Teacher Keys Evaluation System (TKES)
Teacher Sampling Process for Pilot
Districts were allowed to select one of two random sampling options: (1) an across-school model in which approximately 10 percent of teachers across all schools in the district are selected to participate or (2) a whole-school model in which all teachers in selected schools in the district participate (to represent approximately 10 percent of teachers in the district). This process resulted in approximately 5,800 teachers being randomly selected to participate of the more than 48,000 teachers in the 26 RT3 districts.
This document provides a description of the steps that were used to sample teachers in those districts that selected Option 1.
1) An extract from CPI data was generated on October 24, 2011.
a) Only instructional job codes were included in the extract.
b) These data still contained errors since sample selection needed to be completed prior to final sign off of CPI by district superintendents.
c) To compensate for errors, oversampling was used knowing that some teachers would be disqualified from the sample.
2) A file for each district was creating including teacher information. The file was sorted by teacher last name and Excel random function was applied to assign a random value to each record. The file was resorted by this value.
3) The district received an Excel file through the GaDOE Portal for verification.
4) Districts reviewed files for accuracy. Errors were corrected and a final file submitted to the district for approval.