Vote Legitimacy Analyses
Dani Grant & Alex Flores
4/30/2021
- I. Import all datasets used for processing or analyses; perform alphas and correlations for measures
- II. Sample Descriptives
- III. Chi-Square
- 1. Election Expectations
- 2. Perceived Election Legitimacy
- a. Descriptives
- b. ANOVA: Vote Confidence = Partisanship * Timing * Vote Type
- c. Simple Effects of Timing, at each level of Party
- d. Own vs. National Vote Models
- g. Moderators/Mediators
- h. assessing differences in the pDem_Rep*tDur_Post parameter when positive emotion is added a mediator
- 3. Emotions
- 4. Media Trust/Consumption
- a. descriptives
- b. (other - fox perception) ~ party
- c. vote legitimacy ~ party * media type - for 2 df test
- d. vote legitimacy ~ party
- e. anova between simple and augmented models: Interaction of media type and partisanship
- f. other method, using OtherMedia - Fox to account for the two-level factor of media type! Instead of using them both separately.
- g. Effect sizes by party for each news outlet
- h. Does ID strength moderate?
- 5. Media Analyses: Vote legitimacy ~ party * foxPerception * otherMediaPerception
- 9. Plots
I. Import all datasets used for processing or analyses; perform alphas and correlations for measures
II. Sample Descriptives
d$age <- as.numeric(d$age)
nrow(d) #N = 1236
# exclude participants out of proper age range
d$age <- ifelse(d$age > 120, NA,
ifelse(d$age < 18, NA,
d$age))
nrow(d) #N = 1236
psych::describe(d$age)
#M = 49.71, SD = 15.21, Range = 18 - 89
table(d$gender, exclude = F) # F = 1, M = 2, Other = 3
prop.table(table(d$gender, exclude = F)) #, exclude = F))
# 1 2 3 NA
# 538(43.6%) 579(46.9%) 2(0.1%) 116(9.4%)
prop.table(table(d$race))
# 1 2 3 4 5 6 7
# 8.9% 12.7% 7.4% 1.1% 0.5% 68.5% 1.1%
prop.table(table(d$race,exclude = F))
# 1 = Asian, Asian-American
# 2 = Black, African-American
# 3 = Hispanic, Latino-American
# 4 = native American
# 5 = Native Pacific Islander
# 6 = White, Caucasian-American
# 7 = other
# 1 2 3 4 5 6 7 NA
# 8.2% 11.7% 6.8% 0.1% <0.1% 62.7% 0.1% 8.1%
psych::describe(d$education)
# M = 14.48, SD = 2.62
table(d$party_factor, exclude = F)
# Democrat Republican Independent
# 565 466 194
table(d$election_timing, exclude = F)III. Chi-Square
a. Chi-Square for response rates
####
# response rates for during/post election
####
# 584 = total responses for during election; 656 = total responses post-election
# number retrieved from email with vendor
t <- matrix(c((583/836), (653/836)), ncol = 2, byrow = TRUE)
t2 <- matrix(c(583,836, 653,836), ncol = 2, byrow = F)
colnames(t) <- c("during", "post")
t## during post
## [1,] 0.6973684 0.7811005## Warning in chisq.test(t): Chi-squared approximation may be incorrect##
## Chi-squared test for given probabilities
##
## data: t
## X-squared = 0.0047421, df = 1, p-value = 0.9451##
## Fisher's Exact Test for Count Data
##
## data: t2
## p-value = 0.1335
## alternative hypothesis: true odds ratio is not equal to 1
## 95 percent confidence interval:
## 0.7684631 1.0372274
## sample estimates:
## odds ratio
## 0.89283851. Election Expectations
a. Descriptives
##
## Descriptive statistics by group
## : Democrat
## : During-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 187 7.16 2.03 8 7.44 1.48 1 9 8 -1.05 -0.01 0.15
## ------------------------------------------------------------
## : Republican
## : During-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 190 3.63 2.47 3 3.34 2.97 1 9 8 0.77 -0.51 0.18
## ------------------------------------------------------------
## : Independent
## : During-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 71 4.87 2.03 5 4.88 1.48 1 9 8 0.02 -0.48 0.24
## ------------------------------------------------------------
## : Democrat
## : Post-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 307 7.58 2.04 8 8 1.48 1 9 8 -1.58 1.66 0.12
## ------------------------------------------------------------
## : Republican
## : Post-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 226 3.63 2.78 2 3.31 1.48 1 9 8 0.68 -0.95 0.18
## ------------------------------------------------------------
## : Independent
## : Post-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 103 5.73 2.42 6 5.87 2.97 1 9 8 -0.27 -0.91 0.24##
## Pearson's product-moment correlation
##
## data: d$ownvote_conf and d$overallvote_conf
## t = 39.45, df = 1206, p-value < 2.2e-16
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
## 0.7248916 0.7742292
## sample estimates:
## cor
## 0.7506048b. two-way ANOVA: Election Expectation ~ Partisanship * Election Timing
pred.ma <- lm(electPredictTB.plot ~ pDem_Rep + pInd_Not + tDur_Post + pInd_Not:tDur_Post + pDem_Rep:tDur_Post, data = d)
# Effect size calculations
mcSummary(pred.ma)## Loading required package: car## Loading required package: carData##
## Attaching package: 'car'## The following object is masked from 'package:dplyr':
##
## recode## The following object is masked from 'package:psych':
##
## logit## lm(formula = electPredictTB.plot ~ pDem_Rep + pInd_Not + tDur_Post +
## pInd_Not:tDur_Post + pDem_Rep:tDur_Post, data = d)
##
## Omnibus ANOVA
## SS df MS EtaSq F p
## Model 3307.008 5 661.402 0.363 122.74 0
## Error 5808.958 1078 5.389
## Corr Total 9115.966 1083 8.417
##
## RMSE AdjEtaSq
## 2.321 0.36
##
## Coefficients
## Est StErr t SSR(3) EtaSq tol CI_2.5 CI_97.5
## (Intercept) 0.435 0.079 5.500 162.979 0.027 NA 0.280 0.590
## pDem_Rep -3.738 0.157 -23.811 3055.249 0.345 0.967 -4.046 -3.430
## pInd_Not 0.200 0.195 1.025 5.665 0.001 0.965 -0.183 0.584
## tDur_Post 0.423 0.158 2.674 38.530 0.007 0.819 0.113 0.733
## pInd_Not:tDur_Post -0.645 0.391 -1.649 14.650 0.003 0.799 -1.412 0.123
## pDem_Rep:tDur_Post -0.418 0.314 -1.332 9.561 0.002 0.968 -1.034 0.198
## p
## (Intercept) 0.000
## pDem_Rep 0.000
## pInd_Not 0.305
## tDur_Post 0.008
## pInd_Not:tDur_Post 0.099
## pDem_Rep:tDur_Post 0.183elect.aov <- ezANOVA(data = d[!is.na(d$electPredictTB),]
, dv = electPredictTB
, wid = s3
, between = .(party_factor, election_timing)
, type = 3
, return_aov = T)## Warning: Converting "s3" to factor for ANOVA.## Warning: You have removed one or more levels from variable "party_factor".
## Refactoring for ANOVA.## Warning: Data is unbalanced (unequal N per group). Make sure you specified a
## well-considered value for the type argument to ezANOVA().## Coefficient covariances computed by hccm()## $ANOVA
## Effect DFn DFd F p p<.05
## 2 party_factor 2 1078 284.756558 5.137423e-100 *
## 3 election_timing 1 1078 7.173209 7.512570e-03 *
## 4 party_factor:election_timing 2 1078 2.195529 1.117971e-01
## ges
## 2 0.345680475
## 3 0.006610198
## 4 0.004056812
##
## $`Levene's Test for Homogeneity of Variance`
## DFn DFd SSn SSd F p p<.05
## 1 5 1078 146.7436 3106.12 10.18567 1.500728e-09 *
##
## $aov
## Call:
## aov(formula = formula(aov_formula), data = data)
##
## Terms:
## party_factor election_timing party_factor:election_timing
## Sum of Squares 3255.851 27.495 23.662
## Deg. of Freedom 2 1 2
## Residuals
## Sum of Squares 5808.958
## Deg. of Freedom 1078
##
## Residual standard error: 2.321345
## Estimated effects may be unbalanced
## 1 observation deleted due to missingnessc. main effect for partisan identity - 2 df test
pred.mc <- lm(electPredictTB.plot ~ tDur_Post + pInd_Not:tDur_Post + pDem_Rep:tDur_Post, data = d)
summary(pred.mc)##
## Call:
## lm(formula = electPredictTB.plot ~ tDur_Post + pInd_Not:tDur_Post +
## pDem_Rep:tDur_Post, data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -5.2107 -2.3954 0.1795 2.7893 4.0872
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.56178 0.08866 6.337 3.44e-10 ***
## tDur_Post 0.57374 0.19452 2.949 0.00325 **
## tDur_Post:pInd_Not -0.27633 0.47475 -0.582 0.56065
## tDur_Post:pDem_Rep -1.63073 0.38258 -4.262 2.20e-05 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 2.867 on 1080 degrees of freedom
## (152 observations deleted due to missingness)
## Multiple R-squared: 0.02612, Adjusted R-squared: 0.02341
## F-statistic: 9.655 on 3 and 1080 DF, p-value: 2.721e-06## Analysis of Variance Table
##
## Model 1: electPredictTB.plot ~ tDur_Post + pInd_Not:tDur_Post + pDem_Rep:tDur_Post
## Model 2: electPredictTB.plot ~ pDem_Rep + pInd_Not + tDur_Post + pInd_Not:tDur_Post +
## pDem_Rep:tDur_Post
## Res.Df RSS Df Sum of Sq F Pr(>F)
## 1 1080 8877.9
## 2 1078 5809.0 2 3068.9 284.76 < 2.2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1d. Comparison of Reps to midpoint
electOutcome.R <- lm(electPredictTB.plot ~ (pRepD + pRepI) * tDur_Post, data = d)
summary(electOutcome.R)##
## Call:
## lm(formula = electPredictTB.plot ~ (pRepD + pRepI) * tDur_Post,
## data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -6.5798 -1.7282 0.4202 1.4202 5.3684
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -1.367839 0.114242 -11.973 < 2e-16 ***
## pRepD 3.737955 0.156982 23.811 < 2e-16 ***
## pRepI 1.668536 0.212378 7.856 9.52e-15 ***
## tDur_Post 0.001164 0.228484 0.005 0.9959
## pRepD:tDur_Post 0.418212 0.313964 1.332 0.1831
## pRepI:tDur_Post 0.853751 0.424757 2.010 0.0447 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 2.321 on 1078 degrees of freedom
## (152 observations deleted due to missingness)
## Multiple R-squared: 0.3628, Adjusted R-squared: 0.3598
## F-statistic: 122.7 on 5 and 1078 DF, p-value: < 2.2e-16## lm(formula = electPredictTB.plot ~ (pRepD + pRepI) * tDur_Post,
## data = d)
##
## Omnibus ANOVA
## SS df MS EtaSq F p
## Model 3307.008 5 661.402 0.363 122.74 0
## Error 5808.958 1078 5.389
## Corr Total 9115.966 1083 8.417
##
## RMSE AdjEtaSq
## 2.321 0.36
##
## Coefficients
## Est StErr t SSR(3) EtaSq tol CI_2.5 CI_97.5 p
## (Intercept) -1.368 0.114 -11.973 772.500 0.117 NA -1.592 -1.144 0.000
## pRepD 3.738 0.157 23.811 3055.249 0.345 0.813 3.430 4.046 0.000
## pRepI 1.669 0.212 7.856 332.606 0.054 0.818 1.252 2.085 0.000
## tDur_Post 0.001 0.228 0.005 0.000 0.000 0.393 -0.447 0.449 0.996
## pRepD:tDur_Post 0.418 0.314 1.332 9.561 0.002 0.455 -0.198 1.034 0.183
## pRepI:tDur_Post 0.854 0.425 2.010 21.770 0.004 0.690 0.020 1.687 0.045e. t.test for Dems to midpoint
electOutcome.D <- lm(electPredictTB.plot ~ (pDemR + pDemI) * tDur_Post, data = d)
summary(electOutcome.D)##
## Call:
## lm(formula = electPredictTB.plot ~ (pDemR + pDemI) * tDur_Post,
## data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -6.5798 -1.7282 0.4202 1.4202 5.3684
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.3701 0.1077 22.013 <2e-16 ***
## pDemR -3.7380 0.1570 -23.811 <2e-16 ***
## pDemI -2.0694 0.2089 -9.906 <2e-16 ***
## tDur_Post 0.4194 0.2153 1.948 0.0517 .
## pDemR:tDur_Post -0.4182 0.3140 -1.332 0.1831
## pDemI:tDur_Post 0.4355 0.4178 1.042 0.2975
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 2.321 on 1078 degrees of freedom
## (152 observations deleted due to missingness)
## Multiple R-squared: 0.3628, Adjusted R-squared: 0.3598
## F-statistic: 122.7 on 5 and 1078 DF, p-value: < 2.2e-16## lm(formula = electPredictTB.plot ~ (pDemR + pDemI) * tDur_Post,
## data = d)
##
## Omnibus ANOVA
## SS df MS EtaSq F p
## Model 3307.008 5 661.402 0.363 122.74 0
## Error 5808.958 1078 5.389
## Corr Total 9115.966 1083 8.417
##
## RMSE AdjEtaSq
## 2.321 0.36
##
## Coefficients
## Est StErr t SSR(3) EtaSq tol CI_2.5 CI_97.5 p
## (Intercept) 2.370 0.108 22.013 2611.273 0.310 NA 2.159 2.581 0.000
## pDemR -3.738 0.157 -23.811 3055.249 0.345 0.853 -4.046 -3.430 0.000
## pDemI -2.069 0.209 -9.906 528.733 0.083 0.845 -2.479 -1.659 0.000
## tDur_Post 0.419 0.215 1.948 20.439 0.004 0.442 -0.003 0.842 0.052
## pDemR:tDur_Post -0.418 0.314 -1.332 9.561 0.002 0.527 -1.034 0.198 0.183
## pDemI:tDur_Post 0.436 0.418 1.042 5.855 0.001 0.713 -0.384 1.255 0.297f. t.test for Inds to midpoint
electOutcome.I <- lm(electPredictTB.plot ~ (pIndD + pIndR) * tDur_Post, data = d)
summary(electOutcome.I)##
## Call:
## lm(formula = electPredictTB.plot ~ (pIndD + pIndR) * tDur_Post,
## data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -6.5798 -1.7282 0.4202 1.4202 5.3684
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.3007 0.1790 1.680 0.0933 .
