Table 2.
Agreement Between Telephone-Interviewer Estimate of Race and Self-reported Race Among Subsample (n = 24) of Refusal-Interview Participants, North Carolina, 2005a
| Category | Black | White | Total |
|---|---|---|---|
| Self-reported race | 4 | 20 | 24 |
| Interviewer-estimated race | 5 | 19 | 24 |
| Correctly identified | 4 | 19 | 23 |
| Incorrectly identified | 1 | 0 | 1 |
PRISM indicates Personally Relevant Information About Screening Mammography.
a
Sensitivity of interviewer estimates of black participants = 100% (4/4); positive predictive value of interviewer estimates of black participants = 80% (4/5); specificity of interviewer estimates of white participants = 95% (19/20). κ = 0.86; 95% confidence interval, 0.60–1.00.