DUPLICATES IN PARCC DATA
AN OVERVIEW
Damian W Betebenner
National Center for the Improvement of Educational Assessment
Dover, New Hampshire March15th, 2017
BACKGROUND
• Duplicate cases are defined as cases in the data set the are redundant on
CONTENT_AREA, YEAR, and ID
• That is, duplicate cases represent situations where more than one test
record exists in a given CONTENT_AREA and YEAR for a student.
• There are both legitimate and illegitimate reasons for duplicates of this sort.
• Currently we accommodate the duplicates (without getting into the issue
of whether they are legitimate or illegitimate).
• The problems with duplicates from a growth perspective is that we have to
accommodate all permutations of current score/prior score duplicates
leading to potentially a large number of SGPs.
EXAMPLE: STUDENT WITH DUPLICATES IN
ALL 3 YEARS: 8 PERMUTATIONS
Spring 2014 Spring 2015 Spring 2016
ANALYSIS
• Across all windows of data received thus far, there are 1,398 records
that would be considered duplicates.
• These records occur in all content areas and years as follows:
Algebra I Algebra II Geometry Mathematics ELA
ANALYSIS
• Students with duplicates in one year often have duplicates in other
years.
• For example, of the 214 ELA duplicates in 2016-2017.1, 106 duplicates
in 2015-2016.2 had IDs in that set. Not all of those duplicates fit into
analyzable grade sequences, but some did.
• As we being preparations for the Spring 2016-2017 analyses, we think
it is necessary to understand:
• Why these duplicates exist?
• What states would like done with these duplicates in terms of
growth analyses.
About this copy
Duplicates in PARCC Data — PARCC, March 2017.
This is an image-based copy of the original Keynote slides; builds and animations are flattened.