The 1990 Post-Enumeration Survey: Operations and Results

Racial and ethnic disparities in breast cancer rates by age: NAACCR Breast Cancer Project

“…Rates may be biased because of errors in reporting race on death certificates [30] and hospital medical records (used to determine numerators for cancer incidence rates) and on censuses and surveys (used to determine denominators). Recent evaluation studies suggest that reporting race for white and black populations is generally reliable, but that biases are more serious for some of the smaller populations, particularly for AI [31,33]. Although inter-censal estimates of denominator data have been criticized as producing inaccurate results, a recent report noted that the effect on cancer rates is modest overall [34].…”

“…Measures of cancer incidence rates for various racial and ethnic populations may be limited by problems in ascertaining race or in differences in the quality of information (resulting from errors in reporting race and ethnicity on records used for cancer incidence and atrisk populations) [29][30][31][32][33]. Rates may be biased because of errors in reporting race on death certificates [30] and hospital medical records (used to determine numerators for cancer incidence rates) and on censuses and surveys (used to determine denominators).…”

Criteria for selecting a suitable method for producing post-2000 county population estimates: A case study of population estimates in Illinois

“…DA is a macro-level approach whereas PES/ACE is a micro-level approach where estimates of coverage are based on case-by-case matching with census records for a sample of the population (Robinson et al 1993). Details on PES (Hogan 1992) and ACE (U.S. Census Bureau 2001) are available elsewhere. The national net undercount rate for 1990 was 1.58 percent (Robinson et al 2001) and the net overcount rate for the 2000 census counts was 0.49 percent (U.S. Census Bureau 2003).…”

Measurement error in racial and ethnic statistics

“…Many epidemiologists call self-reported race or ethnicity ''the gold standard'' when studying how rates of morbidity or mortality vary between blacks and whites or Hispanics and non-Hispanics and recommend that, whenever possible, self-reports 5 According to the published estimates, black males 25-34 years were undercounted in the 1990 census by 13% (Hogan 1993). 6 A response in the postenumeration survey is taken to be better indicator of a person's race or ethnicity because (a) the record or coding of the response is taken to be more reliable or (b) the response is taken to be more thoughtful or considered and closer to the way the person really thinks of her race or ethnicity.…”