Sampling Procedure
The sample for tile TDHS-98 was designed to provide estimates of population and health indicators including fertility and mortality rates for the nation as a who/e, for urban and rural areas, and for tile five major regions of tile country (West, South, Central, North and East). A weighted, multi-stage, stratified cluster sampling approach was used in tile selection of the TDHS-98 sample.
The optimal distribution with a target sample size of I0,000 selected households was based on the provisional results of the 1997 General Population Count. Selection of the TDHS-98 sample was undertaken in three stages. Tile sampling units at tile first stage were tile settlements stratified by population size. The ti'ame for the selection of the primary sampling units (PSU) was prepared using the provisional results of the 1997 Population Count. The fi'ame was divided into two groups, one including those settlements with populations of more than 10,000 and the other including settlements with populations less than 10,000. The selection of the settlement in each group was carried out with probability proportional to size (1997 poptdatiou).
The second stage of selection required the selection of the assigned nnmber of clusters in each selected settlement. For the majority of the settlements (340 clusters), the selection of clusters was based on the household lists that were available from the 1995 Structure Schedules. The State Institute of Statistics (SIS) selected the clusters and provided to Hacettepe Institute of Population Studies a description of each selected cluster. Each cluster included approximately 100 households. For those settlements where SIS was not able to provide information (140 clusters), the lists of households were prepared in the field.
Following the selection of the secondary sampling units (SSUs), a household listing was prepared or updated for each SSU by the TDHS-98 listing teams. Using the household lists, a systematic random sample of fixed number of households (25 in clusters located in settlements over 10,000 and 15 in those less than 10,000) was chosen within each cluster for the TDHS-98. All women at ages 15-49 who were present in the household on the night before the interview were eligible for the survey. In half of the selected households for women interview, husbands of currently married eligible women, who were present in the household on the night before the interview or who usually lived in the household were eligible for the husband survey.
SAMPLE FRAME
Different criteria have been used to describe "urban" and "rural" settlements in Turkey. In the demographic surveys of the 1970s a population size of 2,000 was used to differentiate between urban and rural settlements. In the 1980s, this was increased to 10,00O and, in some surveys in the 1990s, to 20,000. A number of surveys used the administrative status of settlements in combination with population size for the purpose of differentiation.
The urban frame of the 1998 TDHS consisted of a list of provincial centres, district centres, and other settlements with populations larger than 10,000, regardless of administrative status. In turn, the rural frame consists of all district centres, subdistricts and villages not included iF the urban fi'ame. Initial information on these settlements was obtained from the preliminary results of 1997 Population Count. The preliminary results of 1997 Population Count provided a computerized list of all settlements (provincial and district centres, , subdistricts and villages) and their population. The population counts were taken from the cumulative enumeration forms for settlements, which were filled by supervisors during the Population Count.
STRATIFICATION
Currently Turkey is divided administratively into 80 provinces. This figure was 67 for a long time, with new provinces formed since the late 1980s, For purposes of selection in prior surveys in Turkey, these provinces have been grouped into five regions, as described in Chapter 1. This regional breakdown has been popularised as a powerful variable for understanding the demographic, social, cultural, and economic differences between different parts of the country. The five regions, West, South, Central, North, and East regions, include varying numbers of provinces.
One of tile priorities of the TDHS was to produce a sample design that was methodologically and conceptually consistent with the designs of previous demographic surveys carried out by the Hacettepe Institute of Population Studies. In surveys prior to the 1993, the five-region division of the country was used for stratification. In the 1993 TDHS, a more detailed stratification taking into account subregions was employed to obtain a better dispersion of file sample. The criteria for subdividing the five major regions into subregions were the infant mortality rates &each province, estimated from the 1990 Population Census using indirect techniques? Using the infant mortality estimates as well as geographic proximity, the provinces in each region were grouped into 14 subregions at the time of the 1993 TDHS. The sub-regional division developed during the 1993 TDHS was used in the 1998 survey.
