A SAMPLE DESIGN APPENDIX A.1

SAMPLE DESIGN
APPENDIX
A
A. Aliaga
A.1
SAMPLE DESIGN
The sample for the 2000 TDHS was designed to allow statistical analysis at the national
level, for urban and rural areas, and for the six regions of the country (Ashgabad City, Akhal,
Balkan, Dashoguz, Lebap, and Mary).
The sample design was specified in terms of a target number of households in the six
regions of Turkmenistan. The overall target number of households was set at 6,800. This number
was allocated to the regions as follows: 800 to Ashgabad City, 1,000 to each of 4 regions (Akhal,
Balkan, Lebap and Mary) and 2,000 to the remaining region (Dashoguz), for which more intensive
analysis was desired (see Table A.1).
The six regions of the country were further stratified into urban areas (cities, towns and
small settlements) and rural areas (villages). The sampling frame consisted of the list of standard
segments. Each standard segment was created on the basis of contiguous blocks that have clear
boundaries—coinciding to the extent possible with census supervisor areas—and have between 200
and 500 households according to measures of size estimated by projection from to the 1995 Census
data.
Table A.1 Sample allocation
Distribution of standard segments in the DHS sample by region and urban-rural
residence, Turkmenistan 2000
______________________________________________________________________
Number of
Number of Number of
Target
selected
urban
rural
number of
standard
standard
standard
Region
households
segments
segments
segments
_______________________________________________________________________
Ashgabad
Akhal
Balkan
Dashoguz
Lebap
Mary
800
1000
1000
2000
1000
1000
34
35
35
57
35
35
34
11
27
22
15
9
0
24
8
35
20
26
Total
6800
231
118
113
Appendix A * 165
A.2
SAMPLE SELECTION
The sample was designed as a two-stage probability sample. Within regions the sample was
to be self-weighting. The first stage involved the selection of standard segments (PSUs) by
systematic sampling with probability proportional to size. This resulted in the selection of 231
standard segments:118 in urban areas and 113 in rural areas. A household listing operation was
conducted in each selected standard segment. In the second stage, households were selected with
probability proportional to the inverse of the first stage selection probability. On average, the
number of households selected per standard segment was 28.
Since the sample for each of the six survey regions was self-weighting, the sampling fraction
for each region was an important design parameter. The sampling fractions were estimated with
projected census figures. The weighting factors for the six survey regions are inversely proportional
to the sampling fractions.
A.3
SAMPLE IMPLEMENTATION
Implementation of the sample design resulted in the selection of 6,850 households. The
data on household membership and age collected in the Household Questionnaire identified 8,250
women eligible for the Women’s Questionnaire (i.e., women age 15-49 who were usual household
members or who stayed in the household the night before the interviewer’s visit) (Table A.2).
From the 6,850 selected households, 6,391 were identified as current households and
household interviews were completed in 6,302. This yields a household response rate of
98.6 percent. Of the 8,250 women who were eligible respondents, a total of 7,919 were
interviewed. This yields an eligible woman response rate of 96.0 percent.
The overall response rate (94.7 percent) is the product of the household response rate and
the eligible woman response rate. The overall response rate varies by region from 85.6 percent in
Ashgabad City to 97.4 percent in the Balkan Region.
166
* Appendix A
A.2 Sample implementation
Percent distribution of households and eligible women in the DHS sample by results of the interview and household,
eligible woman, and overall response rates, according to sample domain and urban-rural residence, Turkmenistan 2000
_____________________________________________________________________________________________________
Result of
Region
Residence
interview and
________________________________________________ _____________
response rate
Ashgabad Akhal Balkan Dashoguz Lebap Mary
Urban
Rural
Total
______________________________________________________________________________________________________
Selected households
Completed (C)
Household present but
no competent respondent
at home (HP)
Postponed (P)
Refused (R)
Dwelling not found (DNF)
Household absent (HA)
Dwelling vacant/address
not a dwelling (DV)
Dwelling destroyed (DD)
Other (O)
80.5
2.1
89.2
0.3
91.8
0.9
95.6
0.1
94.4
0.3
93.9
0.1
88.9
0.8
95.7
0.1
92.0
0.5
0.3
3.3
0.4
8.6
0.0
0.3
0.9
4.8
0.0
0.0
0.1
4.0
0.0
0.2
0.0
2.3
0.0
0.2
0.0
2.7
0.1
0.3
0.0
3.5
0.1
0.8
0.2
5.5
0.0
0.2
0.2
1.9
0.0
0.6
0.2
3.9
4.4
0.5
0.0
4.2
0.2
0.0
2.8
0.2
0.1
1.6
0.1
0.1
1.8
0.6
0.0
2.0
0.0
0.1
3.5
0.3
0.0
1.5
0.2
0.1
2.5
0.2
0.1
Total
Number of households
100.0
780
100.0
956
100.0
992
100.0
2,032
100.0
1,040
100.0
1,050
100.0
3,688
100.0
3,162
100.0
6,850
Household response
1
rate (HRR)
93.0
98.3
98.9
99.7
99.5
99.5
97.9
99.4
98.6
Eligible women
Completed (EWC)
Not at home (EWNH)
Postponed (EWP)
Refused (EWR)
Partly completed (EWPC)
Incapacitated
Other
92.0
4.6
0.0
2.2
0.0
0.8
0.5
93.8
5.4
0.0
0.1
0.0
0.8
0.0
98.4
0.6
0.0
0.0
0.0
1.0
0.0
96.9
2.1
0.0
0.2
0.0
0.8
0.0
96.8
2.0
0.1
0.2
0.1
0.9
0.0
95.2
3.4
0.0
0.7
0.1
0.7
0.0
96.3
2.3
0.0
0.5
0.0
0.8
0.1
95.7
3.1
0.0
0.2
0.0
0.8
0.0
96.0
2.7
0.0
0.4
0.0
0.8
0.0
100.0
636
100.0
1,153
100.0
1,016
100.0
2,925
100.0
1,305
100.0
1,215
100.0
3,836
100.0
4,414
100.0
8,250
92.0
93.8
98.4
96.9
96.8
95.2
96.3
95.7
96.0
Total
Number of women
Eligible woman response
2
rate (EWRR)
Overall response
3
rate (ORR)
85.6
92.1
97.4
96.6
96.3
94.7
94.3
95.2
94.7
__________________________________________________________________________________________________
Note: The household response rate is calculated for completed households as a proportion of completed, no competent
respondent, refused, and dwelling not found. The eligible woman response rate is calculated for completed interviews as
a proportion of completed, not at home, postponed, refused, partially completed, incapacitated and "other." The overall
response rate is the product of the household and eligible woman response rates.
1
Using the number of households falling into specific response categories, the household response rate (HRR) is
calculated as:
C
_______________________ * 100
C + HP +P+ R + DNF
2
Using the number of eligible women falling into specific response categories, the eligible woman response rate
(EWRR) is calculated as:
EWC
___________________________________ * 100
EWC + EWNH + EWR + EWI + EWO
3
The overall response rate (ORR) is calculated as:
ORR = (HRR * EWRR) ÷ 100
Appendix A * 167