BushApproval: U.S. President George Bush approval ratings

In jverzani/UsingR: Data Sets, Etc. for the Text "Using R for Introductory Statistics", Second Edition

Description

Approval ratings as reported by six different polls.

Usage

data(BushApproval)

Format

A data frame with 323 observations on the following 3 variables.

date

The date poll was begun (some take a few days)

approval

a numeric number between 0 and 100

who

a factor with levels fox gallup newsweek time.cnn upenn zogby

Details

A data set of approval ratings of George Bush over the time of his presidency, as reported by several agencies. Most polls were of size approximately 1,000 so the margin of error is about 3 percentage points.

Source

This data was found at http://www.pollingreport.com/BushJob.htm. The idea came from an article in Salon http://salon.com/opinion/feature/2004/02/09/bush_approval/index.html by James K. Galbraith.

Examples

data(BushApproval)
attach(BushApproval)

## Plot data with confidence intervals. Each poll gets different line type
## no points at first
plot(strptime(date,"%m/%d/%y"),approval,type="n",
     ylab = "Approval Rating",xlab="Date",
     ylim=c(30,100)
     )

## plot line for CI. Margin or error about 3
## matlines has trouble with dates from strptime()
colors = rainbow(6)

for(i in 1:nrow(BushApproval)) {
  lines(rep(strptime(date[i],"%m/%d/%y"),2),
        c(approval[i]-3,approval[i]+3),
        lty=as.numeric(who[i]),
        col=colors[as.numeric(who[i])]
        )
  
}

## plot points
points(strptime(date,"%m/%d/%y"),approval,pch=as.numeric(who))

## add legend
legend((2003-1970)*365*24*60*60,90,legend=as.character(levels(who)),lty=1:6,col=1:6)
detach(BushApproval)

jverzani/UsingR documentation built on Aug. 3, 2020, 11:57 a.m.