A poll was taken by 538 on people’s feelings about Star Wars movies and Star Wars characters. The most surprising result from our analysis is that respondents’ feelings about Han Solo seem not to be colored by whether they believe Han shot first or Greedo did.
The most statistically significant variable we found was a multiple of people’s feelings on Anakin Skywalker and their feelings about the Phantom Menace. We put these variables together because they are likely to be correlated with each other.
We also found out that age is a statistically significant factor and that significance is strengthened if we put groups in a non-ordinal order. The two lowest age groups are most likely to have experienced one of the trilogies as a child. When we give them the furthest separated scores, we get the most significance.
Both models have very high significance by the chi-squared test. Our larger model has a lower AIC value. But the more spare model has significance at .05 for all three variables. It has a p-value of 8.094944e-08 for its chi-square test.
for (i in 1:1186){anakin.times.menace[i]<-(feelings.on.anakin[i]*phantom.menace.ranking[i])}
glmResults<-glm(who.shot.first ~ feelings.on.vader+feelings.on.lando+feelings.on.han+feelings.on.jarjar+
star.trek.fan+family.income+empire.strikes.back.ranking+rearranged.age.group+anakin.times.menace, family="binomial")
summary(glmResults)##
## Call:
## glm(formula = who.shot.first ~ feelings.on.vader + feelings.on.lando +
## feelings.on.han + feelings.on.jarjar + star.trek.fan + family.income +
## empire.strikes.back.ranking + rearranged.age.group + anakin.times.menace,
## family = "binomial")
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -2.0477 -1.2126 0.7214 0.9893 1.5020
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.129921 1.459340 0.089 0.9291
## feelings.on.vader 0.150473 0.082569 1.822 0.0684 .
## feelings.on.lando -0.081140 0.132237 -0.614 0.5395
## feelings.on.han -0.081558 0.217789 -0.374 0.7080
## feelings.on.jarjar 0.059467 0.090010 0.661 0.5088
## star.trek.fan -0.136520 0.235906 -0.579 0.5628
## family.income -0.004766 0.096987 -0.049 0.9608
## empire.strikes.back.ranking 0.047023 0.086379 0.544 0.5862
## rearranged.age.group -0.168207 0.100297 -1.677 0.0935 .
## anakin.times.menace 0.079291 0.019584 4.049 5.15e-05 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 479.23 on 361 degrees of freedom
## Residual deviance: 452.52 on 352 degrees of freedom
## (824 observations deleted due to missingness)
## AIC: 472.52
##
## Number of Fisher Scoring iterations: 41-pchisq(479.23 , 361)## [1] 2.923399e-051-pchisq(452.47 ,352)## [1] 0.00023166821-pchisq(479.23-452.47 ,361-352)## [1] 0.001532475glmResults<-glm(who.shot.first ~feelings.on.vader+rearranged.age.group+anakin.times.menace, family="binomial")
summary(glmResults)##
## Call:
## glm(formula = who.shot.first ~ feelings.on.vader + rearranged.age.group +
## anakin.times.menace, family = "binomial")
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -2.0274 -1.2245 0.7174 0.9817 1.4447
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.34702 0.34259 -1.013 0.31109
## feelings.on.vader 0.19670 0.06394 3.077 0.00209 **
## rearranged.age.group -0.19856 0.08573 -2.316 0.02056 *
## anakin.times.menace 0.06919 0.01511 4.580 4.65e-06 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 650.36 on 492 degrees of freedom
## Residual deviance: 612.35 on 489 degrees of freedom
## (693 observations deleted due to missingness)
## AIC: 620.35
##
## Number of Fisher Scoring iterations: 41-pchisq(650.36 , 492)## [1] 2.005601e-061-pchisq(614.52 , 489 )## [1] 9.389717e-051-pchisq(650.36-614.52 , 492-489)## [1] 8.094944e-08