glucose: Glucose control

In ggm: Graphical Markov Models with Mixed Graphs

Glucose control

Description

Data on glucose control of diabetes patients.

Usage

data(glucose)

Format

A data frame with 68 observations on the following 8 variables.

Y

a numeric vector, Glucose control (glycosylated haemoglobin), values up to about 7 or 8 indicate good glucose control.

X

a numeric vector, a score for knowledge about the illness.

Z

a numeric vector, a score for fatalistic externality (mere chance determines what occurs).

U

a numeric vector, a score for social externality (powerful others are responsible).

V

a numeric vector, a score for internality (the patient is him or herself responsible).

W

a numeric vector, duration of the illness in years.

A

a numeric vector, level of education, with levels -1: at least 13 years of formal schooling, 1: less then 13 years.

B

a numeric vector, gender with levels -1: females, 1: males.

Details

Data on 68 patients with fewer than 25 years of diabetes. They were collected at the University of Mainz to identify psychological and socio-economic variables possibly important for glucose control, when patients choose the appropriate dose of treatment depending on the level of blood glucose measured several times per day.

The variable of primary interest is Y, glucose control, measured by glycosylated haemoglobin. X, knowledge about the illness, is a response of secondary interest. Variables Z, U and V measure patients' type of attribution, called fatalistic externality, social externality and internality. These are intermediate variables. Background variables are W, the duration of the illness, A the duration of formal schooling and B, gender. The background variables A and B are binary variables with coding -1, 1.

Source

Cox & Wermuth (1996), p. 229.

References

Cox, D. R. & Wermuth, N. (1996). Multivariate dependencies. London: Chapman & Hall.

Examples

data(glucose)
## See Cox & Wermuth (1996), Figure 6.3 p. 140
coplot(Y ~ W | A, data=glucose)

ggm documentation built on May 29, 2024, 7:27 a.m.