## pIndD 2.0694 0.2089 9.906 < 2e-16 ***
## pIndR -1.6685 0.2124 -7.856 9.52e-15 ***
## tDur_Post 0.8549 0.3581 2.388 0.0171 *
## pIndD:tDur_Post -0.4355 0.4178 -1.042 0.2975
## pIndR:tDur_Post -0.8538 0.4248 -2.010 0.0447 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 2.321 on 1078 degrees of freedom
## (152 observations deleted due to missingness)
## Multiple R-squared: 0.3628, Adjusted R-squared: 0.3598
## F-statistic: 122.7 on 5 and 1078 DF, p-value: < 2.2e-16## lm(formula = electPredictTB.plot ~ (pIndD + pIndR) * tDur_Post,
## data = d)
##
## Omnibus ANOVA
## SS df MS EtaSq F p
## Model 3307.008 5 661.402 0.363 122.74 0
## Error 5808.958 1078 5.389
## Corr Total 9115.966 1083 8.417
##
## RMSE AdjEtaSq
## 2.321 0.36
##
## Coefficients
## Est StErr t SSR(3) EtaSq tol CI_2.5 CI_97.5 p
## (Intercept) 0.301 0.179 1.680 15.201 0.003 NA -0.051 0.652 0.093
## pIndD 2.069 0.209 9.906 528.733 0.083 0.459 1.659 2.479 0.000
## pIndR -1.669 0.212 -7.856 332.606 0.054 0.466 -2.085 -1.252 0.000
## tDur_Post 0.855 0.358 2.388 30.718 0.005 0.160 0.152 1.558 0.017
## pIndD:tDur_Post -0.436 0.418 -1.042 5.855 0.001 0.257 -1.255 0.384 0.297
## pIndR:tDur_Post -0.854 0.425 -2.010 21.770 0.004 0.288 -1.687 -0.020 0.045g. Does ID strength moderate?
2. Perceived Election Legitimacy
a. Descriptives
##
## Descriptive statistics by group
## group: Democrat
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 557 4.17 0.98 4.5 4.32 0.74 1 5 4 -1.15 0.67 0.04
## ------------------------------------------------------------
## group: Republican
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 463 2.89 1.21 3 2.87 1.48 1 5 4 0.07 -0.94 0.06
## ------------------------------------------------------------
## group: Independent
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 187 2.97 1.37 3 2.96 1.48 1 5 4 -0.06 -1.22 0.1##
## Descriptive statistics by group
## : Democrat
## : During-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 250 3.92 1 4 4.04 1.48 1 5 4 -0.77 -0.02 0.06
## ------------------------------------------------------------
## : Republican
## : During-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 237 2.99 1.17 3 2.99 1.48 1 5 4 0.04 -0.93 0.08
## ------------------------------------------------------------
## : Independent
## : During-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 86 2.96 1.28 3 2.95 1.48 1 5 4 -0.15 -1.12 0.14
## ------------------------------------------------------------
## : Democrat
## : Post-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 307 4.37 0.92 5 4.55 0 1 5 4 -1.61 2.07 0.05
## ------------------------------------------------------------
## : Republican
## : Post-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 226 2.79 1.24 3 2.74 1.48 1 5 4 0.12 -0.97 0.08
## ------------------------------------------------------------
## : Independent
## : Post-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 101 2.98 1.45 3 2.97 2.22 1 5 4 -0.01 -1.34 0.14b. ANOVA: Vote Confidence = Partisanship * Timing * Vote Type
i. Prep: Creating the long dataset to use the repeated measures anova command
voteconf <- as.data.frame(cbind(d[d$election_timing != "Pre-election",]$party_factor,
d[d$election_timing != "Pre-election",]$ownvote_conf,
d[d$election_timing != "Pre-election",]$overallvote_conf,
d[d$election_timing != "Pre-election",]$s3,
d[d$election_timing != "Pre-election",]$election_timing,
d[d$election_timing != "Pre-election",]$IDstrength))
names(voteconf) <- c("party_factor", "Own","Nationwide", "s3", "election_timing", "IDstrength")
voteconf$party_factor <- recode_factor(voteconf$party_factor, `1` = "Democrat", `2` = "Republican", `3` = "Independent")
d2 <- tidyr::gather(voteconf, Vote_Type, Confidence, Own:Nationwide, factor_key=TRUE)
d3 <- d2[!is.na(d2$party_factor),]
d3 <- na.omit(d3)
d4 <- d3[d3$s3 != 29149117, ]
d4$s3 <- as.factor(as.character(d4$s3))ii. Model
### Main model w/o repeated factors: Contrast coded predictors
vote.party.time <- lm(voteLegit ~ (pDem_Rep + pInd_Not) * tDur_Post, data = d)
### Main ANOVA (& effect size)
main.aov <- ezANOVA(data = d4
, dv = Confidence
, wid = s3
, within = Vote_Type
, between = .(party_factor, election_timing)
, type = 3
, return_aov = T)## Warning: You have removed one or more levels from variable "party_factor".
## Refactoring for ANOVA.## Warning: "election_timing" will be treated as numeric.## Warning: Data is unbalanced (unequal N per group). Make sure you specified a
## well-considered value for the type argument to ezANOVA().Yes–There is a 3-way interaction between vote type, partisanship (Dem vs. Rep) and election timing. There is NO interaction of timing with Independents vs. Dems/Reps. In the context of the full model, there is also no main effect of timing.
c. Simple Effects of Timing, at each level of Party
Parameter of interest: tDur_Post in each model
# Democrats
vote.party.time.D <- lm(voteLegit ~ (pDemR + pDemI) * (tDur_Post), data = d)
summary(vote.party.time.D)##
## Call:
## lm(formula = voteLegit ~ (pDemR + pDemI) * (tDur_Post), data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.3664 -0.8664 0.0760 0.6336 2.2146
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 4.14522 0.04803 86.307 < 2e-16 ***
## pDemR -1.25569 0.07109 -17.662 < 2e-16 ***
## pDemI -1.17795 0.09565 -12.315 < 2e-16 ***
## tDur_Post 0.44245 0.09606 4.606 4.54e-06 ***
## pDemR:tDur_Post -0.65072 0.14219 -4.577 5.22e-06 ***
## pDemI:tDur_Post -0.42650 0.19131 -2.229 0.026 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.128 on 1201 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.2483, Adjusted R-squared: 0.2452
## F-statistic: 79.36 on 5 and 1201 DF, p-value: < 2.2e-16## lm(formula = voteLegit ~ (pDemR + pDemI) * (tDur_Post), data = d)
##
## Omnibus ANOVA
## SS df MS EtaSq F p
## Model 504.474 5 100.895 0.248 79.357 0
## Error 1526.958 1201 1.271
## Corr Total 2031.433 1206 1.684
##
## RMSE AdjEtaSq
## 1.128 0.245
##
## Coefficients
## Est StErr t SSR(3) EtaSq tol CI_2.5 CI_97.5 p
## (Intercept) 4.145 0.048 86.307 9470.640 0.861 NA 4.051 4.239 0.000
## pDemR -1.256 0.071 -17.662 396.631 0.206 0.881 -1.395 -1.116 0.000
## pDemI -1.178 0.096 -12.315 192.810 0.112 0.879 -1.366 -0.990 0.000
## tDur_Post 0.442 0.096 4.606 26.974 0.017 0.458 0.254 0.631 0.000
## pDemR:tDur_Post -0.651 0.142 -4.577 26.629 0.017 0.543 -0.930 -0.372 0.000
## pDemI:tDur_Post -0.427 0.191 -2.229 6.319 0.004 0.744 -0.802 -0.051 0.026vote.party.time.D2 <- lm(Own_Nat_conf_diff ~ (pDemR + pDemI) * (tDur_Post), data = d)
summary(vote.party.time.D2) # This model is technically testing whether the difference between Own & National vote confidence is changing over time, not whether confidence is changing in general##
## Call:
## lm(formula = Own_Nat_conf_diff ~ (pDemR + pDemI) * (tDur_Post),
## data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -4.0891 -0.7131 -0.0814 0.0929 3.9109
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.15289 0.03916 3.904 9.99e-05 ***
## pDemR 0.65719 0.05797 11.337 < 2e-16 ***
## pDemI -0.06763 0.07800 -0.867 0.386037
## tDur_Post -0.28623 0.07832 -3.654 0.000269 ***
## pDemR:tDur_Post 0.48023 0.11594 4.142 3.68e-05 ***
## pDemI:tDur_Post 0.29394 0.15599 1.884 0.059762 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.9194 on 1201 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.128, Adjusted R-squared: 0.1244
## F-statistic: 35.27 on 5 and 1201 DF, p-value: < 2.2e-16# Republicans
vote.party.time.R <- lm(voteLegit ~ (pRepD + pRepI) * (tDur_Post), data = d)
summary(vote.party.time.R)##
## Call:
## lm(formula = voteLegit ~ (pRepD + pRepI) * (tDur_Post), data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.3664 -0.8664 0.0760 0.6336 2.2146
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.88953 0.05242 55.126 < 2e-16 ***
## pRepD 1.25569 0.07109 17.662 < 2e-16 ***
## pRepI 0.07774 0.09793 0.794 0.4275
## tDur_Post -0.20827 0.10483 -1.987 0.0472 *
## pRepD:tDur_Post 0.65072 0.14219 4.577 5.22e-06 ***
## pRepI:tDur_Post 0.22422 0.19586 1.145 0.2525
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.128 on 1201 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.2483, Adjusted R-squared: 0.2452
## F-statistic: 79.36 on 5 and 1201 DF, p-value: < 2.2e-16## lm(formula = voteLegit ~ (pRepD + pRepI) * (tDur_Post), data = d)
##
## Omnibus ANOVA
## SS df MS EtaSq F p
## Model 504.474 5 100.895 0.248 79.357 0
## Error 1526.958 1201 1.271
## Corr Total 2031.433 1206 1.684
##
## RMSE AdjEtaSq
## 1.128 0.245
##
## Coefficients
## Est StErr t SSR(3) EtaSq tol CI_2.5 CI_97.5 p
## (Intercept) 2.890 0.052 55.126 3863.595 0.717 NA 2.787 2.992 0.000
## pRepD 1.256 0.071 17.662 396.631 0.206 0.839 1.116 1.395 0.000
## pRepI 0.078 0.098 0.794 0.801 0.001 0.839 -0.114 0.270 0.427
## tDur_Post -0.208 0.105 -1.987 5.018 0.003 0.384 -0.414 -0.003 0.047
## pRepD:tDur_Post 0.651 0.142 4.577 26.629 0.017 0.454 0.372 0.930 0.000
## pRepI:tDur_Post 0.224 0.196 1.145 1.666 0.001 0.710 -0.160 0.608 0.253# Independents
vote.party.time.I <- lm(voteLegit ~ (pIndD + pIndR) * (tDur_Post), data = d)
summary(vote.party.time.I)##
## Call:
## lm(formula = voteLegit ~ (pIndD + pIndR) * (tDur_Post), data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.3664 -0.8664 0.0760 0.6336 2.2146
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.96727 0.08272 35.870 <2e-16 ***
## pIndD 1.17795 0.09565 12.315 <2e-16 ***
## pIndR -0.07774 0.09793 -0.794 0.427
## tDur_Post 0.01595 0.16544 0.096 0.923
## pIndD:tDur_Post 0.42650 0.19131 2.229 0.026 *
## pIndR:tDur_Post -0.22422 0.19586 -1.145 0.253
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.128 on 1201 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.2483, Adjusted R-squared: 0.2452
## F-statistic: 79.36 on 5 and 1201 DF, p-value: < 2.2e-16## lm(formula = voteLegit ~ (pIndD + pIndR) * (tDur_Post), data = d)
