18
EMPIRICAL RESULTS – ORDERED PROBIT ANALYSIS
3
X
Y = Protest
Y = Rebellion
Economic Grievances
.3151***
(.0710)
.0219
(.0897)
Political Grievances
.3696***
(.0802)
.0509
(.0998)
Autonomy Grievances
.1176*
(.0615)
.7222***
(.0865)
Religious Grievances
-.0399
(.0795)
.3696***
(.0939)
Language Grievances
1.253***
(.3633)
1.195**
(.4764)
Language Grievances * Religious Grievances
-.0519
(.0442)
-.1301**
(.0597)
Language Differential
.4873***
(.1173)
.1689
(.1613)
Level of Democracy
.0584***
(.0160)
.0189
(.0192)
GDP 1990
.0294***
(.0101)
.0069
(.0137)
GDP Growth % 1990’s
-1.264
(4.650)
-30.105***
(6.262)
Group Concentration
-.0479
(.1229)
.4629**
(.1827)
Organizational Cohesion
.2692***
(.0894)
-.0637
(.1105)
Overtness of Language Discrimination
-.2458
(.2218)
.1193
(.2576)
Cultural Differential
-.0211
(.0879)
-.1983*
(.1084)
Economic Differential
-.0811
(.0497)
-.0013
(.0621)
Political Differential
.0943
(.0618)
.0232
(.0775)
Ethno-linguistic Fractionalization
.6351*
(.3570)
1.488***
(.5025)
Group Population 1995
2.28e-5***
(6.28e-6)
-1.73e-5*
(9.13e-6)
Group Population Proportion 1995
3.570***
(.8401)
1.916*
(.9969)
Group has Exclusive Region within its Country?
.5091**
(.8401)
-.9271***
(.3077)
Group Primarily Urban?
.5022**
(.2505)
.0605
(.3961)
LG * Level of Democracy
-.0219**
(.0102)
.0026
(.0122)
LG * Relative Group Size
-1.191***
(.4195)
-.7212
(.5697)
LG * Ethno-linguistic Fractionalization
-.3375
(.2086)
-.9010***
(.2849)
LG * Group Concentration
-.1023**
(.0521)
.0840
(.0869)
LG * Overtness of Language Discrimination
.2055**
(.0954)
-.0675
(.1102)
LG * Organizational Cohesion
.0120
(.0551)
-.1532**
(.0670)
N = 314
N = 314
LRTS
4
= 221.49
LRTS = 204.83
P > Chi-Squared
5
= .0000
P > Chi-Squared = .0000
R-Squared
6
= .2134
R-Squared = .2422
*** =
Significant at 99%
** =
Significant at 95%
* =
Significant at 90%
3
2000
4
Likelihood Ratio Test Statistic
5
This is a measure of “goodness of fit.” Specifically, it is the probability of Type I Error – the probability that these independent variables,
when added to the model, explain no significant additional amount of variance in the dependent variable. The value cannot be exactly zero, but
in this case is extremely small.
6
McFadden’s Psuedo R-Squared