| Datasets for Statistical Analysis Home | Techniques Index | Application Index |
In the first stage, he randomly selected 7 of the 49 collector districts (CDs) in the West and North West Toowoomba SLAs. (CDs are the smallest geographic divisions used by the Australian Bureau of Statistics. There is an average of about 225 households per CD.) In the second stage, he took a 1 in 2 systematic random sample of households in each of the selected CDs. Specifically, he selected the starting point at random and took every second household from there.
The interviewer visited each target household up to a maximum of 4 attempts. If the household had not been contacted and interviewed after these four attempts no further visits were attempted.
Paul's main question is whether a maximum of four call-backs was sufficient. If he had insisted on a fifth call back to households who had not yet been contacted, how much would the success rate improve for the survey?
There are 1384 observations on 5 variables.
| CD | The collector distruct (see Description above) |
| VISITNUM | The number of the visit to the household (an integer from 1 to 4) |
| WEND | The day the household was visited: 1 means a weekend; 2 means a weekday |
| TIME | The time the household was visited; 1 means before 11:00; 1 means before 11:00; 2 means between 11:00 and 11:59; 3 means between 12:00 and 12:59; 4 means between 13:00 and 13:00; 5 means between 14:00 and 14:59; 6 means between 15:00 and 15:59; 7 means after 15:59. |
| SUCCESS | records the outcome of the visit: 1 means the interview was completed; 2 means the interview was not completed (usually, there was no-one home) |