What does school rating mean: The murky middle: What an average school rating means

The murky middle: What an average school rating means

At first, Erin Cervelli wasn’t even considering this school. She and her husband were going on all the tours together — but not this one. “I told my husband he didn’t need to come,” she says. “I was only checking it out because it’s [relatively] close to home.”

But that tour changed everything. They’d been looking at all the “good” schools — schools with strong reputations and Spanish immersion programs. This school was neither trendy nor language-based, yet it seemed ideally suited to how their son Jack learns. Cervelli did her research, talking to friends and other parents and attending orientations and pre-school play dates. When it came to filling out the district application and ranking her favorite schools, she put San Francisco Community Alternative School first and easily nabbed a spot for her son. But she admits she was shocked when she later learned that its GreatSchools Rating — 4 out of 10 — was so low. The mediocre rating didn’t jibe with what she saw — or with what the school promises.

What’s the big deal about this rating?

Every year, 44 million people visit GreatSchools. Their goals vary, but many want to learn more about schools available in their area or where they might move. Some also want to gain more insight into their child’s school. With the exception of certain areas, including Indiana and Washington, DC, where we have more substantial data about things like student improvement rates and college readiness, in most of the country the GreatSchools Rating is based on averaging state test scores.

Here at GreatSchools, we’re constantly reminding people that test scores never tell the whole story, but they are one way for a school to measure the performance of their students on core academic subjects. What’s more, they’re often the only available way to objectively compare to what extent schools are delivering on their promise of teaching certain academic standards.

Clarity around 1 or 2 and 9 or 10

While state test scores are far from perfect, the results of very high-rated or low-rated schools do tell you something about the baseline academic performance of most of the students in that school. “When a school’s rating is a 2, you know that students at this school are scoring lower on state standardized tests on average,” explains Sam Brown Olivieri, who heads the GreatSchools data team. “Across all grades and subjects, students are scoring lower than most other students in the state.”

When you look at Flynn Elementary School (GreatSchools Rating 2), for example, their GreatSchools Rating by grade is consistently low: 1 for 2nd grade, 3 for 3rd grade, and 2 for 4th and 5th grades. Where there is variation in test scores, it reflects a wide achievement gap: Non-economically disadvantaged kids far outscore their poorer classmates (7 vs. 1) — the same score gap holds true for white students versus their Hispanic and black classmates. Does this mean you shouldn’t send your child here? Not at all. It just means you should find out what the school’s own standards of success are and how that dovetails with your expectations for your child. Flynn has both a Spanish immersion program and a general ed program so that complicates things, but the bottom line is that the test scores are consistently low.

Similarly, a high GreatSchools Rating signals another kind of consistency. “When a school’s rating is a 9 or a 10, you know that the vast majority of students, across grades and subjects, are scoring proficient or above — and better than most other students in the state – on that state’s tests,” Olivieri says. At John Yehall Chin (GreatSchools Rating 10), for example, all grade-by-grade ratings are a 10. Boys and girls perform at the same rates. And on the page that breaks out ratings by ethnic and socioeconomic subgroups, you see rows of perfect scores. “This doesn’t mean parents don’t need to visit, though,” Olivieri cautions — because you need to like how they’re teaching the students to achieve those test scores. But when you visit, you can focus on questions about safety, teaching styles, discipline practices, bullying, and… well, the list goes on and on.

The muddled middle: GreatSchools Ratings 4 – 7

When scores are really high or really low, the data tells a clearer story. It’s in the muddled middle — school ratings of 4 through 7 — where the data can suggest many different stories. “You really need to look more deeply at how students across the school are performing in different grades and in different subjects,” Olivieri says. “If a school has a steep decline year-over-year, that’s cause for alarm. Persistent principal turnover can cause challenges, too.”

Old success, new leadership

Case in point: Moscone Elementary School (GreatSchools Rating 6) in San Francisco’s Mission District. At first blush, it feels like a really strict urban Catholic school. Though it’s public, kids wear uniforms. At the early morning drop off, students and parents appear subdued. The school’s first principal, Patricia Martel, led with a singular focus: literacy — quickly and in English. She crafted simple, straightforward mission statements that the kids repeated daily and were translated into Spanish and Chinese for parents to learn as well. Kids don’t miss class, are respectful, and aim high. Unapologetically, Martel would tell parent on school tours, “If you don’t like a highly disciplined school, don’t choose this,” she recalls. Test scores reflected her vision’s success, rising to a GreatSchools Rating 9. During those years the school attracted a devoted following of local families and Martel became a model for principal training programs. But when Martel left, the school floundered a bit.

