Skip to main content

Table 2 Summary of results from different census tract assignment methods (ZIP code centroids and geo-imputation).

From: Estimating the accuracy of geographical imputation

Population

Cases

Randoma

Geographic Centroidb

Population Centroidc

Overall Population Distribution

Population Distribution by Race/Ethnicity

Population Distribution by Race/Ethnicity-Age

 

N

Mean % (Min, Max %)

Total %

Total %

Mean % (Min, Max %)

Mean % (Min, Max %)

Mean % (Min, Max %)

All Cases

95,303

14.1 (13.5,14.7)

20.7

25.9

25.8 (25.3, 26.3)

26.7 (26.0, 27.4)

27.7 (26.5, 29.0)

Non-Hispanic White

75,577

15.0 (14.6, 15.4)

22.0

27.9

27.8 (27.4, 28.1)

28.2 (27.7, 28.7)

29.3 (28.4, 30.3)

Non-Hispanic Black

11,123

9.8 (9.0, 10.8)

15.3

16.6

16.5 (15.7, 17.4)

20.1 (18.7, 21.0)

21.3 (19.5, 23.2)

Hispanic

6,368

11.1 (10.0, 12.4)

16.3

19.2

18.8 (17.4, 20.0)

19.8 (18.7, 21.3)

20.2 (17.5, 22.7)

Asian

2,235

13.7 (11.6, 15.8)

17.3

25.4

26.0 (24.2, 28.5)

28.8 (26.4, 31.5)

28.8 (24.5, 33.5)

Age

       

20–49

11,931

14.2 (13.4, 15.2)

22.0

26.5

26.5 (26.4, 27.4)

27.5 (26.4, 28.4)

27.7 (25.6, 29.5)

50–64

29,452

14.2 (13.6, 14.8)

21.1

26.4

26.2 (25.6, 26.6)

27.0 (26.2, 27.3)

27.4 (26.2, 28.6)

65–84

47,798

14.1 (13.6, 14.5)

20.3

25.4

25.6 (25.1, 26.0)

26.5 (26.1, 27.2)

27.8 (26.8, 28.8)

>85

6,122

13.9 (12.8, 15.3)

19.4

25.9

25.2 (23.9, 26.4)

25.9 (24.1, 27.4)

29.1 (26.9, 31.3)

Population Density

       

<1,132

17,262

16.5 (15.9, 17.2)

29.6

29.6

30.8 (30.1, 31.2)

31.4 (30.3, 32.1)

31.8 (30.5, 33.1)

1,133–2,882

23,036

15.3 (14.9, 15.8)

24.1

29.0

30.2 (29.7, 31.0)

30.8 (30.2, 31.2)

32.4 (31.5, 33.4)

2,883–5,078

21,744

14.1 (13.4, 14.6)

19.2

27.1

25.6 (24.9, 26.2)

26.4 (26.7, 27.4)

27.8 (26.7, 28.7)

5,079–11,579

18,822

14.7 (13.9, 15.5)

17.1

26.3

24.9 (24.8, 25.3)

26.4 (25.5, 27.2)

26.8 (25.3, 28.2)

>11,579

14,439

8.6 (7.9, 9.3)

12.0

14.3

14.3 (13.6, 15.0)

15.5 (14.3, 16.2)

16.3 (14.9, 17.6)

  1. aRandom assignment was based on census tracts having equal probability within each postal ZIP code.
  2. bGeographic postal ZIP centroids were based on the center of the bounding rectangle for each ZIP code area.
  3. cPopulation weighted postal ZIP centroids were calculated as the mean of the census block centroid coordinates weighted by the number of persons per block.
  4. Source: Data from the New Jersey State Cancer Registry, New Jersey Department of Health and Senior Services, 2005.
Back to article page

International Journal of Health Geographics

ISSN: 1476-072X