Soon after the U.S. News 2012 Best High Schools rankings were published in May 2012, it became clear that there were errors in the 2009-2010 enrollment data used in the calculations. The data came from the U.S. Department of Education's Common Core of Data (CCD) and the errors affected a small number of the ranked high schools.
[See the detailed methodology used in Best High Schools rankings.]
One reason why U.S. News ranks the Best High Schools and publishes data on nearly every U.S. public high school is to help consumers by standardizing the reporting and transparency of high school-level data from every state. As part of this goal, I recently attended the U.S. Department of Education's National Center for Education Statistics' (NCES) 25th Annual STATS-DC 2012 Data Conference and was on a panel titled "High School Rankings by the Media: What We Learned About the Importance of Data Quality in the Common Core of Data and Opportunities for Improvements."
Other speakers on the panel included Marilyn Seastrom, Acting Deputy Commissioner and Chief Statistician, NCES; Marie Stetser, NCES Program Director, CCD, NCES; Julian Montoya, Nevada Department of Education; and, John Gonzalez, New York City Public Schools.
The following are some of the key points I discussed.
• The Best High Schools ranking variables that were affected by the CCD data errors were:
1. The percent of economically disadvantaged students in each high school, which relied on 2009-2010 CCD data. It was the basis for determining top-performing schools in each state in an analysis of how students did on state assessments in reading and mathematics when controlling for poverty.
2. The College Readiness Index, which was the sole factor used to decide the national rankings and used each high school's 2009-2010 12th grade enrollment, provided by the CCD, as the denominator.
• On May 30, 2012, U.S. News changed the status of the first four high schools with CCD data errors from ranked to unranked. Read more about this in a previous blog post, Adjusting the Best High Schools Rankings for Government Data Errors.
• NCES started rechecking the entire CCD database to ensure the accuracy of the data and issued its final list of high schools with CCD data errors in early June. Based on that final NCES list, U.S. News found that a total of 17 nationally ranked gold, silver, or bronze medal high schools, or only around 0.003 percent of the total that were ranked, had incorrect CCD data that impacted their 2012 Best High Schools ranking. U.S. News will change the remaining 13 nationally ranked schools with CCD data errors from being ranked to unranked on usnews.com.
• U.S. News plans to publish new Best High Schools rankings in 2013 and intends to use the CCD's 2010-2011 school year data in the analysis as it becomes available.
We are encouraged that the NCES says that it and its state partners are implementing additional quality control steps on the new CCD data in order to improve its accuracy.