| Weighting [message #28019] |
Thu, 02 November 2023 09:58  |
lmr
Messages: 3
Registered: November 2023
|
Member |
|
|
Can I ask for help, to weight the 2007 IDHS data using strata, so what number of variables that shown strata? It the variable 022, but it seems impossible because the numbers listed are in the hundreds? This number is very different from strata in 2012 and 2012 DHS.
Thank you in advance
|
|
|
|
|
|
| Re: Weighting [message #28037 is a reply to message #28033] |
Fri, 03 November 2023 20:52  |
lmr
Messages: 3
Registered: November 2023
|
Member |
|
|
oh thank you for the answer.
but, im not clear at all. what should i do in spss?
Can the numbering of strata be equated with 2012 and 2016 SDKI? i mean, 1 for Aceh urban, 2 = Aceh rural, 3 = north Sumatra urban, 4 = north Sumatra rural, etc?
Thank you for regarding
[Updated on: Fri, 03 November 2023 21:13] Report message to a moderator |
|
|
|
| Re: Weighting [message #28134 is a reply to message #28037] |
Thu, 16 November 2023 14:36 |
Janet-DHS
Messages: 893
Registered: April 2022
|
Senior Member |
|
|
Following is a response from DHS staff member, Tom Pullum:
Weights are cluster-level. They have the same value for every household in a cluster. The average weight is 1, but the values have been multiplied by 1,000,000 just to get rid of a decimal point. We cannot provide support for SPSS but if you google something like "SPSS weights" you will find several websites that give the syntax.
An adjustment for strata is completely different from an adjustment for weights. (There is also an adjustment for clusters, and that is also completely different.) As I said before, the strata in these surveys (v023) are combinations of region (v024) and place of residence (v025). In the 2012 survey, v023 does not have a value label. I have constructed the label for v023 in the 2012 survey and will paste it below. It has 65 values. The label for v023 in the 2017 survey has 68 values. You will have to go through and find why and where there are differences. I suspect that in the 2012 survey there were 3 strata with no cases. You can do a recode of v023 in one (or both) of these surveys to get the labels to match. Let us know if you have other questions.
Value label for v023 (or hv023) in the Indonesia 2012 survey (not given in the data files):
1 aceh urban
2 aceh rural
3 north sumatera urban
4 north sumatera rural
5 west sumatera urban
6 west sumatera rural
7 riau urban
8 riau rural
9 jambi urban
10 jambi rural
11 south sumatera urban
12 south sumatera rural
13 bengkulu urban
14 bengkulu rural
15 lampung urban
16 lampung rural
17 bangka belitung urban
18 bangka belitung rural
19 riau islands urban
20 riau islands rural
21 jakarta urban
22 west java urban
23 west java rural
24 central java urban
25 central java rural
26 yogyakarta urban
27 yogyakarta rural
28 east java urban
29 east java rural
30 banten urban
31 banten rural
32 bali urban
33 bali rural
34 west nusa tenggara urban
35 west nusa tenggara rural
36 east nusa tenggara urban
37 east nusa tenggara rural
38 west kalimantan urban
39 west kalimantan rural
40 central kalimantan urban
41 central kalimantan rural
42 south kalimantan urban
43 south kalimantan rural
44 east kalimantan urban
45 east kalimantan rural
46 north sulawesi urban
47 north sulawesi rural
48 central sulawesi urban
49 central sulawesi rural
50 south sulawesi urban
51 south sulawesi rural
52 southeast sulawesi urban
53 southeast sulawesi rural
54 gorontalo urban
55 gorontalo rural
56 west sulawesi urban
57 west sulawesi rural
58 maluku urban
59 maluku rural
60 north maluku urban
61 north maluku rural
62 west papua urban
63 west papua rural
64 papua urban
65 papua rural
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