##
## Omnibus ANOVA
## SS df MS EtaSq F p
## Model 504.474 5 100.895 0.248 79.357 0
## Error 1526.958 1201 1.271
## Corr Total 2031.433 1206 1.684
##
## RMSE AdjEtaSq
## 1.128 0.245
##
## Coefficients
## Est StErr t SSR(3) EtaSq tol CI_2.5 CI_97.5 p
## (Intercept) 2.967 0.083 35.870 1635.889 0.517 NA 2.805 3.130 0.000
## pIndD 1.178 0.096 12.315 192.810 0.112 0.463 0.990 1.366 0.000
## pIndR -0.078 0.098 -0.794 0.801 0.001 0.465 -0.270 0.114 0.427
## tDur_Post 0.016 0.165 0.096 0.012 0.000 0.154 -0.309 0.341 0.923
## pIndD:tDur_Post 0.427 0.191 2.229 6.319 0.004 0.251 0.051 0.802 0.026
## pIndR:tDur_Post -0.224 0.196 -1.145 1.666 0.001 0.286 -0.608 0.160 0.253i. by just timing effects?
vote.party.time.dur <- lm(Own_Nat_conf_diff ~ (pDem_Rep + pInd_Not) * (tDur), data = d)
summary(vote.party.time.dur)##
## Call:
## lm(formula = Own_Nat_conf_diff ~ (pDem_Rep + pInd_Not) * (tDur),
## data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -4.0891 -0.7131 -0.0814 0.0929 3.9109
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.36490 0.04301 8.483 < 2e-16 ***
## pDem_Rep 0.41708 0.08336 5.004 6.46e-07 ***
## pInd_Not 0.42314 0.10755 3.935 8.81e-05 ***
## tDur -0.02835 0.05908 -0.480 0.631
## pDem_Rep:tDur 0.48023 0.11594 4.142 3.68e-05 ***
## pInd_Not:tDur -0.05383 0.14683 -0.367 0.714
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.9194 on 1201 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.128, Adjusted R-squared: 0.1244
## F-statistic: 35.27 on 5 and 1201 DF, p-value: < 2.2e-16vote.party.time.post <- lm(Own_Nat_conf_diff ~ (pDem_Rep + pInd_Not) * (tPost), data = d)
summary(vote.party.time.post)##
## Call:
## lm(formula = Own_Nat_conf_diff ~ (pDem_Rep + pInd_Not) * (tPost),
## data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -4.0891 -0.7131 -0.0814 0.0929 3.9109
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.33655 0.04050 8.310 2.56e-16 ***
## pDem_Rep 0.89731 0.08058 11.135 < 2e-16 ***
## pInd_Not 0.36932 0.09997 3.694 0.00023 ***
## tPost 0.02835 0.05908 0.480 0.63140
## pDem_Rep:tPost -0.48023 0.11594 -4.142 3.68e-05 ***
## pInd_Not:tPost 0.05383 0.14683 0.367 0.71398
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.9194 on 1201 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.128, Adjusted R-squared: 0.1244
## F-statistic: 35.27 on 5 and 1201 DF, p-value: < 2.2e-16d. Own vs. National Vote Models
i. Descriptives for own & national vote confidence
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 1209 3.68 1.35 4 3.85 1.48 1 5 4 -0.7 -0.73 0.04##
## Descriptive statistics by group
## : Democrat
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 557 4.24 1.01 5 4.4 0 1 5 4 -1.26 0.93 0.04
## ------------------------------------------------------------
## : Republican
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 463 3.3 1.39 3 3.37 1.48 1 5 4 -0.3 -1.17 0.06
## ------------------------------------------------------------
## : Independent
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 188 3 1.49 3 3 1.48 1 5 4 -0.06 -1.38 0.11##
## Descriptive statistics by group
## : Democrat
## : During-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 250 4.07 1.03 4 4.21 1.48 1 5 4 -0.93 0.21 0.07
## ------------------------------------------------------------
## : Republican
## : During-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 237 3.35 1.34 3 3.43 1.48 1 5 4 -0.29 -1.12 0.09
## ------------------------------------------------------------
## : Independent
## : During-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 86 3 1.36 3 3 1.48 1 5 4 -0.11 -1.18 0.15
## ------------------------------------------------------------
## : Democrat
## : Post-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 307 4.37 0.97 5 4.56 0 1 5 4 -1.6 1.97 0.06
## ------------------------------------------------------------
## : Republican
## : Post-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 226 3.24 1.44 3 3.3 1.48 1 5 4 -0.29 -1.26 0.1
## ------------------------------------------------------------
## : Independent
## : Post-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 102 3 1.59 3 3 2.97 1 5 4 -0.03 -1.55 0.16## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 1208 3.3 1.43 3 3.37 1.48 1 5 4 -0.28 -1.25 0.04##
## Descriptive statistics by group
## : Democrat
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 557 4.1 1.06 4 4.27 1.48 1 5 4 -1.05 0.33 0.04
## ------------------------------------------------------------
## : Republican
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 463 2.49 1.3 2 2.37 1.48 1 5 4 0.49 -0.91 0.06
## ------------------------------------------------------------
## : Independent
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 187 2.93 1.41 3 2.91 1.48 1 5 4 -0.01 -1.24 0.1##
## Descriptive statistics by group
## : Democrat
## : During-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 250 3.78 1.1 4 3.88 1.48 1 5 4 -0.6 -0.46 0.07
## ------------------------------------------------------------
## : Republican
## : During-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 237 2.64 1.25 2 2.56 1.48 1 5 4 0.31 -0.94 0.08
## ------------------------------------------------------------
## : Independent
## : During-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 86 2.92 1.29 3 2.9 1.48 1 5 4 -0.05 -1 0.14
## ------------------------------------------------------------
## : Democrat
## : Post-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 307 4.36 0.94 5 4.54 0 1 5 4 -1.58 2.12 0.05
## ------------------------------------------------------------
## : Republican
## : Post-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 226 2.33 1.33 2 2.17 1.48 1 5 4 0.69 -0.76 0.09
## ------------------------------------------------------------
## : Independent
## : Post-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 101 2.93 1.51 3 2.91 1.48 1 5 4 0.01 -1.44 0.15ii. Main Effect of Vote Type: Test of intercept in this model, or main effect from ezANOVA command above
vote.party.time <- lm(Own_Nat_conf_diff ~ (pDem_Rep + pInd_Not) * tDur_Post, data = d)
summary(vote.party.time)##
## Call:
## lm(formula = Own_Nat_conf_diff ~ (pDem_Rep + pInd_Not) * tDur_Post,
## data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -4.0891 -0.7131 -0.0814 0.0929 3.9109
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.35073 0.02954 11.873 < 2e-16 ***
## pDem_Rep 0.65719 0.05797 11.337 < 2e-16 ***
## pInd_Not 0.39623 0.07342 5.397 8.15e-08 ***
## tDur_Post -0.02835 0.05908 -0.480 0.631
## pDem_Rep:tDur_Post 0.48023 0.11594 4.142 3.68e-05 ***
## pInd_Not:tDur_Post -0.05383 0.14683 -0.367 0.714
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.9194 on 1201 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.128, Adjusted R-squared: 0.1244
## F-statistic: 35.27 on 5 and 1201 DF, p-value: < 2.2e-16## $ANOVA
## Effect DFn DFd F p p<.05
## 2 party_factor 1 1004 1.4537950 2.282041e-01
## 3 election_timing 1 1004 2.7955868 9.483574e-02
## 5 Vote_Type 1 1004 35.1044622 4.289830e-09 *
## 4 party_factor:election_timing 1 1004 23.5400573 1.417505e-06 *
## 6 party_factor:Vote_Type 1 1004 0.0974653 7.549577e-01
## 7 election_timing:Vote_Type 1 1004 0.6936821 4.051129e-01
## 8 party_factor:election_timing:Vote_Type 1 1004 16.7795336 4.536486e-05 *
## ges
## 2 1.221155e-03
## 3 2.345586e-03
## 5 5.412152e-03
## 4 1.941296e-02
## 6 1.510804e-05
## 7 1.075173e-04
## 8 2.594276e-03
##
## $aov
##
## Call:
## aov(formula = formula(aov_formula), data = data)
##
## Grand Mean: 3.587302
##
## Stratum 1: s3
##
## Terms:
## party_factor election_timing party_factor:election_timing
## Sum of Squares 814.9752 9.7773 54.4294
## Deg. of Freedom 1 1 1
## Residuals
## Sum of Squares 2321.4529
## Deg. of Freedom 1004
##
## Residual standard error: 1.520593
## 3 out of 6 effects not estimable
## Estimated effects may be unbalanced
##
## Stratum 2: s3:Vote_Type
##
## Terms:
## Vote_Type party_factor:Vote_Type election_timing:Vote_Type
## Sum of Squares 99.5556 56.8400 0.5691
## Deg. of Freedom 1 1 1
## party_factor:election_timing:Vote_Type Residuals
## Sum of Squares 7.1511 427.8842
## Deg. of Freedom 1 1004
##
## Residual standard error: 0.6528242
## Estimated effects may be unbalancediii. National Vote Confidence ~ Partisanship
##
## Call:
## lm(formula = overallvote_conf ~ (pDem_Rep + pInd_Not), data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.09874 -1.09874 -0.09874 0.90126 2.51188
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.17189 0.03884 81.658 < 2e-16 ***
## pDem_Rep -1.61062 0.07626 -21.121 < 2e-16 ***
## pInd_Not 0.36830 0.09652 3.816 0.000143 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.213 on 1204 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.2795, Adjusted R-squared: 0.2783
## F-statistic: 233.6 on 2 and 1204 DF, p-value: < 2.2e-161. Confidence in National Vote ~ Partisanship x Timing
# ANOVA cmd approach
natvote.party.time <- lm(overallvote_conf ~ (pDem_Rep + pInd_Not) * (tDur_Post), data = d)
natvote.party.time.c <- lm(overallvote_conf ~ tDur_Post + pDem_Rep + pInd_Not, data = d)
# Effect size calcs
aov(natvote.party.time)## Call:
## aov(formula = natvote.party.time)
##
## Terms:
## pDem_Rep pInd_Not tDur_Post pDem_Rep:tDur_Post
## Sum of Squares 665.4162 21.4071 7.0342 50.3570
## Deg. of Freedom 1 1 1 1
## pInd_Not:tDur_Post Residuals
## Sum of Squares 0.6430 1712.1714
## Deg. of Freedom 1 1201
##
## Residual standard error: 1.193994
## Estimated effects may be unbalanced
## 29 observations deleted due to missingness## lm(formula = overallvote_conf ~ (pDem_Rep + pInd_Not) * (tDur_Post),
## data = d)
##
## Omnibus ANOVA
## SS df MS EtaSq F p
## Model 744.858 5 148.972 0.303 104.496 0
## Error 1712.171 1201 1.426
## Corr Total 2457.029 1206 2.037
##
## RMSE AdjEtaSq
## 1.194 0.3
##
## Coefficients
## Est StErr t SSR(3) EtaSq tol CI_2.5 CI_97.5
## (Intercept) 3.160 0.038 82.387 9676.470 0.850 NA 3.085 3.236
## pDem_Rep -1.584 0.075 -21.045 631.378 0.269 0.994 -1.732 -1.437
## pInd_Not 0.352 0.095 3.692 19.431 0.011 0.992 0.165 0.539
## tDur_Post 0.098 0.077 1.276 2.321 0.001 0.805 -0.053 0.248
## pDem_Rep:tDur_Post -0.891 0.151 -5.917 49.906 0.028 0.990 -1.186 -0.595
## pInd_Not:tDur_Post 0.128 0.191 0.672 0.643 0.000 0.804 -0.246 0.502
## p
## (Intercept) 0.000
## pDem_Rep 0.000
## pInd_Not 0.000
## tDur_Post 0.202
## pDem_Rep:tDur_Post 0.000
## pInd_Not:tDur_Post 0.5022. Confidence in National Vote ~ Partisanship x Timing - by parties
#Dems
natvote.party.time.D <- lm(overallvote_conf ~ (pDemR + pDemI) * (tDur_Post), data = d)
summary(natvote.party.time.D)##
## Call:
## lm(formula = overallvote_conf ~ (pDemR + pDemI) * (tDur_Post),
## data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.3616 -0.7760 0.2240 0.6384 2.6681
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 4.06878 0.05086 80.003 < 2e-16 ***
## pDemR -1.58429 0.07528 -21.045 < 2e-16 ***
## pDemI -1.14413 0.10129 -11.296 < 2e-16 ***
## tDur_Post 0.58556 0.10172 5.757 1.09e-08 ***
## pDemR:tDur_Post -0.89084 0.15056 -5.917 4.28e-09 ***
## pDemI:tDur_Post -0.57348 0.20258 -2.831 0.00472 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.194 on 1201 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.3032, Adjusted R-squared: 0.3003
## F-statistic: 104.5 on 5 and 1201 DF, p-value: < 2.2e-16## lm(formula = overallvote_conf ~ (pDemR + pDemI) * (tDur_Post),
## data = d)
##
## Omnibus ANOVA
## SS df MS EtaSq F p
## Model 744.858 5 148.972 0.303 104.496 0
## Error 1712.171 1201 1.426
## Corr Total 2457.029 1206 2.037
##
## RMSE AdjEtaSq
## 1.194 0.3
##
## Coefficients
## Est StErr t SSR(3) EtaSq tol CI_2.5 CI_97.5 p
## (Intercept) 4.069 0.051 80.003 9124.561 0.842 NA 3.969 4.169 0.000
## pDemR -1.584 0.075 -21.045 631.378 0.269 0.881 -1.732 -1.437 0.000
## pDemI -1.144 0.101 -11.296 181.898 0.096 0.879 -1.343 -0.945 0.000
## tDur_Post 0.586 0.102 5.757 47.247 0.027 0.458 0.386 0.785 0.000
## pDemR:tDur_Post -0.891 0.151 -5.917 49.906 0.028 0.543 -1.186 -0.595 0.000
## pDemI:tDur_Post -0.573 0.203 -2.831 11.425 0.007 0.744 -0.971 -0.176 0.005# Reps
natvote.party.time.R <- lm(overallvote_conf ~ (pRepD + pRepI) * (tDur_Post), data = d)
summary(natvote.party.time.R)##
## Call:
## lm(formula = overallvote_conf ~ (pRepD + pRepI) * (tDur_Post),
## data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.3616 -0.7760 0.2240 0.6384 2.6681
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.48449 0.05551 44.761 < 2e-16 ***
## pRepD 1.58429 0.07528 21.045 < 2e-16 ***
## pRepI 0.44015 0.10370 4.244 2.36e-05 ***
## tDur_Post -0.30527 0.11101 -2.750 0.00605 **
## pRepD:tDur_Post 0.89084 0.15056 5.917 4.28e-09 ***
## pRepI:tDur_Post 0.31736 0.20740 1.530 0.12624
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.194 on 1201 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.3032, Adjusted R-squared: 0.3003
## F-statistic: 104.5 on 5 and 1201 DF, p-value: < 2.2e-16## lm(formula = overallvote_conf ~ (pRepD + pRepI) * (tDur_Post),
## data = d)
##
## Omnibus ANOVA
## SS df MS EtaSq F p
## Model 744.858 5 148.972 0.303 104.496 0
## Error 1712.171 1201 1.426
## Corr Total 2457.029 1206 2.037
##
## RMSE AdjEtaSq
## 1.194 0.3
##
## Coefficients
## Est StErr t SSR(3) EtaSq tol CI_2.5 CI_97.5 p
## (Intercept) 2.484 0.056 44.761 2856.353 0.625 NA 2.376 2.593 0.000
## pRepD 1.584 0.075 21.045 631.378 0.269 0.839 1.437 1.732 0.000
## pRepI 0.440 0.104 4.244 25.683 0.015 0.839 0.237 0.644 0.000
## tDur_Post -0.305 0.111 -2.750 10.781 0.006 0.384 -0.523 -0.087 0.006
## pRepD:tDur_Post 0.891 0.151 5.917 49.906 0.028 0.454 0.595 1.186 0.000
## pRepI:tDur_Post 0.317 0.207 1.530 3.338 0.002 0.710 -0.090 0.724 0.126# Inds
natvote.party.time.I <- lm(overallvote_conf ~ (pIndD + pIndR) * (tDur_Post), data = d)
summary(natvote.party.time.I)##
## Call:
## lm(formula = overallvote_conf ~ (pIndD + pIndR) * (tDur_Post),
## data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.3616 -0.7760 0.2240 0.6384 2.6681
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.92465 0.08760 33.388 < 2e-16 ***
## pIndD 1.14413 0.10129 11.296 < 2e-16 ***
## pIndR -0.44015 0.10370 -4.244 2.36e-05 ***
## tDur_Post 0.01209 0.17519 0.069 0.94500
## pIndD:tDur_Post 0.57348 0.20258 2.831 0.00472 **
## pIndR:tDur_Post -0.31736 0.20740 -1.530 0.12624
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.194 on 1201 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.3032, Adjusted R-squared: 0.3003
## F-statistic: 104.5 on 5 and 1201 DF, p-value: < 2.2e-16iv. Own Vote Confidence ~ Partisanship
##
## Call:
## lm(formula = ownvote_conf ~ (pDem_Rep + pInd_Not), data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.2370 -1.2370 0.7041 0.7630 2.0000
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.51351 0.03982 88.243 < 2e-16 ***
## pDem_Rep -0.94109 0.07829 -12.021 < 2e-16 ***
## pInd_Not 0.76644 0.09887 7.752 1.91e-14 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.245 on 1205 degrees of freedom
## (28 observations deleted due to missingness)
## Multiple R-squared: 0.1494, Adjusted R-squared: 0.148
## F-statistic: 105.8 on 2 and 1205 DF, p-value: < 2.2e-161. Confidence in Own Vote ~ Partisanship x Timing
ownvote.party.time <- lm(ownvote_conf ~ (pDem_Rep + pInd_Not) * (tDur_Post), data = d)
