A COMPUTER-GENERATED SAMPLING TABLE FOR SELECTING RESPONDENTS WITHIN HOUSEHOLDS.
It is desirable to select a respondent within each household by the least complicated, least time-consuming, and least rapport-straining method available. The technique developed by Carter and Troldahi meets each of these conditions very well, but more than the four original versions of the sampling...
| Publicado en: | Public Opinion Quarterly Vol. 29; no. 3; pp. 431 - 437 |
|---|---|
| Autores principales: | , |
| Formato: | Artículo |
| Publicado: |
Oxford University Press / USA
Fall65
|
| Materias: | |
| Acceso en línea: | Ver este registro en EBSCOhost |
| Sumario: | It is desirable to select a respondent within each household by the least complicated, least time-consuming, and least rapport-straining method available. The technique developed by Carter and Troldahi meets each of these conditions very well, but more than the four original versions of the sampling table are required if selection error is to be minimized. The "Stanford modification" of the Carter-Troldahi technique uses the computer to circumvent the question of practicality (which dictated that only four versions be prepared) and to permit the preparation of as many versions as are needed-sixty. Be- sides providing an excellent sampling table for use in telephone and doorstep interviews, this project illustrates one way in which the computer may step out of its role as data analyst and help at the other end of a survey as well. |
|---|