Proximity Mapping for Multivariate data
Proximity mapping is a visualization technique that is used to reduce the dimensionality of multivariate data and to display relationships among objects (cases, items, or other entities) in a spatial configuration. The data to be analyzed includes variables that either represent a proximity matrix (or matrices), or represent multivariate data that is converted into proximity matrix or matrices.
This case study illustrates the use of PROXMAP for analyzing multivariate data. The study uses
the sample file states9_labels_num.sav, containing a well-known dataset that
involves the 50 states in the United States. The original dataset was introduced by Wainer and
Thissen (1981), who reexamined Angoff and Mencken’s (1931) search for the Worst American State by
using seven social indicators. Meulman (1984) later extended the dataset by adding an eighth
variable: the percentage of students who fail a nationwide achievement test, which is taken from
Walberg and Rasher (1977).
For the current case study, a categorical variable is further included that indicates the region (US Census division) that each state belongs to.
| Variable | Description |
|---|---|
| POPUL | 1975 population (in thousands) |
| INCOME | Per capita income (in US dollars) |
| LIFE | Life expectancy (in years) |
| SCHOOL | Percentage of population over age 25 with a high school diploma |
| ILLIT | Illiteracy rate (percentage of the population) |
| FAIL | Percentage of the population who fails in a nationwide test |
| HOMIC | 1976 homicide and nonnegligent manslaughter rate (per 1,000) |
| FREEZE | Average number of days per year with sub-zero temperatures (°C) |
| REGION | Categorical classification into one of nine US regions (Census division) |
| STATE | Name of each US state (used as object label) |
Go to .