International Scholarly Research Notices

Research Article

Entrepreneurial Environment and the Prevalence of Diabetes in U.S. Counities

Table 2

Classic OLS, spatial lag model, and spatial error model of the county-level diabetes prevalence rate (2007).
Classic OLSSpatial lagSpatial error
𝑏 s.e. 𝑏 s.e. 𝑏 s.e.
Number of small businesses per 1,000 people−0.0129***0.0013−0.0063***0.0011−0.0094***0.0012
Percent of population uninsured−0.0096+0.00580.0109*0.00460.0312***0.0057
Income inequality (gini coefficient)0.0983***0.00920.01030.0074−0.0191*0.0081
Per capita income (in thousand dollars)−0.0585***0.0103−0.0155+0.0082−0.0263**0.0093
Percent of population nonhispanic black0.0434***0.00210.0249***0.00180.0417***0.0027
Percent of population hispanic−0.0215***0.0027−0.0093***0.0021−0.0248***0.0034
Number of physicians per 1,000 people−0.0236**0.0073−0.0174**0.0057−0.0102*0.0051
Population size, natural log−0.0834***0.0242−0.0346+0.0191−0.0876***0.0211
Percent of population age 65 and older (2000)0.1157***0.00690.1090***0.00550.1502***0.0063
Age adjusted mortality rate, 1994−19980.0052***0.00020.0035***0.00020.0024***0.0002
County part of metropolitan area 0.2428***0.06170.0919+0.04860.03650.0470
Spatial parameter  0.5596***0.01450.7779***0.0136
Intercept1.7286***0.4573−0.27350.36967.2912***0.4097
R-square0.63020.77080.8049
Log likelihood−4955.88−4327.05−4211.06
Akaike info criterion (AIC)9935.778680.098446.12
Schwarz criterion (SC)10008.108758.438518.43
Wald  1482.253249.42
Likelihood ratio (LR)  1257.67 1489.64
Lagrange multiplier (LM) 1429.44 (236.32)1477.84 (284.72)
Two-tailed tests: *** 𝑃 < . 0 0 1 ; ** 𝑃 < . 0 1 ; * 𝑃 < . 0 5 ; + 𝑃 < 0 .1.
𝑁 = 3 , 0 6 0 .
Numbers in the parentheses are robust LM values.