Skip to main content

Table 1 The association between double kernel density (DKD) of lung and NHL morbidity rates (cases per 100,000 residents) and distance to the revealed exposure sources (Method—bivariate regression, distance variables—linear and quadratic wind-adjusted distance terms)c

From: Spatial identification of potential health hazards: a systematic areal search approach

Variables

Model 1

Model 2

Ba and (tb)

Ba and (tb)

A. Lung cancer

 (Constant)

13.935 (58.947*)

1.131 (7.965*)

 Distance

−5.500E−0.40 (−19.350*)

0.002 (4.364*)

 Distance2

−1.115E−07 (−4.152*)

 No. of reference points

1000

1000

 \( {\text{R}}^{2} \)

0.286

0.301

 \( {\text{R}}_{\text{adjusted}}^{2} \)

0.285

0.299

 F

374.419*

133.819*

B. NHL cancer

 (Constant)

4.656 (17.237*)

−3.697 (−5.219*)

 Distance

3.380E−04 (8.409*)

0.003 (13.916*)

 Distance2

−2.189E−07 (−12.791*)

 No. of reference points

1000

1000

 \( {\text{R}}^{2} \)

0.070

0.205

 \( {\text{R}}_{\text{adjusted}}^{2} \)

0.069

0.204

 F

70.714*

120.722*

  1. Model 1: Bivariate linear model
  2. Model 2: Bivariate quadratic model
  3. * indicates a 0.01 two-tailed significance level
  4. aRegression coefficient
  5. b t-statistics in the parentheses
  6. cThe models reported in the table are estimated for the distances to the “best performing” source locations, marked by small triangles in Fig. 4, that is, source locations distances to which help to improve the models’ fits most significantly (see text for explanations)