Now, seven years, two principals, one recession, and a much more impoverished community later, the school’s scores have fallen. For a high-poverty (84%) school, this shouldn’t be equated with failure, especially since the gap between some students and their economically-disadvantaged peers is narrow (7 vs. 5). And, while their GreatSchools Rating fell in 2011 (7 to 6) and 2012 (6 to 5), they reversed the trend last year (5 to 6). “This is [Principal] Valerie [Hoshino]’s fourth year,” Martel says, adding, “Change takes four to seven years to really become institutionalized. It took me a while, too. ”

We’re just not that into tests?

When Erin Cervelli saw San Francisco Community Alternative School’s GreatSchools Rating 4, she was surprised but she didn’t regret her decision. Test scores “are not an issue for me,” she says. The parent reviews (4-star average out of 45 reviews) of this progressive mecca with a welcoming yard, lush garden, project-based learning, teacher “looping” (teachers stay with a class for two years so children are well-known), social-emotional curriculum, and a late start (9:15 am vs. 7:50 am at Moscone) tend to match her enthusiasm. A handful of reviews rail against academic shortcomings, such as a notable complaint in 2005 about kids needing to repeat Algebra I in high school. At orientation this summer, says Cervelli, test scores were addressed and dismissed quickly but efficiently: the project-based learning school does not teach to the test, administrators said, because it would require about two and a half months of valuable instruction time.

“When a school dismisses test scores, it’s time to ask more questions,” Olivieri says. “That’s when you need to ask pointed questions about what the school’s learning goals are — and how they measure them. Ask, ‘How do you know if your students are prepared for success?’ and ‘How do you measure progress and ensure that all kids are learning?’“ she suggests. You can also ask if the school teaches to the state standards — and if and when your school will be teaching to the new Common Core State Standards. Don’t just let it go, Olivieri says, firing off more questions to ask: “If you don’t use your state’s standardized test as a true measure, then what do you use? Do you track your students in high school and college to see how your students do? How do you hold yourselves (teacher, principal, and school) accountable for student success?”

Taking Olivieri’s advice, I asked Nora Houseman, San Francisco Community Alternative School’s principal, these questions. “We’re not anti-assessment — not at all,” she says. “We believe in assessments that are linked to what the kids are learning … but we’re very opposed to tests that aren’t.” Their project-based approach emphasizes depth over breadth, critical thinking, and presentation skills. “I think, and I think the district thinks, we’re very much aligned to the new Common Core Standards,” she says — and she’s optimistic that these new assessments will better match what they’re teaching — and that as a result the school’s test scores will go up.

Another yellow flag at San Francisco Community School is an apparent equity gap. The year before last, about a third of kids got into Lowell, one of the country’s leading high schools. But, only about one third of students aren’t from lower socioeconomic backgrounds — and the test score divide between the haves and have nots is striking: 10 vs. 3. So, does this school serve only the privileged kids? That’s not where Houseman sees the line. “There’s a big divide with our K-8 students versus our transfer students,” she says, adding that the district often recommends it to students when things don’t work out at other, more traditional, schools. “We have less time with our transfer students,” she says. “Our K-8 kids do extremely well and often get into the school of their choice.”

It’s all about your child – and getting answers

So what’s a parent to do? Look at the data to see how the school is serving children like yours. Will your son really thrive at a school where the girls are performing significantly better than the boys? Is there an equity gap? Do the test scores jump around erratically? These data points will rarely provide answers, but they can offer a starting point for figuring out the hard questions all parents need to ask.

Editor’s note: For this story, we tried to pick schools with similar levels of free and reduced lunch students — a shorthand for socioeconomic levels. But such markers — like test scores – are limited measures. There’s a clear connection in U.S. public schools between race, wealth, and student outcomes. Beyond that fact, there are other factors to consider, such as how schools serve boys versus girls, as well as kids from different backgrounds, students at different grade levels, and students with different learning needs, all of which may be crucial to consider when finding the right fit for your child.  

California School Ratings – FAQ

Where do you get the school ratings data?

The school testing scores are from California’s
2016 CAASPP Paper-based Test Results.
Percentiles and ranks are calculated by California School Ratings, based on the 2016 CAASPP test scores.

What is the 1-10 rank given for schools on this site?

The California School Ratings Rank is determined by a school’s percentile in comparison to
other schools of the same type in California (1 is the worst, 10 is the best).

Percentiles are calculated by California School Ratings, based on the 2016 CAASPP test scores.

Example: Schools in the 90th percentile and above have rank 10, 80%-89.999% rank 9 and so on.
A similar number of schools occupy each rank.

I see a school on your site with a lower rank and higher
test score compared to another school on the site, how is that possible?

Because only similar grade-range schools are compared when ranking, the test score
required to achieve a given rank is different for High Schools, Middle Schools, and Elementary schools.

