closely guards the attendance records of its more than 400,000 students, citing strict laws on student privacy.
When the Tribune asked the district in May for attendance data and other information on all K-8 students, reporters framed the request in ways that would make it impossible to identify or track any individual child.
Over the next five months, CPS provided databases that fulfilled the reporters' basic requests. The data used in the Tribune's analysis are from the 2010-11 school year, the most recent period for which complete records were available.
The district eliminated certain information, such as the schools the students attended. Officials also redacted records on roughly 4,000 children whom they determined could potentially be identified.
Information provided included students' race, gender, special-education status, birth year, number of days enrolled, number of days attended, number of days missed because of excused or unexcused absences and in- or out-of-school suspensions.
Specific enrollment and exit details for the students also were provided, showing whether they had arrived in Chicago from another district midyear, for example, or left a CPS school for any reason. Dates for these changes in status were included.
The district also provided special education information for two disability categories, out of more than a dozen.
Other data sources the Tribune used to bolster its analysis included separate student-level enrollment data that CPS reported to the Illinois State Board of Education, the district's claims for attendance-based state aid, and a 2012 grant application that CPS submitted to the state for anti-truancy funding.
To reach the most accurate conclusions possible about elementary attendance patterns, and to eliminate weaknesses in the data, the Tribune excluded from its final analysis roughly 12 percent of the initial database of about 280,000 K-8 students.
Among the omitted students were those who had enrolled but left the district before the start of the school year, those whose records were inadvertently duplicated in the attendance database, and students at charter schools that report enrollment to CPS but not daily attendance.
Also excluded were 20,200 students whose enrollment exceeded the 170-day school year. Although Chicago's database did not include summer school, some students enrolled for as many as 221 days, or nearly every weekday of the year, and school officials offered only hypothetical scenarios to explain such implausibly high tallies.
Finally, the Tribune's analysis excluded 2,300 more students who were enrolled for fewer than six days during the school year.
For the remaining 246,600 students, the Tribune's goal was to calculate for the first time how many days of school each child had missed for any of three reasons: truancy, excused absences and enrollment gaps.
The CPS data explicitly listed days lost to truancy and excused absences. Missed enrollment days required further calculation to account for student mobility; for example, students who enroll late because they started the year in another district may have not missed any school time.
The entry codes and exit codes provided in the database allowed the Tribune to discount missed enrollment days for students who spent part of the year enrolled in private school, in another district or being home-schooled.