Skip to main content

Table 3 Prediction ability of neighborhood audit item responses by spatial scale (large/small) and rater adjustment

From: Spatial predictive properties of built environment characteristics assessed by drop-and-spin virtual neighborhood auditing

Audit item

Detrending

ROC AUC prediction*

Large-scale/Rater

+ Small-scale Kriging

%Change of rater adjustment, full equation

%Change large- vs small-scale

Garbage

3rd order spatial

0.755

0.856

 

13.4

3rd order spatial + rater

0.874

0.919

7.36

5.1

Abandoned cars

3rd order spatial

0.881

0.996

 

13.1

3rd order spatial + rater

0.952

0.994

− 0.2

4.4

Building conditions ≥ moderate

3rd order spatial

0.852

0.902

 

5.9

3rd order spatial + rater

0.880

0.916

1.55

4.1

Yard conditions ≥ moderate

3rd order spatial

0.897

0.934

 

4.1

3rd order spatial + rater

0.921

0.946

1.28

2.7

Dumpster

3rd order spatial

0.714

0.842

 

17.9

3rd order spatial + rater

0.714

0.849

0.83

18.9

Graffiti

3rd order spatial

0.871

0.914

 

4.9

3rd order spatial + rater

0.887

0.923

0.98

4.1

Boarded/burned building

3rd order spatial

0.851

0.914

 

7.4

3rd order spatial + rater

0.850

0.915

0.11

7.6

Outdoor seating

3rd order spatial

0.562

0.735

 

30.8

3rd order spatial + rater

0.679

0.767

4.35

13

Team sports

3rd order spatial

0.591

0.915

 

54.8

3rd order spatial + rater

0.683

0.894

− 2.3

30.9

Yard decorations

3rd order spatial

0.667

0.825

 

23.7

3rd order spatial + rater

0.708

0.852

3.27

20.3

Fences

3rd order spatial

0.710

0.884

 

24.5

3rd order spatial + rater

0.736

0.897

1.47

21.9

Sidewalk present

3rd order spatial

0.814

0.950

 

16.7

3rd order spatial + rater

0.814

0.943

− 0.74

15.8

Complete sidewalk

3rd order spatial

0.683

0.842

 

23.3

3rd order spatial + rater

0.726

0.855

1.54

17.8

Sidewalk condition

3rd order spatial

0.662

0.771

 

16.5

3rd order spatial + rater

0.697

0.775

0.52

11.2

Sidewalk width

3rd order spatial

0.830

0.897

 

8.1

3rd order spatial + rater

0.831

0.901

0.45

8.4

Sidewalk from curb distance

3rd order spatial

0.586

0.978

 

66.9

3rd order spatial + rater

0.724

0.914

− 6.54

26.2

Car obstruction

3rd order spatial

0.626

0.879

 

40.4

3rd order spatial + rater

0.663

0.881

0.23

32.9

Garbage can obstruction

3rd order spatial

0.674

0.929

 

37.8

3rd order spatial + rater

0.695

0.917

− 1.29

31.9

Pole or sign obstruction

3rd order spatial

0.592

0.900

 

52

3rd order spatial + rater

0.726

0.876

− 2.67

20.7

Other obstruction

3rd order spatial

0.689

0.870

 

26.3

3rd order spatial + rater

0.694

0.871

0.11

25.5

Curb cuts

3rd order spatial

0.559

0.710

 

27

3rd order spatial + rater

0.669

0.716

0.85

7

Clear intersection

3rd order spatial

0.681

0.851

 

25

3rd order spatial + rater

0.684

0.852

0.12

24.6

Pedestrian crossing sign

3rd order spatial

0.669

0.827

 

23.6

3rd order spatial + rater

0.664

0.809

− 2.18

21.8

Pedestrian signal

3rd order spatial

0.651

0.908

 

39.5

3rd order spatial + rater

0.679

0.897

− 1.21

32.1

Pedestrian crossing marks

3rd order spatial

0.731

0.871

 

19.2

3rd order spatial + rater

0.725

0.860

− 1.26

18.6

Type of pedestrian crosswalk marks

3rd order spatial

0.697

0.931

 

33.6

3rd order spatial + rater

0.709

0.934

0.32

31.7

traffic signal type

3rd order spatial

0.661

0.841

 

27.2

3rd order spatial + rater

0.699

0.838

− 0.36

19.9

One-way street

3rd order spatial

0.789

0.931

 

18

3rd order spatial + rater

0.793

0.923

− 0.86

16.4

Number of lanes

3rd order spatial

0.676

0.920

 

36.1

3rd order spatial + rater

0.674

0.948

3.04

40.7

Presence of highway

3rd order spatial

0.838

0.993

 

18.5

3rd order spatial + rater

0.838

0.993

0

18.5

Highway is barrier

3rd order spatial

0.563

0.833

 

48

3rd order spatial + rater

0.479

0.583

− 30.01

21.7

  1. *The area under the curve (AUC) of receiver operator curves (ROC) resulting from each item-specific logistic regression of the validation dataset was calculated where observed audit item response was the dependent variable and predicted response probability was the single independent variable measured as a continuous variable