Why are your rankings or test scores different from the ones on another school rating site?

In cases where other sites display a different rank, it is possible that they have
older or newer data or that the site uses a different formula for producing ranks.

How does California School Ratings (CSR) compute school percentiles?

How do I advertise on California School Ratings?

You can advertise on school-ratings.com through Google’s Adwords site.

The bigger the school, the higher the rating: how the number of students affects the rating of schools

The parameters of the Moscow rating of schools change every year. This year, rating points are almost proportional to the number of students in the school. That is, you can enter a new rating – each student brings the school 1 point. The order of schools in the ranking will hardly change. Natalya Netrusova, a mathematics teacher at the Intellectual school, is looking into what happened.

Moscow rating criteria

The Moscow rating of schools is cumulative: each student can increase it. Among the rating parameters, there are those that directly depend on the number, and others that do not. For example, the number of first graders who transferred from the preschool department depends on the number. The larger the school, the more first graders it has. But the number of winners of the All-Russian Olympiad depends little on the size of the school. nine0003

Parameters proportional to the number

Eleventh-graders who have passed three USE in total at least 250, bring the school 1.5 points (previously it was 1 point).

Ninth-graders who passed three OGEs in total at least 12, bring the school 0.25 points.

Seventh graders who have attended this school since grade 5 earn 0.4 points (new this year).

Fifth graders who transferred from the primary school of the same educational institution – 0. 3 points (new this year). nine0003

Third graders do not earn points. 🙂

First-graders who transferred from the preschool department – 0.4 points (before it was less).

Ideal school model

Consider an ideal school with n students. Let the parallels be the same: there are n/11 students in each class. No one leaves the school, the children pass the OGE and the Unified State Examination perfectly. The parameters from the previous paragraph will bring to the school (1.5 + 0.25 + 0.4 + 0.3 + 0.4) n/11 = 0.26 n points

That is, in an ideal school, each student will certainly bring to the school 0.26 points. It is not difficult to calculate the rating: you need to multiply the number of students by 0.26 and add points for other parameters (Olympiads, sports achievements, etc.). nine0003

What about in the real world?

In a real school, things are a little worse: not everyone scores 250 points in three USEs, and not all first-graders of the school came to it from the preschool department.

Let’s look at the real dependence of rating points on the number of schools. Let’s plot the rating scores of schools on the graph: the number is on the horizontal axis, the rating scores are on the vertical axis. If the parameters listed above significantly affect the rating, then most schools are approximately on the same straight line on this graph. nine0003

We polled 180 schools about their rating scores and took the number from open sources.

Dependence of rating points on the number of schools

A linear dependence is visible. Each student brings about 0.14 points to most schools, just because he is. Growing up in the ranking is easy – increase the number. There are already 178 schools in Moscow with more than 3,000 students.

The rightmost point is interesting – the school is located just below the straight line, which crowds most of the other schools, that is, not everything in this school is ideal. But her contingent is over 10,000, which immediately gives her more than 1000 rating points. nine0003

Schools that can do anything other than be large are visible on the graph – a group of schools located above the main straight line on the left side of the graph. These schools scored a lot of points due to other parameters (mainly Olympiads).

But the enrollment is not large enough to compete with the big schools. 50 first-graders are 2 winners or 4 prize-winners of the All-Russian Olympiad or 17 prize-winners of the Moscow Olympiad.

The rules for calculating the rating are adjusted every year in accordance with the goals of the development of education. It remains only to conclude what these goals are. nine0003

Photo: Shutterstock (Miriam82)

School rating depends not only on USE results , schools, governors and regions.

Sergey Bochenkov and Igor Valdman in the HSE journal Educational Issues emphasize the importance of creating a balanced national system for monitoring the quality of education and explain how to correctly interpret the results of the Unified State Examination

When evaluating the work of teachers, schools and educational systems, the results of unified exams should be used only in combination with other measurements of the “efficiency” of a secondary school, the expert of the independent agency for assessing the quality of education “Leader” Sergey Bochenkov and the head of the laboratory for control and monitoring in education of the Institute are convinced management of education RAO Igor Valdman . In the article “Interpretation and Presentation of USE Results”, published in the journal “Problems of Education”, No. 3, 2013, researchers oppose the perception of unified exams as a “measure of all things.” nine0003

The absolutization of the Unified State Examination is largely a consequence of the imbalance in the all-Russian system for assessing the quality of education. Meanwhile, overloading the Unified State Exam with unjustified expectations is fraught with “erroneous managerial decisions,” experts warn.

