NAME:  Cigarette data for an introduction to multiple regression
TYPE:  A sample of 25 brands of cigarettes
SIZE:  25 observations, 5 variables

DESCRIPTIVE ABSTRACT:
Measurements of weight and tar, nicotine, and carbon monoxide content
are given for 25 brands of domestic cigarettes.

SOURCES: 
Mendenhall, William, and Sincich, Terry (1992), _Statistics for
Engineering and the Sciences_ (3rd ed.), New York: Dellen Publishing
Co. (ISBN: 0 02380552 8)  
(Original source: Federal Trade Commission, USA)

VARIABLE DESCRIPTIONS: 
Brand name 
Tar content (mg) 
Nicotine content (mg) 
Weight (g)
Carbon monoxide content (mg)

Values are delimited by blanks.  There are no missing values.

SPECIAL NOTES:
Observation 3 (Bull Durham) is an outlying point.

STORY BEHIND THE DATA:
The Federal Trade Commission annually rates varieties of domestic
cigarettes according to their tar, nicotine, and carbon monoxide
content.  The United States Surgeon General considers each of these
substances hazardous to a smoker's health.  Past studies have shown
that increases in the tar and nicotine content of a cigarette are
accompanied by an increase in the carbon monoxide emitted from the
cigarette smoke.

The data presented here are taken from Mendenhall and Sincich (1992)
and are a subset of the data produced by the Federal Trade Commission.

For more information, see the article "Using Cigarette Data for an
Introduction to Multiple Regression" by Lauren McIntyre in Volume 2,
Number 1, of the _Journal of Statistics Education_.  

PEDAGOGICAL NOTES:
The dataset presented here contains measurements of weight and tar,
nicotine, and carbon monoxide (CO) content for 25 brands of
cigarettes.  Students familiar with simple linear regression can use
these data to develop an understanding of multiple regression
techniques.  Students will discover that there is an outlier in this
dataset (where we define an outlier as a point not near the rest of the
data) and that tar and nicotine are collinear variables.  These
characteristics of the data can lead to a very good discussion and an
enhanced understanding of regression when introduced carefully in
class.

We use the dataset as part of a computer demonstration.  We lead the
students into a discussion of the outlying point, collinearity, and
multiple regression, using a question and answer format.  The article
by Lauren McIntyre presents a summary of the data presentation.

SUBMITTED BY:
Lauren McIntyre
Department of Statistics, Box 8203
North Carolina State University
Raleigh, NC 27695-8203
mcintyre@stat.ncsu.edu