summary(ownvote.party.time)##
## Call:
## lm(formula = ownvote_conf ~ (pDem_Rep + pInd_Not) * (tDur_Post),
## data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.3713 -1.0720 0.6287 0.9280 2.0000
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.50794 0.03984 88.040 < 2e-16 ***
## pDem_Rep -0.92709 0.07829 -11.841 < 2e-16 ***
## pInd_Not 0.75812 0.09896 7.661 3.79e-14 ***
## tDur_Post 0.06300 0.07969 0.791 0.42934
## pDem_Rep:tDur_Post -0.41061 0.15659 -2.622 0.00885 **
## pInd_Not:tDur_Post 0.09403 0.19793 0.475 0.63482
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.242 on 1202 degrees of freedom
## (28 observations deleted due to missingness)
## Multiple R-squared: 0.1557, Adjusted R-squared: 0.1522
## F-statistic: 44.33 on 5 and 1202 DF, p-value: < 2.2e-16## Call:
## aov(formula = ownvote.party.time)
##
## Terms:
## pDem_Rep pInd_Not tDur_Post pDem_Rep:tDur_Post
## Sum of Squares 234.8961 93.1256 2.6823 10.7485
## Deg. of Freedom 1 1 1 1
## pInd_Not:tDur_Post Residuals
## Sum of Squares 0.3480 1853.4016
## Deg. of Freedom 1 1202
##
## Residual standard error: 1.241745
## Estimated effects may be unbalanced
## 28 observations deleted due to missingness## lm(formula = ownvote_conf ~ (pDem_Rep + pInd_Not) * (tDur_Post),
## data = d)
##
## Omnibus ANOVA
## SS df MS EtaSq F p
## Model 341.800 5 68.360 0.156 44.334 0
## Error 1853.402 1202 1.542
## Corr Total 2195.202 1207 1.819
##
## RMSE AdjEtaSq
## 1.242 0.152
##
## Coefficients
## Est StErr t SSR(3) EtaSq tol CI_2.5 CI_97.5
## (Intercept) 3.508 0.040 88.040 11951.661 0.866 NA 3.430 3.586
## pDem_Rep -0.927 0.078 -11.841 216.206 0.104 0.994 -1.081 -0.773
## pInd_Not 0.758 0.099 7.661 90.486 0.047 0.992 0.564 0.952
## tDur_Post 0.063 0.080 0.791 0.964 0.001 0.806 -0.093 0.219
## pDem_Rep:tDur_Post -0.411 0.157 -2.622 10.603 0.006 0.990 -0.718 -0.103
## pInd_Not:tDur_Post 0.094 0.198 0.475 0.348 0.000 0.805 -0.294 0.482
## p
## (Intercept) 0.000
## pDem_Rep 0.000
## pInd_Not 0.000
## tDur_Post 0.429
## pDem_Rep:tDur_Post 0.009
## pInd_Not:tDur_Post 0.6352. Confidence in Own Vote ~ Partisanship x Timing - by parties
#Dems
ownvote.party.time.D <- lm(ownvote_conf ~ (pDemR + pDemI) * (tDur_Post), data = d)
mcSummary(ownvote.party.time.D)## lm(formula = ownvote_conf ~ (pDemR + pDemI) * (tDur_Post), data = d)
##
## Omnibus ANOVA
## SS df MS EtaSq F p
## Model 341.800 5 68.360 0.156 44.334 0
## Error 1853.402 1202 1.542
## Corr Total 2195.202 1207 1.819
##
## RMSE AdjEtaSq
## 1.242 0.152
##
## Coefficients
## Est StErr t SSR(3) EtaSq tol CI_2.5 CI_97.5 p
## (Intercept) 4.222 0.053 79.817 9823.161 0.841 NA 4.118 4.325 0.000
## pDemR -0.927 0.078 -11.841 216.206 0.104 0.881 -1.081 -0.773 0.000
## pDemI -1.222 0.105 -11.617 208.089 0.101 0.878 -1.428 -1.015 0.000
## tDur_Post 0.299 0.106 2.830 12.346 0.007 0.457 0.092 0.507 0.005
## pDemR:tDur_Post -0.411 0.157 -2.622 10.603 0.006 0.543 -0.718 -0.103 0.009
## pDemI:tDur_Post -0.299 0.210 -1.423 3.123 0.002 0.742 -0.712 0.113 0.155# Reps
ownvote.party.time.R <- lm(ownvote_conf ~ (pRepD + pRepI) * (tDur_Post), data = d)
summary(ownvote.party.time.R)##
## Call:
## lm(formula = ownvote_conf ~ (pRepD + pRepI) * (tDur_Post), data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.3713 -1.0720 0.6287 0.9280 2.0000
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.29457 0.05773 57.074 < 2e-16 ***
## pRepD 0.92709 0.07829 11.841 < 2e-16 ***
## pRepI -0.29457 0.10767 -2.736 0.00631 **
## tDur_Post -0.11127 0.11545 -0.964 0.33533
## pRepD:tDur_Post 0.41061 0.15659 2.622 0.00885 **
## pRepI:tDur_Post 0.11127 0.21535 0.517 0.60545
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.242 on 1202 degrees of freedom
## (28 observations deleted due to missingness)
## Multiple R-squared: 0.1557, Adjusted R-squared: 0.1522
## F-statistic: 44.33 on 5 and 1202 DF, p-value: < 2.2e-16## lm(formula = ownvote_conf ~ (pRepD + pRepI) * (tDur_Post), data = d)
##
## Omnibus ANOVA
## SS df MS EtaSq F p
## Model 341.800 5 68.360 0.156 44.334 0
## Error 1853.402 1202 1.542
## Corr Total 2195.202 1207 1.819
##
## RMSE AdjEtaSq
## 1.242 0.152
##
## Coefficients
## Est StErr t SSR(3) EtaSq tol CI_2.5 CI_97.5 p
## (Intercept) 3.295 0.058 57.074 5022.668 0.730 NA 3.181 3.408 0.000
## pRepD 0.927 0.078 11.841 216.206 0.104 0.838 0.773 1.081 0.000
## pRepI -0.295 0.108 -2.736 11.541 0.006 0.838 -0.506 -0.083 0.006
## tDur_Post -0.111 0.115 -0.964 1.432 0.001 0.384 -0.338 0.115 0.335
## pRepD:tDur_Post 0.411 0.157 2.622 10.603 0.006 0.454 0.103 0.718 0.009
## pRepI:tDur_Post 0.111 0.215 0.517 0.412 0.000 0.708 -0.311 0.534 0.605# Inds
ownvote.party.time.I <- lm(ownvote_conf ~ (pIndD + pIndR) * (tDur_Post), data = d)
summary(ownvote.party.time.I)##
## Call:
## lm(formula = ownvote_conf ~ (pIndD + pIndR) * (tDur_Post), data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.3713 -1.0720 0.6287 0.9280 2.0000
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.000e+00 9.089e-02 33.006 < 2e-16 ***
## pIndD 1.222e+00 1.052e-01 11.617 < 2e-16 ***
## pIndR 2.946e-01 1.077e-01 2.736 0.00631 **
## tDur_Post 2.563e-16 1.818e-01 0.000 1.00000
## pIndD:tDur_Post 2.993e-01 2.103e-01 1.423 0.15494
## pIndR:tDur_Post -1.113e-01 2.153e-01 -0.517 0.60545
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.242 on 1202 degrees of freedom
## (28 observations deleted due to missingness)
## Multiple R-squared: 0.1557, Adjusted R-squared: 0.1522
## F-statistic: 44.33 on 5 and 1202 DF, p-value: < 2.2e-16v. full model by vote type
##
## Call:
## lm(formula = ownvote_conf ~ (pDem_Rep + pInd_Not) * (tDur_Post),
## data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.3713 -1.0720 0.6287 0.9280 2.0000
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.50794 0.03984 88.040 < 2e-16 ***
## pDem_Rep -0.92709 0.07829 -11.841 < 2e-16 ***
## pInd_Not 0.75812 0.09896 7.661 3.79e-14 ***
## tDur_Post 0.06300 0.07969 0.791 0.42934
## pDem_Rep:tDur_Post -0.41061 0.15659 -2.622 0.00885 **
## pInd_Not:tDur_Post 0.09403 0.19793 0.475 0.63482
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.242 on 1202 degrees of freedom
## (28 observations deleted due to missingness)
## Multiple R-squared: 0.1557, Adjusted R-squared: 0.1522
## F-statistic: 44.33 on 5 and 1202 DF, p-value: < 2.2e-16##
## Call:
## lm(formula = overallvote_conf ~ (pDem_Rep + pInd_Not) * (tDur_Post),
## data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.3616 -0.7760 0.2240 0.6384 2.6681
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.16048 0.03836 82.387 < 2e-16 ***
## pDem_Rep -1.58429 0.07528 -21.045 < 2e-16 ***
## pInd_Not 0.35199 0.09534 3.692 0.000233 ***
## tDur_Post 0.09789 0.07672 1.276 0.202258
## pDem_Rep:tDur_Post -0.89084 0.15056 -5.917 4.28e-09 ***
## pInd_Not:tDur_Post 0.12806 0.19068 0.672 0.501983
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.194 on 1201 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.3032, Adjusted R-squared: 0.3003
## F-statistic: 104.5 on 5 and 1201 DF, p-value: < 2.2e-16g. Moderators/Mediators
i. Does ID strength moderate?
ID strength only for partisans; Independents not included in measure.
ii. Does pol. ideology moderate?
# Own Vote
ownvote.party.time.polid.mod <- lm(ownvote_conf ~ (pDem_Rep) * (tDur_Post) * polIdeology, data = d)
# National Vote
natvote.party.time.polid.mod <- lm(overallvote_conf ~ (pDem_Rep) * (tDur_Post) * polIdeology, data = d)
# All Vote
party.time.polid.mod <- lm(voteLegit ~ (pDem_Rep) * (tDur_Post) * polIdeology, data = d)iii. Emotions
### Vote Legit
# Moderation
vote.emo.mod <- lm(voteLegit ~ negative * positive * (pDem_Rep + pInd_Not) * tDur_Post, data = d)
## Mediation
# Step 1
summary(vote.party.time)##
## Call:
## lm(formula = Own_Nat_conf_diff ~ (pDem_Rep + pInd_Not) * tDur_Post,
## data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -4.0891 -0.7131 -0.0814 0.0929 3.9109
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.35073 0.02954 11.873 < 2e-16 ***
## pDem_Rep 0.65719 0.05797 11.337 < 2e-16 ***
## pInd_Not 0.39623 0.07342 5.397 8.15e-08 ***
## tDur_Post -0.02835 0.05908 -0.480 0.631
## pDem_Rep:tDur_Post 0.48023 0.11594 4.142 3.68e-05 ***
## pInd_Not:tDur_Post -0.05383 0.14683 -0.367 0.714
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.9194 on 1201 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.128, Adjusted R-squared: 0.1244
## F-statistic: 35.27 on 5 and 1201 DF, p-value: < 2.2e-16# Step 2
d$emotions <- d$negative + d$positive # lol
vote.emo.med2 <- lm(emotions ~ (pDem_Rep + pInd_Not) * tDur_Post, data = d)
# Step 3
vote.emo.med3 <- lm(voteLegit ~ negative + positive + (pDem_Rep + pInd_Not) * tDur_Post, data = d)
summary(vote.emo.med3)##
## Call:
## lm(formula = voteLegit ~ negative + positive + (pDem_Rep + pInd_Not) *
## tDur_Post, data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -2.94946 -0.74808 0.09349 0.71634 2.78245
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.091098 0.093732 32.978 < 2e-16 ***
## negative -0.115729 0.020891 -5.540 3.72e-08 ***
## positive 0.190195 0.018571 10.241 < 2e-16 ***
## pDem_Rep -1.052662 0.069706 -15.101 < 2e-16 ***
## pInd_Not 0.399722 0.088112 4.537 6.29e-06 ***
## tDur_Post 0.023641 0.068505 0.345 0.730
## pDem_Rep:tDur_Post -0.191226 0.139199 -1.374 0.170
## pInd_Not:tDur_Post -0.007015 0.170065 -0.041 0.967
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.063 on 1199 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.3326, Adjusted R-squared: 0.3287
## F-statistic: 85.38 on 7 and 1199 DF, p-value: < 2.2e-16##
## Call:
## lm(formula = Own_Nat_conf_diff ~ (pDem_Rep + pInd_Not) * tDur_Post,
## data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -4.0891 -0.7131 -0.0814 0.0929 3.9109
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.35073 0.02954 11.873 < 2e-16 ***
## pDem_Rep 0.65719 0.05797 11.337 < 2e-16 ***
## pInd_Not 0.39623 0.07342 5.397 8.15e-08 ***
## tDur_Post -0.02835 0.05908 -0.480 0.631
## pDem_Rep:tDur_Post 0.48023 0.11594 4.142 3.68e-05 ***
## pInd_Not:tDur_Post -0.05383 0.14683 -0.367 0.714
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.9194 on 1201 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.128, Adjusted R-squared: 0.1244
## F-statistic: 35.27 on 5 and 1201 DF, p-value: < 2.2e-16### Ownvote
# Moderation
ownvote.emo.mod <- lm(ownvote_conf ~ negative * positive * (pDem_Rep + pInd_Not) * tDur_Post, data = d)
# Mediation
ownvote.emo.med <- lm(ownvote_conf ~ negative + positive + (pDem_Rep + pInd_Not) * tDur_Post, data = d)
### National Vote
# Moderation
natvote.emo.mod <- lm(overallvote_conf ~ negative * positive * (pDem_Rep + pInd_Not) * tDur_Post, data = d)
# Mediation
natvote.emo.med <- lm(overallvote_conf ~ negative + positive + (pDem_Rep + pInd_Not) * tDur_Post, data = d)
### And with media measures?
vote.emo.media.med <- lm(voteLegit ~ negative + positive + (pDem_Rep + pInd_Not) * tDur_Post + foxPerception + otherMediaPerception, data = d)
summary(vote.emo.media.med)##
## Call:
## lm(formula = voteLegit ~ negative + positive + (pDem_Rep + pInd_Not) *
## tDur_Post + foxPerception + otherMediaPerception, data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -2.7310 -0.7054 0.0865 0.7016 3.3803
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.60552 0.12290 21.201 < 2e-16 ***
## negative -0.12818 0.02009 -6.381 2.51e-10 ***
## positive 0.13554 0.01911 7.093 2.23e-12 ***
## pDem_Rep -0.59642 0.08071 -7.389 2.76e-13 ***
## pInd_Not 0.47810 0.08469 5.645 2.06e-08 ***
## tDur_Post 0.03424 0.06566 0.521 0.6021
## foxPerception -0.13812 0.02612 -5.288 1.47e-07 ***
## otherMediaPerception 0.41959 0.04237 9.903 < 2e-16 ***
## pDem_Rep:tDur_Post -0.28492 0.13358 -2.133 0.0331 *
## pInd_Not:tDur_Post -0.06742 0.16292 -0.414 0.6791
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.018 on 1197 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.3894, Adjusted R-squared: 0.3848
## F-statistic: 84.83 on 9 and 1197 DF, p-value: < 2.2e-16### Moderated Mediation
vote.emo.mod.med <- lm(voteLegit ~ negative.c + positive.c + (negative.c + positive.c):tDur_Post + (pDem_Rep + pInd_Not)*tDur_Post, data = d)
summary(vote.emo.mod.med)##
## Call:
## lm(formula = voteLegit ~ negative.c + positive.c + (negative.c +
## positive.c):tDur_Post + (pDem_Rep + pInd_Not) * tDur_Post,
## data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -2.85419 -0.77355 0.07655 0.75018 2.87436
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.37407 0.03468 97.299 < 2e-16 ***
## negative.c -0.11480 0.02090 -5.493 4.82e-08 ***
## positive.c 0.18710 0.01871 10.000 < 2e-16 ***
## pDem_Rep -1.02821 0.07164 -14.353 < 2e-16 ***
## pInd_Not 0.39625 0.08814 4.496 7.60e-06 ***
## tDur_Post 0.03390 0.06935 0.489 0.625
## negative.c:tDur_Post -0.03237 0.04180 -0.774 0.439
## positive.c:tDur_Post 0.04363 0.03742 1.166 0.244
## tDur_Post:pDem_Rep -0.15091 0.14328 -1.053 0.292
## tDur_Post:pInd_Not -0.03708 0.17628 -0.210 0.833
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.063 on 1197 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.3339, Adjusted R-squared: 0.3289
## F-statistic: 66.66 on 9 and 1197 DF, p-value: < 2.2e-16##
## Call:
## lm(formula = Own_Nat_conf_diff ~ (pDem_Rep + pInd_Not) * tDur_Post,
## data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -4.0891 -0.7131 -0.0814 0.0929 3.9109
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.35073 0.02954 11.873 < 2e-16 ***
## pDem_Rep 0.65719 0.05797 11.337 < 2e-16 ***
## pInd_Not 0.39623 0.07342 5.397 8.15e-08 ***
## tDur_Post -0.02835 0.05908 -0.480 0.631
## pDem_Rep:tDur_Post 0.48023 0.11594 4.142 3.68e-05 ***
## pInd_Not:tDur_Post -0.05383 0.14683 -0.367 0.714
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.9194 on 1201 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.128, Adjusted R-squared: 0.1244
## F-statistic: 35.27 on 5 and 1201 DF, p-value: < 2.2e-16## Moderated Mediation: Positive
vote.pos.mod.med <- lm(voteLegit ~ positive.c + positive.c:tDur_Post + (pDem_Rep + pInd_Not)*tDur_Post, data = d)
summary(vote.pos.mod.med)##
## Call:
## lm(formula = voteLegit ~ positive.c + positive.c:tDur_Post +
## (pDem_Rep + pInd_Not) * tDur_Post, data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -2.96928 -0.73982 0.08622 0.75839 2.57620
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.38515 0.03502 96.673 < 2e-16 ***
## positive.c 0.19945 0.01880 10.608 < 2e-16 ***
## pDem_Rep -1.02610 0.07210 -14.231 < 2e-16 ***
## pInd_Not 0.32779 0.08829 3.713 0.000215 ***
## tDur_Post 0.05309 0.07003 0.758 0.448524
## positive.c:tDur_Post 0.05087 0.03760 1.353 0.176419
## tDur_Post:pDem_Rep -0.22919 0.14420 -1.589 0.112235
## tDur_Post:pInd_Not -0.05224 0.17658 -0.296 0.767422
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.076 on 1199 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.3166, Adjusted R-squared: 0.3126
## F-statistic: 79.35 on 7 and 1199 DF, p-value: < 2.2e-16##
## Call:
## lm(formula = Own_Nat_conf_diff ~ (pDem_Rep + pInd_Not) * tDur_Post,
## data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -4.0891 -0.7131 -0.0814 0.0929 3.9109
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.35073 0.02954 11.873 < 2e-16 ***
