This is the dataset used in Gelman & Hill's book, Data Analysis Using Regression and Multilevel/Hierarchical Models. They are polling data on the presidential election from 1988, collected one week before the election. You can download all the data from the book here: http://www.stat.columbia.edu/~gelman/arm/examples/ARM_Data.zip. Note that this is only a few of the variables from the original data supplied with the book.

polls

Format

A tibble with 13,544 rows and 7 variables:

state

An integer denoting the state identification number.

edu

An ordered factor stating the education level of the respondent

age

An unordered factor stating the age of range of the respondent

female

A dummy variable (integer) denoting whether the respondent was coded as male (female = 0) or female (female = 1).

black

A dummy variable (integer) denoting whether the respondent was coded as Black (black = 1) or not Black (black = 0).

weight

A sampling weight

bush

Whether the respondent stated they were in favor of voting for George Bush Sr.