Differences in technology innovation R&D performance creation behavior between for-profit institutions and not-for-profit institutions

SpringerPlus

Table 5 Successive binary logistic regression analyses (not-for-profit institutions, n ₂ = 439)

Variable	Level	Coefficient (Z value)	Odds ratio	Coefficient (Z value)	Odds ratio	Coefficient (Z value)	Odds ratio
Intercept		−1.291 (−3.91)		−2.374 (−5.27)		−4.445 (−5.75)
X1		0.337 (7.59***)	1.40	0.110 (3.00***)	1.12	0.092 (1.86*)	1.10
T1	R	−0.122 (−0.52)	0.88	0.650 (1.82*)	1.92	0.569 (1.07)	1.77
T2	Cs	0.766 (1.95*)	2.15	−0.904 (−1.29)	0.40
	Cd	0.607 (1.89*)	1.83	0.269 (0.67)	1.31	0.865 (1.37)	2.38
B1	1			−0.769 (−2.39**)	0.46	−0.443 (−0.97)	0.64
B2	1					2.803 (7.19***)	16.50
Model significance test
Log likelihood		−254.046		−171.189		−92.820
Chi-Sq		98.574		25.799		72.502
DF		4		5		5
P value		0.000***		0.000***		0.000***
Measures of association
Somers’ D		0.55		0.39		0.73
Goodman–Kruskal γ		0.55		0.39		0.73
Kendall’s τ_A		0.27		0.10		0.14

Prediction pair type	Count	Proportion	Count	Proportion	Count	Proportion
Concordant pairs	36,985	0.771	16,712	0.687	11,650	0.858
Discordant pairs	10,759	0.224	7272	0.299	1800	0.133
Tied pairs	226	0.005	326	0.013	121	0.009
Total	47,970	1.000	24,310	1.000	13,571	1.000

^aReference case (i.e., Success)
^bReference level
^cA partial sample composed of 67 observations accompanied with T2 = Cs is discarded because all these observations have B3 = 0 solely