Statistical Safety Performance Models considering Pavement and Roadway Characteristics

Table 2

Summary statistics of key variables (1,811 segments).

Description	Mean	Standard deviation	Minimum	Maximum
Response variables
(1) N_Crash: number of crashes	16.6	46.2	0	642
Continuous variables
(2) Ln_Lgth: natural log of the segment length in feet	8.3	0.9	5.6	10.9
(3) Ln_AADT: natural log of average annual daily traffic	9.8	1.0	7.4	12.0
(4) Lq_MPD: largest quartile of mean profile depth	0.7	0.3	0.2	2.5
(5) Av_Frict: average friction value in skid number	40.1	6.8	16.9	62.0
(6) Av_Grad: average longitudinal grade	1.0	0.6	0	4.5
(7) Lg_Dgcv: largest degree of curve	1.2	3.5	0	66.6
(8) Ln_Lgcv: natural log of the length of curve in feet	1.8	3.1	0	9.7
Indicator variables (described with proportion for each category)
(9) PvType: pavement type	1. AC: 57.8%; 2. JCP: 30.6%; 3. CRCP: 11.7%
(10) RrloUrb: rural or urban	1. Rural: 55.5%; 2. Urban: 44.5%
(11) ExsMdn: existence of median	1. No: 10.7%; 2. Yes: 89.3%
(12) ExsShd: existence of shoulder	1. No: 4.1%; 2. Yes: 95.9%
(13) RnkIRI: rank of international roughness index	1. Good: 4.8%; 2. Acceptable: 19.8%; 3. Poor: 75.4%
(14) RnkRut: rank of maximum rutting depth	1. Good: 6.4%; 2. Acceptable: 23.3%; 3. Poor: 70.3%

Note. AC: asphalt pavement; JCP: jointed concrete pavement; and CRCP: continuously reinforced concrete pavement.