Table of Contents  |  Search Technical Documentation  |  References

NAEP Analysis and Scaling → Initial Activities → Constructed-Response Interrater Reliability → Range of Response Codes, Percent Exact Agreement, and Cohen's Kappa or Intraclass Correlation for the Constructed-Response Items Used in Scaling, by Block and Item, Age 9 Reading Long-Term Trend Bridge Study: 2004
Range of response codes, percent exact agreement, and Cohen's Kappa or intraclass correlation for the constructed-response items used in scaling, by block and item, age 9 reading long-term trend bridge study: 2004
Block Item Range of response codes Sample size Percent exact agreement Cohen's Kappa Intraclass correlation
H N001527 1–2 200 100 1.00
J N008905 1–2 200 100 1.00
L N002804 1–2 200 96 0.89
M N003104 1–2 100 96 0.58
N N003704 1–2 100 97 0.94
R21 N021203 1–5 200 77 0.85
† The intraclass correlation is not reported for dichotomously scored items; Cohen's Kappa is not reported for polytomously scored items.
NOTE: Cohen's Kappa is a measure of reliability that is appropriate for items that are dichotomously scored. The intraclass correlation coefficient is most appropriate for items with more than two categories.
SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), 2004 Reading Long-Term Trend Assessment.

Last updated 06 April 2009 (GF)

Printer-friendly Version