Demographic and Health Surveys Methodology - Measure DHS

At data entry time, contrary to the first type of multiple response questions, all responses are stored in justone field. This makes data entry easy to control, and the data entry operator just type the letters that arecircled. However, dealing with alphabetic codes presents a challenge for analysis purposes. To solve thisproblem, each response for this type of question gets translated into codes “0–No, 1–Yes, 9–Missing” inthe recode as follows:• M2A—Prenatal care: Doctor• M2B—Prenatal care: Nurse/Midwife• M2C— Prenatal care: Auxiliary Midwife• M2D— Prenatal care: CS Health Specialist• M2E—Prenatal care: CS Health Specialist• M2F—Prenatal care: Trained Birth Attendant• M2G—Prenatal care: Traditional Birth Attendant• M2H—Prenatal care: Relative• M2I—Prenatal care: CS Other Person• M2J—Prenatal care: CS Other Person• M2K—Prenatal care: Other Response (uncoded)• M2L—Prenatal care: CS Other• M2N—Prenatal care: No OneThere are several issues that need to be taken into consideration with this type of variable:1. If the question was missing, all applicable variables will get assigned code missing.2. If a response does not exist for a particular country, the variable will be left coded BLANK.3. There are always provisions to include country-specific (CS) responses that are not part of thecore questionnaire. For example, some countries include “health workers” as a possible response.Sampling WeightsDHS users should be aware that, in many cases, the data must be weighted. The following describes howDHS weights are constructed and when they should be used.DefinitionSampling weights are adjustment factors applied to each case in tabulations to adjust for differences inprobability of selection and interview between cases in a sample, either due to design or happenstance. Inthe DHS surveys, many times the sample is selected with unequal probability to expand the number ofcases available (and hence reduce sample variability) for certain areas or subgroups for which statisticsare needed. In this case, weights need to be applied when tabulations are made of statistics to produce theproper representation. When weights are calculated because of sample design, corrections for differentialresponse rates are also made.Guide to DHS Statistics 12 September 2003