| Datasets for Statistical Analysis Home | Techniques Index | Application Index |
social, health;
generalized linear models
The data record details about the Birth to Ten study (BTT), perfomed in the greater Johannesburg/Soweto metropolitan area of South Africa during 1990. In the study, all mothers of singleton births were interviewed during a seven-week period between April and June to women with permanent addresses in a defined area (a total of 4019 births). Five years later, 964 of these mothers were re-interviewed. If the mothers interviewed later and representative of the original populations, the two groups should show similar characteristics. Of of those characteristics is documented here: the proportion with and without medical aid.
There are eight observations on four variables.
| Counts | The number of subjects in the given classification |
| Group | Which group the mother belonged to: 1 refers to the mothers not followed up after the five years; 2 refers to the mothers followed-up five years later. |
| MedicalAid | Whether or not the mother had medical aid; either Yes or No |
| Race | The mothers' race; either White or Black |
There are no missing values.
Morrell, Christopher H. (1999). Simpson's Paradox: An Example From a Longitudinal Study in South Africa Journal of Statistics Education, 7(3).
The data are an example of Simpson's paradox.