## pDem_Rep 0.65719 0.05797 11.337 < 2e-16 ***
## pInd_Not 0.39623 0.07342 5.397 8.15e-08 ***
## tDur_Post -0.02835 0.05908 -0.480 0.631
## pDem_Rep:tDur_Post 0.48023 0.11594 4.142 3.68e-05 ***
## pInd_Not:tDur_Post -0.05383 0.14683 -0.367 0.714
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.9194 on 1201 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.128, Adjusted R-squared: 0.1244
## F-statistic: 35.27 on 5 and 1201 DF, p-value: < 2.2e-16pos.mod.med <- lm(positive ~ (pDem_Rep + pInd_Not) * tDur_Post, data = d)
## Moderated Mediation: Negative
vote.neg.mod.med <- lm(voteLegit ~ negative.c + negative.c:tDur_Post + (pDem_Rep + pInd_Not)*tDur_Post, data = d)
summary(vote.neg.mod.med)##
## Call:
## lm(formula = voteLegit ~ negative.c + negative.c:tDur_Post +
## (pDem_Rep + pInd_Not) * tDur_Post, data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.5026 -0.7946 0.1356 0.8059 2.8460
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.32530 0.03567 93.226 < 2e-16 ***
## negative.c -0.14074 0.02163 -6.508 1.12e-10 ***
## pDem_Rep -1.23597 0.07060 -17.507 < 2e-16 ***
## pInd_Not 0.61619 0.08912 6.914 7.62e-12 ***
## tDur_Post 0.05849 0.07134 0.820 0.412416
## negative.c:tDur_Post -0.05071 0.04325 -1.172 0.241234
## tDur_Post:pDem_Rep -0.51696 0.14120 -3.661 0.000262 ***
## tDur_Post:pInd_Not 0.10797 0.17823 0.606 0.544788
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.108 on 1199 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.2751, Adjusted R-squared: 0.2709
## F-statistic: 65 on 7 and 1199 DF, p-value: < 2.2e-16##
## Call:
## lm(formula = Own_Nat_conf_diff ~ (pDem_Rep + pInd_Not) * tDur_Post,
## data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -4.0891 -0.7131 -0.0814 0.0929 3.9109
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.35073 0.02954 11.873 < 2e-16 ***
## pDem_Rep 0.65719 0.05797 11.337 < 2e-16 ***
## pInd_Not 0.39623 0.07342 5.397 8.15e-08 ***
## tDur_Post -0.02835 0.05908 -0.480 0.631
## pDem_Rep:tDur_Post 0.48023 0.11594 4.142 3.68e-05 ***
## pInd_Not:tDur_Post -0.05383 0.14683 -0.367 0.714
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.9194 on 1201 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.128, Adjusted R-squared: 0.1244
## F-statistic: 35.27 on 5 and 1201 DF, p-value: < 2.2e-16#### Negative - Positive Emotions
d$Neg_minus_Pos <- d$negative - d$positive
d$Pos_minus_Neg <- d$positive - d$negative
vote.posneg.mod.med <- lm(voteLegit ~ Pos_minus_Neg + Pos_minus_Neg:tDur_Post + (pDem_Rep + pInd_Not)*tDur_Post, data = d)
summary(vote.posneg.mod.med)##
## Call:
## lm(formula = voteLegit ~ Pos_minus_Neg + Pos_minus_Neg:tDur_Post +
## (pDem_Rep + pInd_Not) * tDur_Post, data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -2.99004 -0.79102 0.07291 0.72264 3.03111
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.30721 0.03436 96.263 < 2e-16 ***
## Pos_minus_Neg 0.15469 0.01311 11.797 < 2e-16 ***
## pDem_Rep -1.05839 0.07065 -14.981 < 2e-16 ***
## pInd_Not 0.45053 0.08547 5.271 1.6e-07 ***
## tDur_Post 0.01871 0.06871 0.272 0.785
## Pos_minus_Neg:tDur_Post 0.04138 0.02623 1.578 0.115
## tDur_Post:pDem_Rep -0.17558 0.14130 -1.243 0.214
## tDur_Post:pInd_Not -0.01931 0.17093 -0.113 0.910
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.065 on 1199 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.3305, Adjusted R-squared: 0.3266
## F-statistic: 84.56 on 7 and 1199 DF, p-value: < 2.2e-16iv. Media Measures
### Moderated Mediation
vote.media.mod.med <- lm(voteLegit ~ foxPerception.c + otherMediaPerception.c + (foxPerception.c + otherMediaPerception.c):tDur_Post + (pDem_Rep + pInd_Not)*tDur_Post, data = d)
summary(vote.media.mod.med)##
## Call:
## lm(formula = voteLegit ~ foxPerception.c + otherMediaPerception.c +
## (foxPerception.c + otherMediaPerception.c):tDur_Post + (pDem_Rep +
## pInd_Not) * tDur_Post, data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.5602 -0.7826 0.1285 0.8037 2.6642
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.34804 0.03544 94.472 < 2e-16 ***
## foxPerception.c -0.10975 0.02847 -3.856 0.000122 ***
## otherMediaPerception.c 0.28148 0.03791 7.425 2.14e-13 ***
## pDem_Rep -0.97658 0.08040 -12.146 < 2e-16 ***
## pInd_Not 0.53869 0.08835 6.097 1.46e-09 ***
## tDur_Post 0.10520 0.07088 1.484 0.138014
## foxPerception.c:tDur_Post -0.07837 0.05693 -1.377 0.168915
## otherMediaPerception.c:tDur_Post 0.10767 0.07582 1.420 0.155839
## tDur_Post:pDem_Rep -0.48974 0.16080 -3.046 0.002373 **
## tDur_Post:pInd_Not 0.05299 0.17671 0.300 0.764328
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.1 on 1197 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.2865, Adjusted R-squared: 0.2811
## F-statistic: 53.41 on 9 and 1197 DF, p-value: < 2.2e-16##
## Call:
## lm(formula = Own_Nat_conf_diff ~ (pDem_Rep + pInd_Not) * tDur_Post,
## data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -4.0891 -0.7131 -0.0814 0.0929 3.9109
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.35073 0.02954 11.873 < 2e-16 ***
## pDem_Rep 0.65719 0.05797 11.337 < 2e-16 ***
## pInd_Not 0.39623 0.07342 5.397 8.15e-08 ***
## tDur_Post -0.02835 0.05908 -0.480 0.631
## pDem_Rep:tDur_Post 0.48023 0.11594 4.142 3.68e-05 ***
## pInd_Not:tDur_Post -0.05383 0.14683 -0.367 0.714
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.9194 on 1201 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.128, Adjusted R-squared: 0.1244
## F-statistic: 35.27 on 5 and 1201 DF, p-value: < 2.2e-16## Moderated Mediation: Fox
vote.fox.mod.med <- lm(voteLegit ~ foxPerception.c + foxPerception.c:tDur_Post + (pDem_Rep + pInd_Not)*tDur_Post, data = d)
summary(vote.fox.mod.med)##
## Call:
## lm(formula = voteLegit ~ foxPerception.c + foxPerception.c:tDur_Post +
## (pDem_Rep + pInd_Not) * tDur_Post, data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.4505 -0.8991 0.0763 0.7269 2.4043
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.33389 0.03620 92.106 < 2e-16 ***
## foxPerception.c -0.05627 0.02815 -1.999 0.0458 *
## pDem_Rep -1.20336 0.07624 -15.783 < 2e-16 ***
## pInd_Not 0.56455 0.09030 6.252 5.62e-10 ***
## tDur_Post 0.07830 0.07239 1.082 0.2797
## foxPerception.c:tDur_Post -0.06353 0.05630 -1.128 0.2594
## tDur_Post:pDem_Rep -0.59855 0.15248 -3.925 9.15e-05 ***
## tDur_Post:pInd_Not 0.11301 0.18059 0.626 0.5316
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.126 on 1199 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.2517, Adjusted R-squared: 0.2473
## F-statistic: 57.62 on 7 and 1199 DF, p-value: < 2.2e-16##
## Call:
## lm(formula = Own_Nat_conf_diff ~ (pDem_Rep + pInd_Not) * tDur_Post,
## data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -4.0891 -0.7131 -0.0814 0.0929 3.9109
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.35073 0.02954 11.873 < 2e-16 ***
## pDem_Rep 0.65719 0.05797 11.337 < 2e-16 ***
## pInd_Not 0.39623 0.07342 5.397 8.15e-08 ***
## tDur_Post -0.02835 0.05908 -0.480 0.631
## pDem_Rep:tDur_Post 0.48023 0.11594 4.142 3.68e-05 ***
## pInd_Not:tDur_Post -0.05383 0.14683 -0.367 0.714
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.9194 on 1201 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.128, Adjusted R-squared: 0.1244
## F-statistic: 35.27 on 5 and 1201 DF, p-value: < 2.2e-16## Moderated Mediation: Other Media
vote.other.mod.med <- lm(voteLegit ~ otherMediaPerception.c + otherMediaPerception.c:tDur_Post + (pDem_Rep + pInd_Not)*tDur_Post, data = d)
summary(vote.other.mod.med)##
## Call:
## lm(formula = voteLegit ~ otherMediaPerception.c + otherMediaPerception.c:tDur_Post +
## (pDem_Rep + pInd_Not) * tDur_Post, data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.4108 -0.7926 0.1265 0.8034 2.3869
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.34877 0.03565 93.922 < 2e-16 ***
## otherMediaPerception.c 0.24502 0.03687 6.646 4.57e-11 ***
## pDem_Rep -1.10363 0.07335 -15.047 < 2e-16 ***
## pInd_Not 0.51480 0.08859 5.811 7.95e-09 ***
## tDur_Post 0.11082 0.07131 1.554 0.120425
## otherMediaPerception.c:tDur_Post 0.08609 0.07374 1.167 0.243247
## tDur_Post:pDem_Rep -0.56493 0.14669 -3.851 0.000124 ***
## tDur_Post:pInd_Not 0.05154 0.17718 0.291 0.771207
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.107 on 1199 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.2763, Adjusted R-squared: 0.272
## F-statistic: 65.38 on 7 and 1199 DF, p-value: < 2.2e-16##
## Call:
## lm(formula = Own_Nat_conf_diff ~ (pDem_Rep + pInd_Not) * tDur_Post,
## data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -4.0891 -0.7131 -0.0814 0.0929 3.9109
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.35073 0.02954 11.873 < 2e-16 ***
## pDem_Rep 0.65719 0.05797 11.337 < 2e-16 ***
## pInd_Not 0.39623 0.07342 5.397 8.15e-08 ***
## tDur_Post -0.02835 0.05908 -0.480 0.631
## pDem_Rep:tDur_Post 0.48023 0.11594 4.142 3.68e-05 ***
## pInd_Not:tDur_Post -0.05383 0.14683 -0.367 0.714
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.9194 on 1201 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.128, Adjusted R-squared: 0.1244
## F-statistic: 35.27 on 5 and 1201 DF, p-value: < 2.2e-16h. assessing differences in the pDem_Rep*tDur_Post parameter when positive emotion is added a mediator
𝑍 = β1−β2/sqrt((SEβ1)2+(SEβ2)2)
##
## Call:
## lm(formula = Own_Nat_conf_diff ~ (pDem_Rep + pInd_Not) * tDur_Post,
## data = d)
##
## Standardized Coefficients::
## (Intercept) pDem_Rep pInd_Not tDur_Post
## 0.00000000 0.30645352 0.14597571 -0.01441462
## pDem_Rep:tDur_Post pInd_Not:tDur_Post
## 0.11215925 -0.01101437##
## Call:
## lm(formula = voteLegit ~ positive.c + positive.c:tDur_Post +
## (pDem_Rep + pInd_Not) * tDur_Post, data = d)
##
## Standardized Coefficients::
## (Intercept) positive.c pDem_Rep
## 0.000000000 0.283953578 -0.362238354
## pInd_Not tDur_Post positive.c:tDur_Post
## 0.091423650 0.020436633 0.035974262
## tDur_Post:pDem_Rep tDur_Post:pInd_Not
## -0.040525360 -0.008091972##
## Call:
## lm(formula = Own_Nat_conf_diff ~ (pDem_Rep + pInd_Not) * tDur_Post,
## data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -4.0891 -0.7131 -0.0814 0.0929 3.9109
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.35073 0.02954 11.873 < 2e-16 ***
## pDem_Rep 0.65719 0.05797 11.337 < 2e-16 ***
## pInd_Not 0.39623 0.07342 5.397 8.15e-08 ***
## tDur_Post -0.02835 0.05908 -0.480 0.631
## pDem_Rep:tDur_Post 0.48023 0.11594 4.142 3.68e-05 ***
## pInd_Not:tDur_Post -0.05383 0.14683 -0.367 0.714
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.9194 on 1201 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.128, Adjusted R-squared: 0.1244
## F-statistic: 35.27 on 5 and 1201 DF, p-value: < 2.2e-16##
## Call:
## lm(formula = voteLegit ~ positive.c + positive.c:tDur_Post +
## (pDem_Rep + pInd_Not) * tDur_Post, data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -2.96928 -0.73982 0.08622 0.75839 2.57620
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.38515 0.03502 96.673 < 2e-16 ***
## positive.c 0.19945 0.01880 10.608 < 2e-16 ***
## pDem_Rep -1.02610 0.07210 -14.231 < 2e-16 ***
## pInd_Not 0.32779 0.08829 3.713 0.000215 ***
## tDur_Post 0.05309 0.07003 0.758 0.448524
## positive.c:tDur_Post 0.05087 0.03760 1.353 0.176419
## tDur_Post:pDem_Rep -0.22919 0.14420 -1.589 0.112235
## tDur_Post:pInd_Not -0.05224 0.17658 -0.296 0.767422
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.076 on 1199 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.3166, Adjusted R-squared: 0.3126
## F-statistic: 79.35 on 7 and 1199 DF, p-value: < 2.2e-16B1 <- -0.11505814
B2 <- -0.040525360
SEB1 <- 0.14219
SEB2 <- 0.14420
(Z <- (B1 - B2)/(sqrt((SEB1)^2 + (SEB2^2))))## [1] -0.36803893. Emotions
A. Negative Emotions
a. Descriptives
##
## Descriptive statistics by group
## group: Democrat
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 559 2.97 1.54 2.71 2.84 1.91 1 7 6 0.57 -0.58 0.07
## ------------------------------------------------------------
## group: Republican
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 464 3.06 1.45 2.86 2.97 1.69 1 7 6 0.47 -0.58 0.07
## ------------------------------------------------------------
## group: Independent
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 191 2.55 1.47 2.14 2.38 1.69 1 7 6 0.74 -0.44 0.11##
## Descriptive statistics by group
## : Democrat
## : During-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 251 3.34 1.58 3.14 3.26 1.91 1 7 6 0.32 -0.93 0.1
## ------------------------------------------------------------
## : Republican
## : During-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 238 2.93 1.38 2.86 2.83 1.48 1 7 6 0.57 -0.28 0.09
## ------------------------------------------------------------
## : Independent
## : During-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 87 2.6 1.52 2.29 2.44 1.91 1 6.86 5.86 0.64 -0.63 0.16
## ------------------------------------------------------------
## : Democrat
## : Post-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 308 2.67 1.44 2.43 2.51 1.69 1 7 6 0.81 -0.04 0.08
## ------------------------------------------------------------
## : Republican
## : Post-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 226 3.2 1.5 3 3.13 1.69 1 7 6 0.35 -0.84 0.1
## ------------------------------------------------------------
## : Independent
## : Post-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 104 2.51 1.44 1.86 2.34 1.27 1 7 6 0.81 -0.32 0.14b. Main ANOVA: Negative Emotions ~ Partisanship * Timing
negEmotion <- lm(negative ~ (pDem_Rep + pInd_Not) * (tDur_Post), data = d)
negEmotion.c <- lm(negative ~ (pDem_Rep + pInd_Not) + (tDur_Post), data = d)i. Interaction of partisanship and timing in full ANOVA model & Effect size
## Analysis of Variance Table
##
## Model 1: negative ~ (pDem_Rep + pInd_Not) + (tDur_Post)
## Model 2: negative ~ (pDem_Rep + pInd_Not) * (tDur_Post)
## Res.Df RSS Df Sum of Sq F Pr(>F)
## 1 1210 2696.2
## 2 1208 2640.1 2 56.098 12.834 3.051e-06 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1## lm(formula = negative ~ (pDem_Rep + pInd_Not) * (tDur_Post),
## data = d)
##
## Omnibus ANOVA
## SS df MS EtaSq F p
## Model 105.974 5 21.195 0.039 9.698 0
## Error 2640.049 1208 2.185
## Corr Total 2746.023 1213 2.264
##
## RMSE AdjEtaSq
## 1.478 0.035
##
## Coefficients
## Est StErr t SSR(3) EtaSq tol CI_2.5 CI_97.5
## (Intercept) 2.875 0.047 60.914 8109.126 0.754 NA 2.783 2.968
## pDem_Rep 0.058 0.093 0.622 0.845 0.000 0.994 -0.125 0.240
## pInd_Not 0.479 0.117 4.090 36.551 0.014 0.991 0.249 0.708
## tDur_Post -0.162 0.094 -1.713 6.410 0.002 0.810 -0.347 0.024
## pDem_Rep:tDur_Post 0.936 0.186 5.029 55.265 0.021 0.990 0.571 1.301
## pInd_Not:tDur_Post -0.103 0.234 -0.441 0.425 0.000 0.809 -0.563 0.356
## p
## (Intercept) 0.000
## pDem_Rep 0.534
## pInd_Not 0.000
## tDur_Post 0.087
## pDem_Rep:tDur_Post 0.000
## pInd_Not:tDur_Post 0.659negemo.aov <- ezANOVA(data = d[!is.na(d$negative),]
, dv = negative
, wid = s3
, between = .(party_factor, election_timing)
, type = 3
, return_aov = T)## Warning: Converting "s3" to factor for ANOVA.## Warning: You have removed one or more levels from variable "party_factor".
