • Some schools have pass rates listed as as “>=98%” or “<2%” in the state-released file. Such schools were coded with pass rates of “99%” and “1%”, respectively, and included in the rankings. This small adjustment in the reported pass rates is the reason why the x- and y-axes are listed as “Math Pass Rate Proxy” and “ELA Pass Rate Proxy”.
  • Free and Reduced Lunch rates are sometimes reported to ADE as ranges, such as “40-49%”. Such reported FRL rates were recoded with an “FRL Proxy” that is a number, such as 45%. Additionally, several schools in the state do not report FRL rates. These schools were assigned an FRL Proxy of 0%.
Categories
Education Blog

THE CONUNDRUM OF MEASURING SCHOOL QUALITY: RAISE ACHIEVEMENT OR CLOSE THE GAP?

Results of the 2015 AzMERIT testing data show that students in charter schools consistently outperform students in district schools, across all racial and ethnic groups. But does the higher student achievement lead to a closing of the achievement gap between different groups of students? The Association’s analysis shows that the differences in achievement of student subgroups within charter schools actually serves to widen the already large achievement gap.

CHARTER AND DISTRICT DEMOGRAPHICS[1]

Figure 1 shows the racial and ethnic composition of all Arizona public schools during the 2014-2015 school year.  Latino students comprise the largest group (44 percent) in the state, followed by White students (40 percent). The next largest group, African American students, is only slightly above five percent of the state, followed by Native American enrollment at slightly below five percent, with all other groups reporting fewer than five percent each.

FIGURE 1: DISTRIBUTION BY ETHNICITY IN ARIZONA PUBLIC SCHOOLS FY2015

Figure 2 illustrates notable differences in the composition of district and charter school students. Charter schools serve a larger share of White students, 47 percent compared to 39 percent in district schools. Charters also serve a larger share of Asian and African American students than district schools. Conversely, charters serve a smaller share of Latino students, 36 percent compared to 46 percent in district schools, and a significantly smaller share of Native American students.

FIGURE 2: DISTRIBUTION BY ETHNICITY IN DISTRICT AND CHARTER SCHOOLS

These data suggest that White, Asian, and to a smaller extent, African American families are more likely to choose charter schools than other students and families.  When evaluating these students’ overall academic performance, this decision appears to pay off.

AzMERIT RESULTS – THE CHARTER ADVANTAGE

The figure below (Figure 3) shows the percent of students passing the AzMERIT exam in 2015, broken out by school type (district or charter) and student subgroup. These data combine the passing rates across grade levels in English Language Arts (ELA) and Math.  In both content areas, charter students in all subgroups have higher passing rates than their district counterparts.  These data demonstrate that charter schools are providing increased academic outcomes for the students enrolled – at least for the 2014-15 school year. For Arizona charter schools this is good news, given that a charter, by definition, is a contract to improve student achievement.

FIGURE 3: DISTRICT VS. CHARTER PASS RATES BY ETHNICITY ON 2015 AzMERIT

The charter advantage—the difference between charter and district performance—is particularly large for Asian students, a relatively small group of students in both sectors.  Charter schools enroll a significantly higher percentage of Asian students than district schools and provide a 20 point pass rate advantage over district schools in both ELA and Math. The charter advantage for White students is also significant, nine points in ELA and five points in Math.  We see a similar charter advantage for students from two or more races and Native Hawaiian/Pacific Island students.  These four groups of students account for over 55 percent of charter enrollment; they are all scoring above the state average in ELA and Math and their results help explain charter schools’ overall performance.

Latino, African American, and Native American students also see a charter advantage, but a somewhat smaller one. Despite the gains experienced in charter schools, none of these groups perform at or above the state averages in ELA or Math.

AzMERIT RESULTS – THE ACHIEVEMENT GAP

The initial results indicate there is an achievement gap between districts and charters – where overall charter students are outperforming their district peers.  The Association also wanted to investigate achievement gaps between subgroups of students within charters and districts.  The table below shows achievement gaps within charters and districts for each of the major subgroups of students, in both ELA and Math.

For example, the ELA Asian/White achievement gap (the difference between the passing rates for these two subgroups in ELA) in charter schools is 20 percentage points compared to 9 percentage points in district schools.  This means that the passing rate for Asian students in charter schools is 20 percentage points higher than their White charter peers, whereas Asian students in district schools outperform their White district peers by only nine percentage points.

FIGURE 4: ACHIEVEMENT GAP BY RACE/ETHNICITY AND SCHOOL TYPE

The most striking finding in this analysis is the relative size of the achievement gaps between subgroups, regardless of school type. With only two exceptions, all gaps are at least 20 percentage points.  Of all groups, Native American students have the lowest performance in both ELA and Math, and have the widest achievement gaps compared to their peers. The Asian/Native American gap shows the size of the extremes.

White and Latino students represent the two largest groups in the state and their achievement gap is illustrated in the figure below. Since the charter advantage for White students is greater than it is for Latino students (see Figure 3), the size of the achievement gap is greater in charter schools than in district schools.

FIGURE 5: WHITE/LATINO ACHIEVEMENT GAP BY SCHOOL TYPE

 

IMPLICATIONS

Policy makers are beginning to contemplate school accountability measures to evaluate the quality of Arizona’s public schools.  The new A-F formula will include a combination of proficiency and growth measures.  These data suggest that charter schools are likely to fare well in the overall evaluation when proficiency rates are compared to their district counterparts.  The new A-F system must also integrate federal requirements to evaluate subgroup performance on AzMERIT and other indicators of school quality. The evaluation of achievement gaps in subgroups is a critical component in the identification of schools for targeted and comprehensive support (federal intervention).  These data suggest that charter schools are likely to fare worse in the evaluation of achievement gaps, despite their overall higher performance.

