Skip to main content

Table 2 Overall distribution of self-rated health in original census data versus data from or imputed from survey data

From: Standard multiple imputation of survey data didn’t perform better than simple substitution in enhancing an administrative dataset: the example of self-rated health in England

  Census data Raw survey data Imputed from survey data
Standard logistic regression Poisson regression Ordinal logistic regression Multinomial logistic regression
% (N) % (N) % (95% CI) % (95% CI) % (95% CI) % (95% CI)
Self-rated health
 Very good 42.6 (592,633) 37.9 (50,792) 38.0 (37.7, 38.3) 38.2 (37.9, 38.5)
 Good 40.2 (558,445) 40.5 (54,281)    40.2 (39.8, 40.5) 40.1 (39.8, 40.4)
 Fair 12.1 (167,813) 15.4 (20,707)    15.6 (15.4, 15.8) 15.2 (14.9, 15.4)
 Bad 4.0 (55,535) 4.8 (6440)    4.8 (4.7, 4.9) 4.9 (4.8, 5.0)
 Very bad 1.1 (15,668) 1.4 (1863)    1.4 (1.3, 1.4) 1.6 (1.5, 1.8)
Self-rated health
 Very good/good/fair 94.9 (1,318,891) 93.8 (125,780) 93.7 (93.6, 93.9) 93.8 (93.7, 93.9) 93.8 (93.7, 94.0) 93.5 (93.3, 93.7)
 Bad/very bad 5.1 (71,203) 6.2 (8303) 6.3(6.1, 6.4) 6.2 (6.0, 6.3) 6.2 (6.0, 6.3) 6.5 (6.3, 6.7)