## Refactoring for ANOVA.## Warning: Data is unbalanced (unequal N per group). Make sure you specified a
## well-considered value for the type argument to ezANOVA().## Coefficient covariances computed by hccm()## $ANOVA
## Effect DFn DFd F p p<.05 ges
## 2 party_factor 2 1208 8.476903 2.208628e-04 * 0.01384036
## 3 election_timing 1 1208 2.899911 8.884167e-02 0.00239484
## 4 party_factor:election_timing 2 1208 12.834387 3.051341e-06 * 0.02080686
##
## $`Levene's Test for Homogeneity of Variance`
## DFn DFd SSn SSd F p p<.05
## 1 5 1208 6.438373 906.1957 1.716529 0.1278207
##
## $aov
## Call:
## aov(formula = formula(aov_formula), data = data)
##
## Terms:
## party_factor election_timing party_factor:election_timing
## Sum of Squares 36.0294 13.8460 56.0984
## Deg. of Freedom 2 1 2
## Residuals
## Sum of Squares 2640.0493
## Deg. of Freedom 1208
##
## Residual standard error: 1.478334
## Estimated effects may be unbalanced
## 1 observation deleted due to missingnessc. Effects for Democrats
##
## Call:
## lm(formula = negative ~ (pDemR + pDemI) * (tDur_Post), data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -2.3364 -1.2387 -0.2119 1.0738 4.4915
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.00446 0.06285 47.800 < 2e-16 ***
## pDemR 0.05787 0.09308 0.622 0.534232
## pDemI -0.44973 0.12444 -3.614 0.000314 ***
## tDur_Post -0.66383 0.12571 -5.281 1.53e-07 ***
## pDemR:tDur_Post 0.93614 0.18616 5.029 5.69e-07 ***
## pDemI:tDur_Post 0.57131 0.24887 2.296 0.021869 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.478 on 1208 degrees of freedom
## (22 observations deleted due to missingness)
## Multiple R-squared: 0.03859, Adjusted R-squared: 0.03461
## F-statistic: 9.698 on 5 and 1208 DF, p-value: 4.284e-09## lm(formula = negative ~ (pDemR + pDemI) * (tDur_Post), data = d)
##
## Omnibus ANOVA
## SS df MS EtaSq F p
## Model 105.974 5 21.195 0.039 9.698 0
## Error 2640.049 1208 2.185
## Corr Total 2746.023 1213 2.264
##
## RMSE AdjEtaSq
## 1.478 0.035
##
## Coefficients
## Est StErr t SSR(3) EtaSq tol CI_2.5 CI_97.5 p
## (Intercept) 3.004 0.063 47.800 4993.489 0.654 NA 2.881 3.128 0.000
## pDemR 0.058 0.093 0.622 0.845 0.000 0.880 -0.125 0.240 0.534
## pDemI -0.450 0.124 -3.614 28.546 0.011 0.877 -0.694 -0.206 0.000
## tDur_Post -0.664 0.126 -5.281 60.943 0.023 0.457 -0.910 -0.417 0.000
## pDemR:tDur_Post 0.936 0.186 5.029 55.265 0.021 0.544 0.571 1.301 0.000
## pDemI:tDur_Post 0.571 0.249 2.296 11.517 0.004 0.740 0.083 1.060 0.022d. Effects for Republicans
##
## Call:
## lm(formula = negative ~ (pRepD + pRepI) * (tDur_Post), data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -2.3364 -1.2387 -0.2119 1.0738 4.4915
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.06233 0.06865 44.606 < 2e-16 ***
## pRepD -0.05787 0.09308 -0.622 0.5342
## pRepI -0.50760 0.12746 -3.982 7.23e-05 ***
## tDur_Post 0.27231 0.13731 1.983 0.0476 *
## pRepD:tDur_Post -0.93614 0.18616 -5.029 5.69e-07 ***
## pRepI:tDur_Post -0.36483 0.25493 -1.431 0.1527
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.478 on 1208 degrees of freedom
## (22 observations deleted due to missingness)
## Multiple R-squared: 0.03859, Adjusted R-squared: 0.03461
## F-statistic: 9.698 on 5 and 1208 DF, p-value: 4.284e-09## lm(formula = negative ~ (pRepD + pRepI) * (tDur_Post), data = d)
##
## Omnibus ANOVA
## SS df MS EtaSq F p
## Model 105.974 5 21.195 0.039 9.698 0
## Error 2640.049 1208 2.185
## Corr Total 2746.023 1213 2.264
##
## RMSE AdjEtaSq
## 1.478 0.035
##
## Coefficients
## Est StErr t SSR(3) EtaSq tol CI_2.5 CI_97.5 p
## (Intercept) 3.062 0.069 44.606 4348.410 0.622 NA 2.928 3.197 0.000
## pRepD -0.058 0.093 -0.622 0.845 0.000 0.836 -0.240 0.125 0.534
## pRepI -0.508 0.127 -3.982 34.659 0.013 0.836 -0.758 -0.258 0.000
## tDur_Post 0.272 0.137 1.983 8.596 0.003 0.383 0.003 0.542 0.048
## pRepD:tDur_Post -0.936 0.186 -5.029 55.265 0.021 0.453 -1.301 -0.571 0.000
## pRepI:tDur_Post -0.365 0.255 -1.431 4.476 0.002 0.705 -0.865 0.135 0.153e. Effects for Independents
##
## Call:
## lm(formula = negative ~ (pIndR + pIndD) * (tDur_Post), data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -2.3364 -1.2387 -0.2119 1.0738 4.4915
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.55473 0.10739 23.788 < 2e-16 ***
## pIndR 0.50760 0.12746 3.982 7.23e-05 ***
## pIndD 0.44973 0.12444 3.614 0.000314 ***
## tDur_Post -0.09251 0.21479 -0.431 0.666747
## pIndR:tDur_Post 0.36483 0.25493 1.431 0.152658
## pIndD:tDur_Post -0.57131 0.24887 -2.296 0.021869 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.478 on 1208 degrees of freedom
## (22 observations deleted due to missingness)
## Multiple R-squared: 0.03859, Adjusted R-squared: 0.03461
## F-statistic: 9.698 on 5 and 1208 DF, p-value: 4.284e-09## lm(formula = negative ~ (pIndR + pIndD) * (tDur_Post), data = d)
##
## Omnibus ANOVA
## SS df MS EtaSq F p
## Model 105.974 5 21.195 0.039 9.698 0
## Error 2640.049 1208 2.185
## Corr Total 2746.023 1213 2.264
##
## RMSE AdjEtaSq
## 1.478 0.035
##
## Coefficients
## Est StErr t SSR(3) EtaSq tol CI_2.5 CI_97.5 p
## (Intercept) 2.555 0.107 23.788 1236.712 0.319 NA 2.344 2.765 0.000
## pIndR 0.508 0.127 3.982 34.659 0.013 0.469 0.258 0.758 0.000
## pIndD 0.450 0.124 3.614 28.546 0.011 0.468 0.206 0.694 0.000
## tDur_Post -0.093 0.215 -0.431 0.405 0.000 0.156 -0.514 0.329 0.667
## pIndR:tDur_Post 0.365 0.255 1.431 4.476 0.002 0.290 -0.135 0.865 0.153
## pIndD:tDur_Post -0.571 0.249 -2.296 11.517 0.004 0.254 -1.060 -0.083 0.022f. Does ID strength moderate?
Marginal interaction effect of ID strength and partisan ID.
B. Positive Emotions
a. Descriptives
##
## Descriptive statistics by group
## group: Democrat
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 559 3.97 1.83 4 3.95 2.37 1 7 6 0.1 -1.13 0.08
## ------------------------------------------------------------
## group: Republican
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 464 2.88 1.69 2.4 2.7 1.78 1 7 6 0.68 -0.56 0.08
## ------------------------------------------------------------
## group: Independent
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 191 2.34 1.52 1.8 2.11 1.19 1 7 6 1.02 0.23 0.11##
## Descriptive statistics by group
## : Democrat
## : During-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 251 3.26 1.62 3 3.14 1.78 1 7 6 0.52 -0.62 0.1
## ------------------------------------------------------------
## : Republican
## : During-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 238 3.14 1.64 3 3.03 1.78 1 7 6 0.43 -0.79 0.11
## ------------------------------------------------------------
## : Independent
## : During-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 87 2.4 1.43 1.8 2.24 1.19 1 6.2 5.2 0.74 -0.6 0.15
## ------------------------------------------------------------
## : Democrat
## : Post-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 308 4.55 1.78 4.8 4.64 2.08 1 7 6 -0.27 -1.01 0.1
## ------------------------------------------------------------
## : Republican
## : Post-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 226 2.61 1.71 2 2.36 1.48 1 7 6 1 -0.04 0.11
## ------------------------------------------------------------
## : Independent
## : Post-election
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 104 2.28 1.6 1.5 2.01 0.74 1 7 6 1.2 0.67 0.16b. Main ANOVA: Positive Emotions ~ Partisanship * Timing
##
## Call:
## lm(formula = positive ~ (pDem_Rep + pInd_Not) * (tDur_Post),
## data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.5545 -1.2846 -0.2846 1.2455 4.7154
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.04482 0.05329 57.135 < 2e-16 ***
## pDem_Rep -1.03139 0.10508 -9.815 < 2e-16 ***
## pInd_Not 1.04681 0.13214 7.922 5.26e-15 ***
## tDur_Post 0.21709 0.10658 2.037 0.0419 *
## pDem_Rep:tDur_Post -1.83039 0.21016 -8.709 < 2e-16 ***
## pInd_Not:tDur_Post 0.49966 0.26427 1.891 0.0589 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.669 on 1208 degrees of freedom
## (22 observations deleted due to missingness)
## Multiple R-squared: 0.1876, Adjusted R-squared: 0.1842
## F-statistic: 55.78 on 5 and 1208 DF, p-value: < 2.2e-16##
## Call:
## lm(formula = positive ~ (pDem_Rep + pInd_Not) + (tDur_Post),
## data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.1389 -1.5006 -0.3006 1.2994 4.4926
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.05737 0.05488 55.710 < 2e-16 ***
## pDem_Rep -1.06689 0.10835 -9.846 < 2e-16 ***
## pInd_Not 1.09802 0.13588 8.081 1.54e-15 ***
## tDur_Post 0.37139 0.09918 3.745 0.000189 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.722 on 1210 degrees of freedom
## (22 observations deleted due to missingness)
## Multiple R-squared: 0.1333, Adjusted R-squared: 0.1312
## F-statistic: 62.03 on 3 and 1210 DF, p-value: < 2.2e-16i. Interaction of partisanship and timing in main model & Effect size
## Analysis of Variance Table
##
## Model 1: positive ~ (pDem_Rep + pInd_Not) + (tDur_Post)
## Model 2: positive ~ (pDem_Rep + pInd_Not) * (tDur_Post)
## Res.Df RSS Df Sum of Sq F Pr(>F)
## 1 1210 3589.4
## 2 1208 3364.7 2 224.72 40.34 < 2.2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1## lm(formula = positive ~ (pDem_Rep + pInd_Not) * (tDur_Post),
## data = d)
##
## Omnibus ANOVA
## SS df MS EtaSq F p
## Model 776.781 5 155.356 0.188 55.776 0
## Error 3364.715 1208 2.785
## Corr Total 4141.496 1213 3.414
##
## RMSE AdjEtaSq
## 1.669 0.184
##
## Coefficients
## Est StErr t SSR(3) EtaSq tol CI_2.5 CI_97.5
## (Intercept) 3.045 0.053 57.135 9092.649 0.730 NA 2.940 3.149
## pDem_Rep -1.031 0.105 -9.815 268.334 0.074 0.994 -1.238 -0.825
## pInd_Not 1.047 0.132 7.922 174.812 0.049 0.991 0.788 1.306
## tDur_Post 0.217 0.107 2.037 11.556 0.003 0.810 0.008 0.426
## pDem_Rep:tDur_Post -1.830 0.210 -8.709 211.279 0.059 0.990 -2.243 -1.418
## pInd_Not:tDur_Post 0.500 0.264 1.891 9.957 0.003 0.809 -0.019 1.018
## p
## (Intercept) 0.000
## pDem_Rep 0.000
## pInd_Not 0.000
## tDur_Post 0.042
## pDem_Rep:tDur_Post 0.000
## pInd_Not:tDur_Post 0.059posemo.aov <- ezANOVA(data = d[!is.na(d$positive),]
, dv = positive
, wid = s3
, between = .(party_factor, election_timing)
, type = 3
, return_aov = T)## Warning: Converting "s3" to factor for ANOVA.## Warning: You have removed one or more levels from variable "party_factor".
