|
pubmed:abstractText |
A methodological experiment was conducted as part of the Demographic and Health Surveys (DHS) project to determine whether different approaches to measuring the same variables would yield similar results. The experiment consisted of the use of a new questionnaire, incorporating variants of many of the traditional approaches to the collection of demographic and health data. The experimental questionnaire was applied in 2 countries, Peru and the Dominican Republic at the same time as the standard survey. The most salient feature of this new questionnaire pregnancies, contraceptive practice, and postpartum, marriage, employment, and migration information. In this paper, the authors evaluate whether the use of the calendar in the experimental questionnaire improved the quality of the resulting information on contraception in the DHS surveys from the abovementioned countries. Specifically, the author determine whether the monthly calendar led to better estimates of contraceptive prevalence, failure, and discontinuation, and discuss the advantages and drawbacks of the calendar over the more standard approach. Results from this evaluation suggest that the major advantages of the calendar over the more standard questionnaire are 3-fold: it allows for more complete reports of use for periods prior to the survey; it allows for a detailed study of contraceptive use patterns; and it obtains information which is more internally consistent with other types of information. On the bases of these findings, the DHS project will include a calendar in the new standard questionnaire to be used in the 2nd round of surveys (DHS II).
|