This situation, according to the authors of the work, can be overcome only through the introduction of other evaluation procedures. That is why one of the priorities of the state policy in the field of secondary education is the formation of a balanced system for measuring the quality of education. In addition to unified examinations, it should include monitoring of educational achievements and socialization of students, as well as assessment of learning outcomes at the school level (see the State Program of the Russian Federation “Development of Education” for 2013-2020). nine0003

How to use the USE correctly

USE scores should be “just one feature of the complex portrait of results” of the school system, Bochenkov and Waldman note. They give a general recipe for interpreting the results of unified exams:

• The results of the USE alone, without involving other data, can only be used to confirm the mastery of the secondary school program and the selection of candidates to continue their studies at the university.
• When evaluating the work of teachers, schools and educational systems, the results of the USE can only be used in combination with other “measurements”. nine0074
• For an objective comparison of learning outcomes in different territorial systems and schools, it is necessary to interpret the picture of the USE as correctly as possible. This will allow you to track the dynamics of changes.

However, in practice, the results of the USE are often interpreted incorrectly, which researchers explain by the lack of the necessary guidelines.

Frequent mistakes in interpreting USE results

USE is a procedure with high stakes. The student’s future depends on its results. This forces parents to reinforce school lessons with extracurricular ones – hire tutors for the child or send him to courses. Thus, it is not entirely correct to unequivocally link the successful passing of the USE with the high quality of school work, the authors of the article are convinced. nine0003

If a region or municipality has a low percentage of students taking a particular elective subject, it is impossible to draw conclusions about the quality of the local educational system or school based on the results obtained. The sample is not representative. As a rule, the most motivated students with the best results pass such exams, Sergey Bochenkov and Igor Valdman explain.

In addition, the system for interpreting USE results has its own inertia – a stable tradition of exam data analysis: analysis formats are usually fixed by law. This makes it easier for managers to keep track of ongoing changes. nine0003

And, finally, the average USE score is not as informative as it is sometimes presented. The variety of single data is difficult to express using the “average temperature in the hospital”. Alumni groups are too heterogeneous.

The “price” of the same scores in subjects is different

Sometimes the interpreters of the USE results compare numerically equal scores in different subjects and draw conclusions about the success of the exams based on them. However, this is not true. In the case of the USE, the arithmetic is not simple: the same scores in different subjects have different “costs”. So, 70 points in mathematics is a higher result than 70 points in Russian. It is necessary to take into account these differences, the researchers specify. nine0003

Threshold values ​​indicating the mastery of subjects at the basic level are also different for different subjects. So, to overcome the minimum threshold in mathematics, 24 points out of 100 are enough, in the Russian language – 36, and in computer science you need to score 40 points. The minimum score indicating a high level of knowledge is 63 in mathematics, 73 in Russian, and 84 out of 100 in computer science.

we are sure that they measure comparable characteristics, and their results are presented in comparable values,” emphasize Bochenkov and Waldman. nine0053 It makes sense, for example, to use indicators of the proportion of graduates who successfully coped with the minimum requirements of the standard in the subject, and the proportion of graduates who successfully passed the threshold of a high level of knowledge.

Evaluation for the teacher

The authors of the article list the questions that are important for the teacher when analyzing the results of the exam.

The general level of requirements for control measuring materials (CMM) for the subject. The source of information for the teacher is the open segment of the bank of tasks, demo versions, etc. (see, for example, on the website of the Federal Institute for Pedagogical Measurements). Tasks that occupy the same position in CIM (A1 or B5, for example) check the same content elements in all options within the subject. The more accurately the teacher imagines the level of tests, the more successfully he organizes the preparation of students. nine0003

Mastering certain content blocks. Data on the mastering of certain segments of the subject is important for the teacher – on the basis of their analysis, a solvability profile is compiled. So, for example, students can solve physical problems in mechanics successfully, but they find it difficult to solve problems in quantum physics. Identification of bottlenecks allows the teacher to adjust his work.

Estimation of demand for the subject among students. “Demand” for a subject is determined both by external factors (the prestige of the future specialty, etc.) and by internal factors related to the school and the teacher. The choice of a profile subject, one way or another, speaks about the work of a teacher. nine0003

Mastering the subject at the minimum and high (profile) level – what proportion of students have overcome both thresholds.

School grades

Evaluating the activities of schools with the help of the Unified State Examination, one must take into account, firstly, that examinations allow one to judge only certain aspects of the quality of education. Secondly, a simple comparison of the results of different schools does not reflect the real situation and can lead to incorrect management decisions. Bochenkov and Waldman emphasize that school indicators should be interpreted in relation to the indicators of a group of similar schools located in territories with similar socio-economic conditions. nine0054 The results of the USE are also affected by the presence of a filter in the selection of children for school, as well as the level of education of parents and their socio-economic status.