Jeff Dominitz and Robert Sherman, "Identification and Estimation of Bounds
on School Performance Measures: A Nonparametric Analysis of a Mixture Model
with Verification", Journal of Applied Econometrics, Vol 21, No. 8, 
2006, pp. 1295-1326.

The file ds-data.txt, which is zipped in ds-data.zip, is an ASCII file in
DOS format. It has 1638 rows and 4 columns.

The data describe the performance of 1638 9th-grade students on a math test 
taken in the year 2000 in the Pasadena Unified School District. The data 
are organized as follows:

variable 1: grade level (all 1638 observations take the value 9)
variable 2: test score (coded as -9 if test not taken)
variable 3: Limited English Proficiency indicator (0 if no, 1 if yes)
variable 4: school identifier (takes values 64, 80, 82, 84, 90)

Valid test scores range from 1 to 99.
The LEP indicator is an administrative classification.
The school identifier is an administrative code.