electors.RdThis is an artificial data set on electoral choice as influenced by class and party positions.
data(electors)A data frame containing the following variables:
class position of voters
party that runs for election
freqency by which each party list is chosen by members of each class
time variable, runs from zero to one
economic-policy "leftness" of each party
emphasis of welfare expansion of each party
position on authoritarian issues
data(electors)
summary(mclogit(
cbind(Freq,interaction(time,class))~econ.left+welfare+auth,
data=electors))
#>
#> Iteration 1 - deviance = 85051.49 - criterion = 0.9989204
#> Iteration 2 - deviance = 76759.94 - criterion = 0.108019
#> Iteration 3 - deviance = 74896.56 - criterion = 0.02487934
#> Iteration 4 - deviance = 74890.9 - criterion = 7.559543e-05
#> Iteration 5 - deviance = 74890.9 - criterion = 1.726814e-09
#> converged
#>
#> Call:
#> mclogit(formula = cbind(Freq, interaction(time, class)) ~ econ.left +
#> welfare + auth, data = electors)
#>
#> Estimate Std. Error z value Pr(>|z|)
#> econ.left -0.507265 0.007495 -67.679 < 2e-16 ***
#> welfare 0.564650 0.010700 52.769 < 2e-16 ***
#> auth 0.030305 0.005749 5.271 1.36e-07 ***
#> ---
#> Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
#>
#> Null Deviance: 80580
#> Residual Deviance: 74890
#> Number of Fisher Scoring iterations: 5
#> Number of observations: 37500
#>
#>
summary(mclogit(
cbind(Freq,interaction(time,class))~econ.left/class+welfare/class+auth/class,
data=electors))
#>
#> Iteration 1 - deviance = 7377.939 - criterion = 0.9875551
#> Iteration 2 - deviance = 4589.544 - criterion = 0.6075407
#> Iteration 3 - deviance = 4293.485 - criterion = 0.06895374
#> Iteration 4 - deviance = 4277.887 - criterion = 0.00364612
#> Iteration 5 - deviance = 4277.808 - criterion = 1.852771e-05
#> Iteration 6 - deviance = 4277.808 - criterion = 5.890781e-10
#> converged
#>
#> Call:
#> mclogit(formula = cbind(Freq, interaction(time, class)) ~ econ.left/class +
#> welfare/class + auth/class, data = electors)
#>
#> Estimate Std. Error z value Pr(>|z|)
#> econ.left -0.77851 0.02312 -33.671 < 2e-16 ***
#> welfare 3.43776 0.03170 108.431 < 2e-16 ***
#> auth -0.13740 0.03608 -3.808 0.00014 ***
#> econ.left:classnew.middle 0.44546 0.02588 17.212 < 2e-16 ***
#> econ.left:classold.middle -0.44082 0.10387 -4.244 2.2e-05 ***
#> classnew.middle:welfare -3.12917 0.03696 -84.659 < 2e-16 ***
#> classold.middle:welfare -5.27438 0.07286 -72.393 < 2e-16 ***
#> classnew.middle:auth -0.86676 0.03947 -21.957 < 2e-16 ***
#> classold.middle:auth 1.39435 0.05615 24.831 < 2e-16 ***
#> ---
#> Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
#>
#> Null Deviance: 80580
#> Residual Deviance: 4278
#> Number of Fisher Scoring iterations: 6
#> Number of observations: 37500
#>
#>
if (FALSE) # This takes a bit longer.
summary(mclogit(
cbind(Freq,interaction(time,class))~econ.left/class+welfare/class+auth/class,
random=~1|party.time,
data=within(electors,party.time<-interaction(party,time))))
summary(mclogit(
cbind(Freq,interaction(time,class))~econ.left/(class*time)+welfare/class+auth/class,
random=~1|party.time,
data=within(electors,{
party.time <-interaction(party,time)
econ.left.sq <- (econ.left-mean(econ.left))^2
})))
#>
#> Iteration 1 - deviance = 1071.031 - criterion = 0.1597241
#> Iteration 2 - deviance = 965.6196 - criterion = 0.02540274
#> Iteration 3 - deviance = 948.8356 - criterion = 0.005154655
#> Iteration 4 - deviance = 947.6262 - criterion = 0.0002054859
#> Iteration 5 - deviance = 947.5081 - criterion = 2.557556e-07
#> Iteration 6 - deviance = 947.5042 - criterion = 4.672682e-13
#> converged
#>
#> Call:
#> mclogit(formula = cbind(Freq, interaction(time, class)) ~ econ.left/(class *
#> time) + welfare/class + auth/class, data = within(electors,
#> {
#> party.time <- interaction(party, time)
#> econ.left.sq <- (econ.left - mean(econ.left))^2
#> }), random = ~1 | party.time)
#>
#> Coefficents:
#> Estimate Std. Error z value Pr(>|z|)
#> econ.left -0.13335 0.20837 -0.640 0.522
#> welfare 2.05552 0.21245 9.675 <2e-16 ***
#> auth 0.08071 0.11717 0.689 0.491
#> econ.left:classnew.middle -1.69581 0.11631 -14.580 <2e-16 ***
#> econ.left:classold.middle -3.04338 0.20351 -14.954 <2e-16 ***
#> econ.left:time -0.07782 0.30228 -0.257 0.797
#> classnew.middle:welfare -0.99267 0.06073 -16.346 <2e-16 ***
#> classold.middle:welfare -1.62088 0.12850 -12.614 <2e-16 ***
#> classnew.middle:auth -1.39056 0.04673 -29.754 <2e-16 ***
#> classold.middle:auth 1.45722 0.05814 25.063 <2e-16 ***
#> econ.left:classnew.middle:time 0.06049 0.14449 0.419 0.675
#> econ.left:classold.middle:time 0.14723 0.26232 0.561 0.575
#> ---
#> Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
#>
#> (Co-)Variances:
#> Grouping level: party.time
#> Estimate Std.Err.
#> (Const.) (Const.)
#> (Const.) 1.604 0.3098
#>
#> Approximate residual deviance: 947.5
#> Number of Fisher scoring iterations: 6
#> Number of observations
#> Groups by party.time: 150
#> Individual observations: 37500
#>