## Refactoring for ANOVA.## Warning: Data is unbalanced (unequal N per group). Make sure you specified a
## well-considered value for the type argument to ezANOVA().## Coefficient covariances computed by hccm()## $ANOVA
## Effect DFn DFd F p p<.05
## 2 party_factor 2 1208 82.371104 2.915279e-34 *
## 3 election_timing 1 1208 4.056453 4.422419e-02 *
## 4 party_factor:election_timing 2 1208 40.339863 1.098628e-17 *
## ges
## 2 0.120009574
## 3 0.003346752
## 4 0.062606500
##
## $`Levene's Test for Homogeneity of Variance`
## DFn DFd SSn SSd F p p<.05
## 1 5 1208 11.77065 1290.089 2.204336 0.05167201
##
## $aov
## Call:
## aov(formula = formula(aov_formula), data = data)
##
## Terms:
## party_factor election_timing party_factor:election_timing
## Sum of Squares 510.460 41.600 224.722
## Deg. of Freedom 2 1 2
## Residuals
## Sum of Squares 3364.715
## Deg. of Freedom 1208
##
## Residual standard error: 1.66894
## Estimated effects may be unbalanced
## 1 observation deleted due to missingnessc. dummy coded for dem
##
## Call:
## lm(formula = positive ~ (pDemR + pDemI) * (tDur_Post), data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.5545 -1.2846 -0.2846 1.2455 4.7154
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.90596 0.07096 55.046 < 2e-16 ***
## pDemR -1.03139 0.10508 -9.815 < 2e-16 ***
## pDemI -1.56250 0.14048 -11.123 < 2e-16 ***
## tDur_Post 1.29717 0.14192 9.140 < 2e-16 ***
## pDemR:tDur_Post -1.83039 0.21016 -8.709 < 2e-16 ***
## pDemI:tDur_Post -1.41486 0.28096 -5.036 5.48e-07 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.669 on 1208 degrees of freedom
## (22 observations deleted due to missingness)
## Multiple R-squared: 0.1876, Adjusted R-squared: 0.1842
## F-statistic: 55.78 on 5 and 1208 DF, p-value: < 2.2e-16## lm(formula = positive ~ (pDemR + pDemI) * (tDur_Post), data = d)
##
## Omnibus ANOVA
## SS df MS EtaSq F p
## Model 776.781 5 155.356 0.188 55.776 0
## Error 3364.715 1208 2.785
## Corr Total 4141.496 1213 3.414
##
## RMSE AdjEtaSq
## 1.669 0.184
##
## Coefficients
## Est StErr t SSR(3) EtaSq tol CI_2.5 CI_97.5 p
## (Intercept) 3.906 0.071 55.046 8439.715 0.715 NA 3.767 4.045 0
## pDemR -1.031 0.105 -9.815 268.334 0.074 0.880 -1.238 -0.825 0
## pDemI -1.563 0.140 -11.123 344.583 0.093 0.877 -1.838 -1.287 0
## tDur_Post 1.297 0.142 9.140 232.707 0.065 0.457 1.019 1.576 0
## pDemR:tDur_Post -1.830 0.210 -8.709 211.279 0.059 0.544 -2.243 -1.418 0
## pDemI:tDur_Post -1.415 0.281 -5.036 70.635 0.021 0.740 -1.966 -0.864 0d. dummy coded for rep
##
## Call:
## lm(formula = positive ~ (pRepD + pRepI) * (tDur_Post), data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.5545 -1.2846 -0.2846 1.2455 4.7154
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.8746 0.0775 37.089 < 2e-16 ***
## pRepD 1.0314 0.1051 9.815 < 2e-16 ***
## pRepI -0.5311 0.1439 -3.691 0.000233 ***
## tDur_Post -0.5332 0.1550 -3.440 0.000602 ***
## pRepD:tDur_Post 1.8304 0.2102 8.709 < 2e-16 ***
## pRepI:tDur_Post 0.4155 0.2878 1.444 0.149044
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.669 on 1208 degrees of freedom
## (22 observations deleted due to missingness)
## Multiple R-squared: 0.1876, Adjusted R-squared: 0.1842
## F-statistic: 55.78 on 5 and 1208 DF, p-value: < 2.2e-16## lm(formula = positive ~ (pRepD + pRepI) * (tDur_Post), data = d)
##
## Omnibus ANOVA
## SS df MS EtaSq F p
## Model 776.781 5 155.356 0.188 55.776 0
## Error 3364.715 1208 2.785
## Corr Total 4141.496 1213 3.414
##
## RMSE AdjEtaSq
## 1.669 0.184
##
## Coefficients
## Est StErr t SSR(3) EtaSq tol CI_2.5 CI_97.5 p
## (Intercept) 2.875 0.078 37.089 3831.540 0.532 NA 2.723 3.027 0.000
## pRepD 1.031 0.105 9.815 268.334 0.074 0.836 0.825 1.238 0.000
## pRepI -0.531 0.144 -3.691 37.945 0.011 0.836 -0.813 -0.249 0.000
## tDur_Post -0.533 0.155 -3.440 32.958 0.010 0.383 -0.837 -0.229 0.001
## pRepD:tDur_Post 1.830 0.210 8.709 211.279 0.059 0.453 1.418 2.243 0.000
## pRepI:tDur_Post 0.416 0.288 1.444 5.807 0.002 0.705 -0.149 0.980 0.149e. dummy coded for independent
##
## Call:
## lm(formula = positive ~ (pIndR + pIndD) * (tDur_Post), data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.5545 -1.2846 -0.2846 1.2455 4.7154
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.3435 0.1212 19.329 < 2e-16 ***
## pIndR 0.5311 0.1439 3.691 0.000233 ***
## pIndD 1.5625 0.1405 11.123 < 2e-16 ***
## tDur_Post -0.1177 0.2425 -0.485 0.627532
## pIndR:tDur_Post -0.4155 0.2878 -1.444 0.149044
## pIndD:tDur_Post 1.4149 0.2810 5.036 5.48e-07 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.669 on 1208 degrees of freedom
## (22 observations deleted due to missingness)
## Multiple R-squared: 0.1876, Adjusted R-squared: 0.1842
## F-statistic: 55.78 on 5 and 1208 DF, p-value: < 2.2e-16## lm(formula = positive ~ (pIndR + pIndD) * (tDur_Post), data = d)
##
## Omnibus ANOVA
## SS df MS EtaSq F p
## Model 776.781 5 155.356 0.188 55.776 0
## Error 3364.715 1208 2.785
## Corr Total 4141.496 1213 3.414
##
## RMSE AdjEtaSq
## 1.669 0.184
##
## Coefficients
## Est StErr t SSR(3) EtaSq tol CI_2.5 CI_97.5 p
## (Intercept) 2.343 0.121 19.329 1040.623 0.236 NA 2.106 2.581 0.000
## pIndR 0.531 0.144 3.691 37.945 0.011 0.469 0.249 0.813 0.000
## pIndD 1.563 0.140 11.123 344.583 0.093 0.468 1.287 1.838 0.000
## tDur_Post -0.118 0.242 -0.485 0.656 0.000 0.156 -0.593 0.358 0.628
## pIndR:tDur_Post -0.416 0.288 -1.444 5.807 0.002 0.290 -0.980 0.149 0.149
## pIndD:tDur_Post 1.415 0.281 5.036 70.635 0.021 0.254 0.864 1.966 0.000f. Does ID strength moderate?
4. Media Trust/Consumption
a. descriptives
##
## Descriptive statistics by group
## group: Democrat
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 558 2.89 0.66 2.86 2.86 0.58 1 5 4 0.37 0.6 0.03
## ------------------------------------------------------------
## group: Republican
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 464 1.99 0.84 1.86 1.9 0.9 1 5 4 0.94 0.62 0.04
## ------------------------------------------------------------
## group: Independent
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 190 2.35 0.79 2.3 2.31 0.5 1 5 4 0.64 0.96 0.06##
## Descriptive statistics by group
## group: Democrat
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 558 1.97 1.1 1.5 1.8 0.74 1 5 4 1.01 0.07 0.05
## ------------------------------------------------------------
## group: Republican
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 464 2.96 1.28 3 2.95 1.48 1 5 4 0.04 -1.24 0.06
## ------------------------------------------------------------
## group: Independent
## vars n mean sd median trimmed mad min max range skew kurtosis se
## X1 1 190 2.16 0.97 2 2.06 0.74 1 5 4 0.73 0.02 0.07b. (other - fox perception) ~ party
d$Other_minus_Fox_perc <- NA
d$Other_minus_Fox_perc <- d$otherMediaPerception - d$foxPerception
percept <- lm(Other_minus_Fox_perc ~ (pDem_Rep + pInd_Not), data = d)
summary(percept)##
## Call:
## lm(formula = Other_minus_Fox_perc ~ (pDem_Rep + pInd_Not), data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -4.1343 -0.8499 0.1214 0.9372 3.9316
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.04827 0.03943 1.224 0.221
## pDem_Rep -1.88730 0.07772 -24.285 <2e-16 ***
## pInd_Not -0.21804 0.09779 -2.230 0.026 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.237 on 1209 degrees of freedom
## (24 observations deleted due to missingness)
## Multiple R-squared: 0.3285, Adjusted R-squared: 0.3274
## F-statistic: 295.8 on 2 and 1209 DF, p-value: < 2.2e-16c. vote legitimacy ~ party * media type - for 2 df test
perception.m1 <- lm(voteLegit ~ pDem_Rep + pInd_Not + otherMediaPerception + foxPerception, data = d)
summary(perception.m1)##
## Call:
## lm(formula = voteLegit ~ pDem_Rep + pInd_Not + otherMediaPerception +
## foxPerception, data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.3653 -0.7999 0.1392 0.7739 2.8750
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.40004 0.11331 21.181 < 2e-16 ***
## pDem_Rep -0.68869 0.08439 -8.161 8.34e-16 ***
## pInd_Not 0.56222 0.08568 6.562 7.89e-11 ***
## otherMediaPerception 0.51439 0.04189 12.278 < 2e-16 ***
## foxPerception -0.12612 0.02736 -4.609 4.47e-06 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.072 on 1202 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.3202, Adjusted R-squared: 0.318
## F-statistic: 141.6 on 4 and 1202 DF, p-value: < 2.2e-16##
## Call:
## lm(formula = voteLegit ~ pDem_Rep + pInd_Not + otherMediaPerception +
## foxPerception, data = d)
##
## Standardized Coefficients::
## (Intercept) pDem_Rep pInd_Not
## 0.0000000 -0.2431231 0.1568096
## otherMediaPerception foxPerception
## 0.3408677 -0.1207183d. vote legitimacy ~ party
##
## Call:
## lm(formula = voteLegit ~ pDem_Rep + pInd_Not, data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.1679 -0.8920 0.1080 0.8321 2.1080
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.34447 0.03645 91.749 < 2e-16 ***
## pDem_Rep -1.27585 0.07156 -17.828 < 2e-16 ***
## pInd_Not 0.56202 0.09058 6.205 7.52e-10 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.138 on 1204 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.2326, Adjusted R-squared: 0.2313
## F-statistic: 182.4 on 2 and 1204 DF, p-value: < 2.2e-16##
## Call:
## lm(formula = voteLegit ~ pDem_Rep + pInd_Not, data = d)
##
## Standardized Coefficients::
## (Intercept) pDem_Rep pInd_Not
## 0.0000000 -0.4504079 0.1567542e. anova between simple and augmented models: Interaction of media type and partisanship
## Analysis of Variance Table
##
## Model 1: voteLegit ~ pDem_Rep + pInd_Not
## Model 2: voteLegit ~ pDem_Rep + pInd_Not + otherMediaPerception + foxPerception
## Res.Df RSS Df Sum of Sq F Pr(>F)
## 1 1204 1559.0
## 2 1202 1380.9 2 178.06 77.493 < 2.2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1i. By ez ANOVA approach
### Prep
mediap <- as.data.frame(cbind(d[d$election_timing != "Pre-election",]$party_factor,
d[d$election_timing != "Pre-election",]$otherMediaPerception,
d[d$election_timing != "Pre-election",]$foxPerception,
d[d$election_timing != "Pre-election",]$s3,
d[d$election_timing != "Pre-election",]$election_timing))
names(mediap) <- c("party_factor", "otherMedia","fox", "s3", "election_timing")
mediap$party_factor <- recode_factor(mediap$party_factor, `1` = "Democrat", `2` = "Republican", `3` = "Independent")
m1 <- tidyr::gather(mediap, media_type, perception, otherMedia:fox, factor_key=TRUE)
## Results
m1 <- na.omit(m1)
media.aov <- ezANOVA(data = m1
, dv = perception
, wid = s3
, within = media_type
, between = .(party_factor)
, type = 3
, return_aov = T)## Warning: Converting "s3" to factor for ANOVA.## Warning: Data is unbalanced (unequal N per group). Make sure you specified a
## well-considered value for the type argument to ezANOVA().## $ANOVA
## Effect DFn DFd F p p<.05 ges
## 2 party_factor 2 1209 5.658348 3.581296e-03 * 0.0055689106
## 3 media_type 1 1209 1.533206 2.158723e-01 0.0005091899
## 4 party_factor:media_type 2 1209 295.776634 2.715111e-105 * 0.1642706941
##
## $aov
##
## Call:
## aov(formula = formula(aov_formula), data = data)
##
## Grand Mean: 2.417255
##
## Stratum 1: s3
##
## Terms:
## party_factor Residuals
## Sum of Squares 12.8938 1377.4899
## Deg. of Freedom 2 1209
##
## Residual standard error: 1.06741
## 2 out of 4 effects not estimable
## Estimated effects may be unbalanced
##
## Stratum 2: s3:media_type
##
## Terms:
## media_type party_factor:media_type Residuals
## Sum of Squares 4.2440 452.5647 924.9392
## Deg. of Freedom 1 2 1209
##
## Residual standard error: 0.8746684
## Estimated effects may be unbalancedf. other method, using OtherMedia - Fox to account for the two-level factor of media type! Instead of using them both separately.
perception_diff <- lm(voteLegit ~ (pDem_Rep + pInd_Not) * Other_minus_Fox_perc, data = d)
perception_diff.c <- lm(voteLegit ~ (pDem_Rep + pInd_Not) + Other_minus_Fox_perc, data = d)
anova(perception_diff.c, perception_diff)## Analysis of Variance Table
##
## Model 1: voteLegit ~ (pDem_Rep + pInd_Not) + Other_minus_Fox_perc
## Model 2: voteLegit ~ (pDem_Rep + pInd_Not) * Other_minus_Fox_perc
## Res.Df RSS Df Sum of Sq F Pr(>F)
## 1 1203 1466.7
## 2 1201 1443.3 2 23.399 9.7352 6.397e-05 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1g. Effect sizes by party for each news outlet
## lm(formula = foxPerception ~ (pDem_Rep + pInd_Not), data = d)
##
## Omnibus ANOVA
## SS df MS EtaSq F p
## Model 258.443 2 129.222 0.138 96.962 0
## Error 1611.244 1209 1.333
## Corr Total 1869.687 1211 1.544
##
## RMSE AdjEtaSq
## 1.154 0.137
##
## Coefficients
## Est StErr t SSR(3) EtaSq tol CI_2.5 CI_97.5 p
## (Intercept) 2.361 0.037 64.163 5486.689 0.773 NA 2.289 2.433 0.000
## pDem_Rep 0.989 0.073 13.635 247.782 0.133 0.999 0.847 1.131 0.000
## pInd_Not 0.303 0.091 3.325 14.732 0.009 0.999 0.124 0.482 0.001# Other by party
othermed.mod <- lm(otherMediaPerception ~ (pDem_Rep + pInd_Not), data = d)
mcSummary(othermed.mod)## lm(formula = otherMediaPerception ~ (pDem_Rep + pInd_Not), data = d)