SAMPLE ALLOCATION
The target sample size of 10,000 households was allocated among the five major divisions using the sampling error estimates from the TDHS-93 in combination with the power allocation technique with the ex- pectation that the target sample size would provide about 8,000 completed individual interviews. During the power allocation calculations, the aim was to keep the allocation as similar as possible to the 1993 TDHS. The optimal distribution (with power 0.4) among the five major regions is shown in Table B.I. For purposes of comparison, Table B.I also shows the allocation of the TDHS-93 sample and the allocation if the TDHS-98 sample had been distributed proportional to the size of the population in each region. To have an adequate representation of clusters within each of the five major regions, it was decided to select 25 households per standard urban segments (each consisting of 100 households) and 15 households per standard rural segment. It was also determined that 70 percent of the 10,000 households would be located in urban settlements and 30 percent in rural settlements.
SAMPLE SELECTION - SELECTION PROCEDURES
The lists of settlements of urban settlements (settlements with I 0,000 or more population) and rural settlements (settlements with less than 10,000 population) constituted the frame for the first stage of the sample selection. For tile selection of the first-stage sample, settlements were grouped within each of the 14 subregions, and a systematic random sample of settlements with probability proportional to size (PPS) based on file preliminary 1997 Population Count was selected from the settlement lists. The output from this first stage of the selection was a list of all of the settlements included in the 1998 TDHS sample along with the number of clusters to be drawn from each settlement.
In Turkey, settlements are not divided into small areal units with well-defined boundaries (e.g., census enumeration areas) that can be used for conducting surveys. For some settlements, however, household lists were available from the Structure Schedules that were prepared in 1995 by many municipalities in collaboration with the State Institute of Statistics (SIS). Household lists from the Structure Schedules were available for settlements from which 340 clusters in the TDHS-98 sample were to be drawn. For those settlements, the household lists were subdivided into segments of approximately 100 households. The list of these segments constituted the frame for the selection of the 340 clusters. For each of the selected clusters, SIS provided a list of the dwellings units with their full addresses (quarter, area, avenue/street, building and door number).
SIS was not able to provide a frame from the 1995 Structure Schedules for settlements from which I40 clusters were to be drawn for the TDHS-98. For these settlements, the list of households had to be prepared in file field. In the case of small settlements (less than 250 households), the entire settlement was listed. In the ease of the small number of settlements in which there were more than 250 households, 200 households were listed and an estimate of the remaining number of households in the settlement was obtained through a quick count.
LISTING AND MAPPING ACTIVITIES
Although tile SIS had dwelling lists for many clusters, they did not have the corresponding maps. For this reason, the selected clusters were formed with streets that were not always adjacent to each other. Moreover, the lists provided by the SIS did not reflect changes that may have occurred during the period from the 1995 to tile survey date. Two types of changes were possible: those that could be updated during listing, such as the construction of a new building oil the street, a change in the use of a building (e.g., a fiat can be used as an office instead of a dwelling), or changes in the names of streets, and those that were more problematic, e.g., the appearance of new quarters in urban centres.
In an effort to develop strategies for dealing with these as well as other possible problems that might arise, a pilot listing activity was undertaken in the capital, Ankara, before the actual listing activity began. The final listing forms, sketch map formats, and listing and mapping manuals were developed based on this experience.
Forty (40) university students were trained for the main listing activity. Listing teams were formed following a four-day training program in the beginning of June 1998. Each team was provided with maps describing the location of the settlements they were expected to visit as well as other materials needed for the listing. Sixteenth (16) listing teams were constituted" with" one mapper and one lister. The listing operation started on 8 of June. It was carried under the supervision of the research assistants and regional coordinators from the Hacettepe Institute of Population Studies.
The cluster (standard segment) size was around 100 households for most of the clusters in urban areas. Only two urban clusters had extremely low numbers of households; in order to obtain 100 households in these clusters, adjacent streets were added to the original cluster. In some of the selected villages, the total populations also were small, and, therefore, the original cluster did not include 100 households. In these cases, the village that was nearest to the selected village was included in the sample, and the names of these villages were provided to the listing teams; the lists of 100 households were completed from the two villages.
Most of the listing activity was completed before the training for the main fieldwork began in July. Overall, the quality of the listing work produced by the listers was good although it varied somewhat largely in response to problems the listing teams experienced in working in some geographic areas. In particular, there were some problems with the listing of clusters of Adana province where there had been an earthquake. There also were problems with the lists for Içel province. Finally, three clusters were not listed due to problems of accessibility.