ex1 <- read.table("lecture3ex1.csv", sep=",", header=TRUE) library(psych) describe(ex1) corr.test(ex1) out1 <- lm(jobsat ~ tlleader + security, data=ex1) summary(out1) out1a <- lm(jobsat ~ tlleader*security, data=ex1) summary(out1a) anova(out1, out1a) out1_1 <- lm(jobsat ~ I(tlleader - mean(tlleader)) + I(security - mean(security)), data=ex1) summary(out1_1) out1a_1 <- lm(jobsat ~ I(tlleader - mean(tlleader))*I(security - mean(security)), data=ex1) summary(out1a_1) anova(out1_1, out1a_1) out1_2 <- lm(I(scale(jobsat)) ~ I(scale(tlleader)) + I(scale(security)), data=ex1) summary(out1_2) out1a_2 <- lm(I(scale(jobsat)) ~ I(scale(tlleader))*I(scale(security)), data=ex1) summary(out1a_2) library(interactions) interact_plot(out1a, pred="tlleader", modx="security") interact_plot(out1a, pred="tlleader", modx="security", modx.values = c(30, 40, 50, 60, 70), plot.points=TRUE) interact_plot(out1a, pred="tlleader", modx="security", interval=TRUE, int.width=.9, x.label = "Transformational Leadership", y.label = "Job Satisfaction", main.title = NULL, legend.main = "Job Security", colors = "seagreen") sim_slopes(out1a, pred="tlleader", modx="security") sim_slopes(out1a, pred="tlleader", modx="security", cond.int=TRUE) ss <- sim_slopes(out1a, pred="tlleader", modx="security", modx.values = seq(30, 75, 5)) plot(ss) johnson_neyman(out1a, pred="tlleader", modx="security", alpha = .05) ex2 <- read.table("lecture3ex2.csv", sep=",", header=TRUE) describe(ex2) corr.test(ex2) out2 <- lm(partlike ~ age + ses + guarlike, data=ex2) summary(out2) out2a <- lm(partlike ~ age*ses + guarlike, data=ex2) summary(out2a) anova(out2, out2a) sim_slopes(out2a, pred="ses", modx="age") interact_plot(out2a, pred="ses", modx="age") ex3 <- read.table("lecture3ex3.csv", sep=",", header=TRUE) describe(ex3) corr.test(ex3[,c("latemins", "angry")]) table(ex3[,"country"]) ex3[,"country"] <- relevel(ex3[,"country"], ref="thai") out3 <- lm(angry ~ latemins + country, data=ex3) summary(out3) out3a <- lm(angry ~ latemins*country, data=ex3) summary(out3a) anova(out3, out3a) sim_slopes(out3a, pred="latemins", modx="country") interact_plot(out3a, pred="latemins", modx="country") out3a <- lm(angry ~ latemins*country, data=ex3) out3ax <- lm(angry ~ latemins + latemins:country, data=ex3) anova(out3ax, out3a) summary(out3a) out3a3 <- lm(angry ~ I(latemins - 3)*country, data=ex3) out3a3x <- lm(angry ~ I(latemins - 3) + I(latemins - 3):country, data=ex3) anova(out3a3x, out3a3) summary(out3a3) out3a6 <- lm(angry ~ I(latemins - 6)*country, data=ex3) out3a6x <- lm(angry ~ I(latemins - 6) + I(latemins - 6):country, data=ex3) anova(out3a6x, out3a6) summary(out3a6) out3a9 <- lm(angry ~ I(latemins - 9)*country, data=ex3) out3a9x <- lm(angry ~ I(latemins - 9) + I(latemins - 9):country, data=ex3) anova(out3a9x, out3a9) summary(out3a9) out3a12 <- lm(angry ~ I(latemins - 12)*country, data=ex3) out3a12x <- lm(angry ~ I(latemins - 12) + I(latemins - 12):country, data=ex3) anova(out3a12x, out3a12) summary(out3a12) ex4 <- read.table("lecture3ex4.csv", sep=",", header=TRUE) describe(ex4) corr.test(ex4) out4 <- lm(cigarette ~ fearprob, data=ex4) summary(out4) out4a <- lm(cigarette ~ fearprob + I(fearprob^2), data=ex4) summary(out4a) anova(out4, out4a) with(ex4, plot(fearprob, cigarette, xlab="Probability of the Fear Advertisement", ylab="Number of Cigarettes Smoked")) myorder <- order(ex4[,"fearprob"]) yhat <- predict(out4a) lines(ex4[myorder,"fearprob"], yhat[myorder]) # Probing Instantaneous rate of change out4a2 <- lm(cigarette ~ I(fearprob - 0.2) + I((fearprob - 0.2)^2), data=ex4) summary(out4a2) out4a7 <- lm(cigarette ~ I(fearprob - 0.7) + I((fearprob - 0.7)^2), data=ex4) summary(out4a7) plot(out3) library(car) outlierTest(out3) qqPlot(out3) leveragePlots(out3)