##
## Omnibus ANOVA
## SS df MS EtaSq F p
## Model 207.015 2 103.508 0.23 181.053 0
## Error 691.185 1209 0.572
## Corr Total 898.200 1211 0.742
##
## RMSE AdjEtaSq
## 0.756 0.229
##
## Coefficients
## Est StErr t SSR(3) EtaSq tol CI_2.5 CI_97.5 p
## (Intercept) 2.409 0.024 99.968 5713.316 0.892 NA 2.362 2.457 0.000
## pDem_Rep -0.898 0.048 -18.910 204.443 0.228 0.999 -0.992 -0.805 0.000
## pInd_Not 0.085 0.060 1.429 1.167 0.002 0.999 -0.032 0.203 0.153h. Does ID strength moderate?
# Fox by party
fox.mod1 <- lm(foxPerception ~ (pDem_Rep) * IDstrength, data = d[d$party_factor != "Independent",])
# Other by party
othermed.mod1 <- lm(otherMediaPerception ~ (pDem_Rep) * IDstrength, data = d[d$party_factor != "Independent",])
mcSummary(othermed.mod1)## lm(formula = otherMediaPerception ~ (pDem_Rep) * IDstrength,
## data = d[d$party_factor != "Independent", ])
##
## Omnibus ANOVA
## SS df MS EtaSq F p
## Model 206.784 3 68.928 0.268 122.626 0
## Error 565.475 1006 0.562
## Corr Total 772.260 1009 0.765
##
## RMSE AdjEtaSq
## 0.75 0.266
##
## Coefficients
## Est StErr t SSR(3) EtaSq tol CI_2.5 CI_97.5
## (Intercept) 2.439 0.024 103.011 5964.675 0.913 NA 2.393 2.485
## pDem_Rep -0.906 0.047 -19.134 205.786 0.267 1.000 -0.999 -0.813
## IDstrength 0.027 0.027 1.002 0.564 0.001 0.996 -0.026 0.080
## pDem_Rep:IDstrength -0.021 0.054 -0.381 0.082 0.000 0.996 -0.127 0.086
## p
## (Intercept) 0.000
## pDem_Rep 0.000
## IDstrength 0.317
## pDem_Rep:IDstrength 0.7035. Media Analyses: Vote legitimacy ~ party * foxPerception * otherMediaPerception
a. full model
vote.media.party.int <- lm(voteLegit ~ foxPerception.c * otherMediaPerception.c * (pDem_Rep + pInd_Not), data = d)
tab_model(vote.media.party.int)| vote Legit | |||
|---|---|---|---|
| Predictors | Estimates | CI | p |
| (Intercept) | 3.35 | 3.27 – 3.44 | <0.001 |
| foxPerception.c | -0.19 | -0.27 – -0.12 | <0.001 |
| otherMediaPerception.c | 0.37 | 0.27 – 0.47 | <0.001 |
| pDem_Rep | -1.02 | -1.20 – -0.83 | <0.001 |
| pInd_Not | 0.57 | 0.37 – 0.77 | <0.001 |
|
foxPerception.c * otherMediaPerception.c |
-0.01 | -0.07 – 0.06 | 0.823 |
|
foxPerception.c * pDem_Rep |
0.11 | -0.02 – 0.25 | 0.104 |
|
foxPerception.c * pInd_Not |
0.28 | 0.09 – 0.48 | 0.005 |
|
otherMediaPerception.c * pDem_Rep |
0.18 | -0.04 – 0.39 | 0.114 |
|
otherMediaPerception.c * pInd_Not |
-0.28 | -0.51 – -0.05 | 0.018 |
|
(foxPerception.c otherMediaPerception.c) pDem_Rep |
-0.03 | -0.16 – 0.11 | 0.699 |
|
(foxPerception.c otherMediaPerception.c) pInd_Not |
0.00 | -0.16 – 0.16 | 0.971 |
| Observations | 1207 | ||
| R2 / R2 adjusted | 0.282 / 0.276 | ||
b. main model and simple model
vote.party.media <- lm(voteLegit ~ (pDem_Rep + pInd_Not) + foxPerception + otherMediaPerception, data = d)
summary(vote.party.media) ##
## Call:
## lm(formula = voteLegit ~ (pDem_Rep + pInd_Not) + foxPerception +
## otherMediaPerception, data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.3653 -0.7999 0.1392 0.7739 2.8750
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.40004 0.11331 21.181 < 2e-16 ***
## pDem_Rep -0.68869 0.08439 -8.161 8.34e-16 ***
## pInd_Not 0.56222 0.08568 6.562 7.89e-11 ***
## foxPerception -0.12612 0.02736 -4.609 4.47e-06 ***
## otherMediaPerception 0.51439 0.04189 12.278 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.072 on 1202 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.3202, Adjusted R-squared: 0.318
## F-statistic: 141.6 on 4 and 1202 DF, p-value: < 2.2e-16##
## Call:
## lm(formula = voteLegit ~ (pDem_Rep + pInd_Not) + foxPerception +
## otherMediaPerception, data = d)
##
## Standardized Coefficients::
## (Intercept) pDem_Rep pInd_Not
## 0.0000000 -0.2431231 0.1568096
## foxPerception otherMediaPerception
## -0.1207183 0.3408677##
## Call:
## lm(formula = voteLegit ~ (pDem_Rep + pInd_Not), data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.1679 -0.8920 0.1080 0.8321 2.1080
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.34447 0.03645 91.749 < 2e-16 ***
## pDem_Rep -1.27585 0.07156 -17.828 < 2e-16 ***
## pInd_Not 0.56202 0.09058 6.205 7.52e-10 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.138 on 1204 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.2326, Adjusted R-squared: 0.2313
## F-statistic: 182.4 on 2 and 1204 DF, p-value: < 2.2e-16##
## Call:
## lm(formula = voteLegit ~ (pDem_Rep + pInd_Not), data = d)
##
## Standardized Coefficients::
## (Intercept) pDem_Rep pInd_Not
## 0.0000000 -0.4504079 0.1567542𝑍 = β1−β2/sqrt((SEβ1)2+(SEβ2)2)
c. assessing differences in the pDem_Rep parameter when media measures are added as controls: Using output from lm.beta commands, so that the parameters are standardized
##
## Call:
## lm(formula = voteLegit ~ (pDem_Rep + pInd_Not), data = d)
##
## Standardized Coefficients::
## (Intercept) pDem_Rep pInd_Not
## 0.0000000 -0.4504079 0.1567542##
## Call:
## lm(formula = voteLegit ~ (pDem_Rep + pInd_Not) + foxPerception +
## otherMediaPerception, data = d)
##
## Standardized Coefficients::
## (Intercept) pDem_Rep pInd_Not
## 0.0000000 -0.2431231 0.1568096
## foxPerception otherMediaPerception
## -0.1207183 0.3408677##
## Call:
## lm(formula = voteLegit ~ (pDem_Rep + pInd_Not), data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.1679 -0.8920 0.1080 0.8321 2.1080
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.34447 0.03645 91.749 < 2e-16 ***
## pDem_Rep -1.27585 0.07156 -17.828 < 2e-16 ***
## pInd_Not 0.56202 0.09058 6.205 7.52e-10 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.138 on 1204 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.2326, Adjusted R-squared: 0.2313
## F-statistic: 182.4 on 2 and 1204 DF, p-value: < 2.2e-16##
## Call:
## lm(formula = voteLegit ~ (pDem_Rep + pInd_Not) + foxPerception +
## otherMediaPerception, data = d)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.3653 -0.7999 0.1392 0.7739 2.8750
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.40004 0.11331 21.181 < 2e-16 ***
## pDem_Rep -0.68869 0.08439 -8.161 8.34e-16 ***
## pInd_Not 0.56222 0.08568 6.562 7.89e-11 ***
## foxPerception -0.12612 0.02736 -4.609 4.47e-06 ***
## otherMediaPerception 0.51439 0.04189 12.278 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.072 on 1202 degrees of freedom
## (29 observations deleted due to missingness)
## Multiple R-squared: 0.3202, Adjusted R-squared: 0.318
## F-statistic: 141.6 on 4 and 1202 DF, p-value: < 2.2e-16B1 <- -0.4504079
B2 <- -0.2431231
SEB1 <- 0.07156
SEB2 <- 0.08439
(Z <- (B1 - B2)/(sqrt((SEB1)^2 + (SEB2^2))))## [1] -1.873408d. Does ID strength moderate?
e. Emotions?
# Moderator model
vote.media.party.int.emo.mod <- lm(voteLegit ~ foxPerception.c * otherMediaPerception.c * (pDem_Rep) * positive.c * negative.c, data = d)
# Mediator model
vote.media.party.int.emo.med <- lm(voteLegit ~ foxPerception.c * otherMediaPerception.c * (pDem_Rep) + positive + negative, data = d)9. Plots
Figure 1: Election Legitimacy by Vote Type, Partisanship, and Timing
voteconf <- as.data.frame(cbind(d[d$election_timing != "Pre-election",]$party_factor, d[d$election_timing != "Pre-election",]$ownvote_conf, d[d$election_timing != "Pre-election",]$overallvote_conf))
voteconf <- as.data.frame(cbind(voteconf, d[d$election_timing != "Pre-election",]$election_timing))
names(voteconf) <- c("party_factor", "Own","Nationwide", "election_timing")
voteconf$party_factor <- recode_factor(voteconf$party_factor, `1` = "Democrat", `2` = "Republican", `3` = "Independent")
voteconf$election_timing <- recode_factor(voteconf$election_timing, `During-election` = "Interregnum", `Post-election` = "Declared")
voteconffull.df <- tidyr::gather(voteconf, Vote_Type, Confidence, Own:Nationwide, factor_key=TRUE)
voteconf.df <- voteconffull.df[voteconffull.df$election_timing != "Pre-election",]
voteConf.party_plot <- ggplot(voteconf.df[!is.na(voteconf.df$party_factor),],
aes(x = election_timing, y = Confidence)) +
geom_violin(alpha = .6, aes(fill = party_factor)) +
geom_point(stat = 'summary', fun = 'mean', size = 1) +
geom_path(stat = 'summary', fun = 'mean', aes(group = 1, col = party_factor)) +
stat_summary(fun.data = mean_cl_normal, geom = "errorbar", position = position_dodge(.9), width=.1, fun.args = list(mult = 1)) +
facet_grid(Vote_Type ~ party_factor)
voteConf.party_plot +
scale_fill_manual(values = c('dodgerblue','red3','orchid4')) +
scale_color_manual(values = c('dodgerblue','red3','orchid4')) +
xlab("Election Phase") +
ylab("Vote Legitimacy") +
theme_classic()## Warning: Removed 35 rows containing non-finite values (stat_ydensity).## Warning: Removed 35 rows containing non-finite values (stat_summary).
## Warning: Removed 35 rows containing non-finite values (stat_summary).
## Warning: Removed 35 rows containing non-finite values (stat_summary).
Figure 2: Negative & Positive Emotions by Timing
Affect <- as.data.frame(cbind(d[d$election_timing != "Pre-election",]$party_factor, d[d$election_timing != "Pre-election",]$election_timing, d[d$election_timing != "Pre-election",]$negative, d[d$election_timing != "Pre-election",]$positive))
names(Affect) <- c("party_factor", "election_timing", "negative", "positive")
Affect$party_factor <- recode_factor(Affect$party_factor, `1` = "Democrat", `2` = "Republican", `3` = "Independent")
Affect$election_timing <- recode_factor(Affect$election_timing, `2` = "Interregnum", `3` = "Declared")
Affect.dv <- tidyr::gather(Affect, Emotion_Type, Rating, negative:positive, factor_key=TRUE)
AAffect.party_plot<- ggplot(nffect[.dv!is.na(nAffect.dvparty_factor),],
aes(x = election_timing, y = Rating, fill = party_factor)) +
geom_violin(alpha = .6, aes(fill = party_factor)) +
geom_point(stat = 'summary', fun = 'mean', size = 1) +
geom_path(stat = 'summary', fun = 'mean', aes(group = 1, col = party_factor)) +
stat_summary(fun.data = mean_cl_normal, geom = "errorbar", position = position_dodge(.9), width=.1, fun.args = list(mult = 1)) +
facet_grid(Emotion_Type ~ party_factor)
Affect.party_plot +
scale_fill_manual(values = c('dodgerblue','red3','orchid4')) +
scale_color_manual(values = c('dodgerblue','red3','orchid4')) +
xlab("Election Phase") +
ylab("Average Nffect") +
tscale_y_continuous(breaks = seq(1,7,1)) +
heme_classic()Figure 3: Media Trust/Consumption by Source
# creating df with only media measures, party ID, and election timing
media <- as.data.frame(cbind(d[,c("party_factor",
"election_timing",
"mediaExposure_1",
"mediaExposure_2",
"mediaExposure_3",
"mediaExposure_4",
"mediaExposure_5",
"mediaExposure_6",
"mediaExposure_7",
"mediaExposure_8",
"mediaExposure_9",
"mediaExposure_10",
"mediaExposure_11",
"mediaExposure_12",
"mediaExposure_13",
"mediaExposure_14",
"mediaExposure_15",
"mediaTrust_1",
"mediaTrust_2",
"mediaTrust_3",
"mediaTrust_4",
"mediaTrust_5",
"mediaTrust_6",
"mediaTrust_7",
"mediaTrust_8",
"mediaTrust_9",
"mediaTrust_10",
"mediaTrust_11",
"mediaTrust_12",
"mediaTrust_13",
"mediaTrust_14",
"mediaTrust_15")]))
media[,18:32] <- media[,18:32] + 3
media$media_1 <- rowMeans(media[,c(3,18)], na.rm = T)
media$media_2 <- rowMeans(media[,c(4,19)], na.rm = T)
media$media_3 <- rowMeans(media[,c(5,20)], na.rm = T)
media$media_4 <- rowMeans(media[,c(6,21)], na.rm = T)
media$media_5 <- rowMeans(media[,c(7,22)], na.rm = T)
media$media_6 <- rowMeans(media[,c(8,23)], na.rm = T)
media$media_7 <- rowMeans(media[,c(9,24)], na.rm = T)
media$media_8 <- rowMeans(media[,c(10,25)], na.rm = T)
media$media_9 <- rowMeans(media[,c(11,26)], na.rm = T)
media$media_10 <- rowMeans(media[,c(12,27)], na.rm = T)
media$media_11 <- rowMeans(media[,c(13,28)], na.rm = T)
media$media_12 <- rowMeans(media[,c(14,29)], na.rm = T)
media$media_13 <- rowMeans(media[,c(15,30)], na.rm = T)
media$media_14 <- rowMeans(media[,c(16,31)], na.rm = T)
media$media_15 <- rowMeans(media[,c(17,32)], na.rm = T)
media.df <- tidyr::gather(media, Source, Perception, media_1:media_15, factor_key=TRUE)
media.df$Source <- recode_factor(media.df$Source, "media_1" = "NYTimes", "media_2" = "WSJ", "media_3" = "WashPost", "media_4" = "USAToday", "media_5" = "Fox News", "media_6" = "CNN", "media_7" = "MSNBC", "media_8" = "Yahoo", "media_9" = "HuffPost", "media_10" = "AOL", "media_11" = "NPR", "media_12" = "ABC", "media_13" = "NBC", "media_14" = "CBS", "media_15" = "PBS")
media.df$Source <- factor(media.df$Source, levels = c("Fox News", "ABC", "AOL", "CBS", "CNN", "HuffPost", "MSNBC", "NBC", "NPR", "NYTimes", "PBS", "USAToday", "WashPost", "WSJ", "Yahoo"))
media_plot <- ggplot(media.df[!is.na(media.df$party_factor),],
aes(x = Source, y = Perception, fill = party_factor)) +
geom_bar(stat = 'summary', fun = 'mean', position = position_dodge(.9), alpha = .9) +
stat_summary(fun.data = mean_cl_normal, geom = "errorbar", position = position_dodge(.9), width=.1, fun.args = list(mult = 1))
media_plot +
scale_fill_manual("Participant Partisan ID", values = c("dodgerblue", "red3", "orchid4")) +
xlab("Media Source") +
ylab("Consumption of/Trust in Media Source") +
coord_cartesian(ylim = c(1,5)) +
theme_blank() +
theme(axis.text.x = element_text(angle = 45, hjust = 1)) ## Warning: Removed 180 rows containing non-finite values (stat_summary).
## Warning: Removed 180 rows containing non-finite values (stat_summary).