In addition to accountability implications, these data raise questions about overall academic performance of student subgroups and ways that schools can close achievement gaps that exist among them.  Clearly, the answer is not to reduce the rate at which Asian and White students are performing in schools.  Rather, there needs to be a focus on replicating that same, or greater, performance for other student groups across all public schools.

FOOTNOTES/REFERENCES

[1] The Association used 2015 AzMERIT data for all public schools to evaluate their performance.  The AzMERIT data provided were unredacted, through a research agreement with the Arizona Department of Education; these data included details for small schools and student groups that are not available in the publicly released data file.  The demographic data included in these analyses were obtained from the October 1 Enrollment file for 2014-2015 (unredacted).

Categories
Education Blog

A FIRST LOOK AT AzMERIT: MORE TEST RESULTS THAN STUDENTS

Arizona’s new state test, AzMERIT, shifts high school testing from a tenth grade exam to end-of-course assessments in grades nine through eleven.  This shift presents new challenges and opportunities for policy makers to consider when building a state accountability system. These dynamics will likely play out in a variety of ways when calculating percent passing and percentage tested. It will be important for policy makers to understand the potential impacts prior to the development of any accountability system.

What’s the Big Deal?

With the passage of SB1430 the State Board of Education will begin to develop its A-F methodology.  The cornerstone to any state accountability system is student level test scores that are used to calculate passing rates and growth scores.  In preparation for this work, the Association began analyzing the 2015 AzMERIT data and discovered significant differences in the number of test records between English Language Arts (ELA) and Math, especially at the high school level.  Given the importance of these data, we set out to investigate the causes and understand the impact. Table 1 shows the number of test records for ELA and Math for each grade level.  At nearly every grade level, the number of Math test records is greater than those for ELA.

The greatest difference is in Grade 9, where most students take an ELA test and, typically, Algebra I. In 2015, the nearly 5,000 more Algebra I test records than Grade 9 ELA test records suggest that students other than high school freshmen are taking Algebra I.
In order to understand the effect of these results, the Association evaluated them in context to the overall enrollment by grade. Table 2 presents grade-level enrollment and a calculation of the percent tested within each content area, as a point of reference.  These data are also reported out by district and charter schools in order to identify any potential differences by type of school.Table 2: 

The use of October enrollment, the only publicly available enrollment file, may impact the “percentage tested” calculations; it is unclear how enrollment changes closer to the testing window could affect these results.  These data show that for almost all elementary grades (grades 3-8), test records account for 99 to 100 percent of grade level enrollment.  Until Grade 8, no compelling difference exists between district and charter schools. However, in Grade 8, math test records account for only 96 percent of grade level enrollment for charter school students, while district schools test 99 percent.

The significant pattern changes that can be seen in the high school “percent tested” columns suggest that the guidance from the Assessment and Accountability divisions within the Department were either unclear or, at best, left room for interpretation.  It would appear that district and charter schools implemented the guidance from the Department regarding end-of-course exams very differently; with large differences in percent testing between the types of schools.  These data also suggest that districts and charters had differing interpretations of the guidance regarding which students must test for each of the end-of-course assessments.

Several examples of this can be seen throughout high school testing.  For example, Grade 9 ELA test records account for 104 percent of Grade 9 enrollment in charter schools and 88 percent of Grade 9 enrollment in district schools. Algebra I test records in math account for 130 percent of Grade 9 enrollment in charter schools and 92 percent of Grade 9 enrollment in district schools.  This suggests that district schools have significantly fewer freshmen taking the Grade 9 ELA test than taking the Algebra I test, and that charter schools are likely testing grades other than freshmen in these courses, particularly Algebra I.  This could be the result of the significant number of alternative charter high schools that serve over-age, credit deficient student populations. The October enrollment file shows 118 alternative charter schools serving Grade 9 students.

In order to investigate this further, the Association evaluated the percentage of ninth grade students tested within charter schools by comparing alternative charter high schools to all other charter high schools.  The results shown in Table 3 suggest alternative charter school students are taking the Algebra I exam beyond Grade 9, and are likely taking multiple end-of-course exams in a given year. This finding is consistent with the student population enrolled in alternative schools and likely explains a significant portion of the overall difference between charter and district schools.

Table 3:

WHY IS THIS IMPORTANT?

These results suggest that guidance from multiple divisions within the Arizona Department of Education resulted in districts and charters making choices about which students should be tested and with which test they should be given.  Without clear policies regarding test taking and the use of test scores for school accountability, the technical validity and overall fairness of any accountability system may be jeopardized.

As state policy makers begin to consider the components and methodology that will be used to calculate A-F and federal accountability formulas, it will be important to understand the impact of end-of-course assessments on availability of data and its use. Unlike the previous state assessment, AzMERIT and the use of end-of-course assessments significantly increases the complexity of available testing data.  These new data need to be fully understood prior to their use for high stakes decisions.

Policy makers should focus on the following areas for consideration: understanding the requirements for end -of-course exam administration and how they impact testing; determine how multiple scores associated with individual students will be used in accountability decisions i.e., calculation of percentage passing; determine how a greater than 100% tested rate will impact accountability decisions.  With each of these areas, clear and timely guidance to the field will be necessary to ensure that schools are not adversely impacted by their testing decisions.  Without this, the foundation of our state’s accountability system (the test score) will be under scrutiny and the reliability and validity of the overall rating will be suspect.