Times updates and expands value-added ratings for Los Angeles elementary school teachers

The Los Angeles Times on Sunday is releasing a major update to its elementary school teacher ratings, underscoring the large disparities throughout the nation’s second-largest school district in instructors’ abilities to raise student test scores.

The posting — the only publication of such teacher performance data in the nation — contains value-added ratings for about 11,500 third- through fifth-grade teachers, nearly double the number released last August. It also reflects changes in the way the scores were calculated and displayed.

Overall ratings for about 470 schools also are included in the release, which is based on student standardized test scores from the academic years 2003-04 through 2009-10. To obtain the rating of a teacher or school, go to latimes.com/valueadded and enter the teacher’s or the school’s name.

The initial release of teacher ratings last summer generated intense controversy — and some praise — across the country, and this round has already met with some opposition.

The Los Angeles Unified School District superintendent and other civic leaders, in a letter to the newspaper’s publisher, recently asked The Times to reconsider publishing the ratings, saying in part that individual teachers’ performances should be addressed in private conversations.

More than 1,000 teachers responded to The Times’ invitation to view their scores before publication, but few took the opportunity to write comments alongside their ratings. Instructors were strongly advised not to do so by their union, United Teachers Los Angeles, which has opposed publication of the ratings.Some of those who did comment said they saw the information as valuable but added that it did not reflect the sum total of their performance.

“Being a relatively new teacher, I welcome feedback that will help me to adjust my teaching to best fit my students’ needs,” said Amy Miller, who has taught fifth grade at Park Western Place Elementary School, one of about 140 teachers to write comments in the database. “It is, however, only one data point.”

Others denounced the newspaper, calling its statistics “invalid” and its ratings “a scarlet letter.”

“Once again you have violated the right to privacy of thousands of teachers,” wrote Patricia Hill, who has taught at Windsor Hills Math Science Aerospace Magnet School.

One teacher offered a suggestion: “How about publishing the names of the highly effective and effective teachers only,” said Steven Butts, who has taught fourth grade at Broadway Elementary School. “I do … believe that more positive recognition for the successful teachers would encourage all teachers to strive for excellence and seek the guidance from those who have proven results year after year.”

Value-added analysis attempts to estimate a teacher’s contribution to student learning by tracking students’ progress on standardized tests from year to year. Each student’s performance is compared with his or her own in past years, an approach that experts say largely controls for influences beyond a teacher’s reach, such as poverty, parenting and prior learning.

Because value-added is, like any other statistical approach, subject to error and is based only on test results, most experts agree it should be used as just one gauge of a teacher’s overall performance. But many say it is the most objective measure available, and districts around the country are adopting it, largely because of federal incentives.

In the interest of greater clarity and accuracy, The Times made several changes in its approach. More information is now shown about the precision of each estimate and how a teacher ranks relative to other teachers in the district. The analysis also takes into account additional variables related to a student’s socioeconomic background and the composition of a teacher’s class. As in the first release, The Times used data obtained from the district through the California Public Records Act.

The changes have not altered the broad conclusions reported by The Times last summer, and the ratings for most teachers changed very little overall.

As before, effective teachers were spread more or less evenly throughout the district. But there were often large disparities among instructors who taught similar students in similar schools — even within the same schools. The differences among teachers were more than three times as great as those among schools.

The vast majority of teachers’ ratings were not significantly influenced by the characteristics of the students they taught, and a teacher’s background and training had little to do with his or her performance, according to a study of the updated ratings prepared for The Times by economist Richard Buddin.

Many experts say it’s not enough for a value-added approach simply to compare a student’s performance to that in previous years—one must adjust for other factors such as race and ethnicity, parents’ education or classroom composition. Different value-added formulas adjust for different things—leading to different results.

To convey how much difference such adjustments can make, The Times is publishing a comparison of results from four value-added models for each teacher. Each model uses a slightly different combination of variables to estimate a teacher’s contribution. On average, the results are very similar but, in specific cases, they can vary sharply.

For its analysis, The Times adopted the model that included the most variables available.

More than a week ago, the newspaper received a letter from civic leaders requesting that The Times reconsider publication of the ratings. The letter was signed by John Deasy, L.A. Unified superintendent; Monica Garcia, Board of Education president; Elise Buik, chief executive of United Way of Greater Los Angeles; and Gary L. Toebben, chief executive of the Los Angeles Area Chamber of Commerce.

Among their concerns was that the ratings were likely to confuse educators and parents because the district has done its own value-added analysis of the data. It recently released overall school ratings to the public and it is planning to confidentially release individual ratings to third- through ninth-grade teachers by the end of the school year.

The model used by the district, which is seeking to persuade the union to include the ratings in formal teacher evaluations, is broadly similar to the one adopted by The Times. But The Times analyzed seven years of data, while the district analyzed up to four. The district and The Times adjusted the data to control for slightly different variables.

In addition, the civic leaders wrote that individual teacher evaluations should be conducted privately for the purpose of helping teachers improve.

Times Editor Russ Stanton said the newspaper went ahead with publication because it is confident of the reliability of its analysis and believes the public has a right to the information. Posting the database, Stanton said, “is a service to the people of Los Angeles.”

jason.song@latimes.com

jason.felch@latimes.com