| Creating Sub-Population of Neonates [message #20204] |
Wed, 07 October 2020 11:17 |
shujaat.smc@gmail.com
Messages: 75
Registered: July 2020
|
Senior Member |
|
|
Dear DHS Representative,
I am doing analysis with objectives of estimating the NMR & its determinants for singleton and bw=2500 & above.
For this I have created the subpopulation of neonates using command;
recode m19 (500/2495=-1 "LowBW") (9996/.=-2 "NoRecord") (2500/3000=1 "BW2500-3000") (3001/3500=2 "BW3001/3500") (3540/4000=3 "BW3540-4000") (4011/6000=4 "BW4011-6000") if v210!=1 & b0==0 , gen (subpopulationbw2)
V210= Born in month of month of interview (Yes=1)
b0==0 This include only singleton births
On simple tabulation I got following result;
tab subpopulationbw2
RECODE of |
m19 (birth |
weight in |
kilograms |
(3 |
decimals)) | Freq. Percent Cum.
------------+-----------------------------------
NoRecord | 32,191 96.36 96.36
LowBW | 237 0.71 97.07
BW2500-3000 | 553 1.66 98.72
BW3001/3500 | 219 0.66 99.38
BW3540-4000 | 118 0.35 99.73
BW4011-6000 | 90 0.27 100.00
------------+-----------------------------------
Total | 33,408 100.00
On adding the counts of my subpopulation total newborns= 980
******* My Querry 1 *************
svy linearized : tabulate subpopulationbw2, count
Number of strata = 8 Number of obs = 33408
Number of PSUs = 374 Population size = 3.846e-08
Design df = 366
----------------------
RECODE of |
m19 |
(birth |
weight in |
kilograms |
(3 |
decimals) |
) | count
----------+-----------
NoRecord | 3.7e-08
LowBW | 3.3e-10
BW2500-3 | 6.9e-10
BW3001/3 | 2.9e-10
BW3540-4 | 1.5e-10
BW4011-6 | 1.2e-10
|
Total | 3.8e-08
----------------------
Key: count = weighted counts
Question: Why these numbers are appearing 6.9e-10 ?????? in my subpopulation of interest ????? What should I do ????
******** My Query 2 ***********************************************
. svy linearized : stcox ib(4).v024
(running stcox on estimation sample)
Survey: Cox regression
Number of strata = 8 Number of obs = 34425
Number of PSUs = 374 Population size = 3.961e-08
Subpop. no. of obs = 34065
Subpop. size = 3.910e-08
Design df = 366
F( 3, 364) = 1.45
Prob > F = 0.2293
------------------------------------------------------------ ------------------
| Linearized
_t | Haz. Ratio Std. Err. t P>|t| [95% Conf. Interval]
-------------+---------------------------------------------- ------------------
v024 |
punjab | 1.36725 .2537737 1.69 0.093 .9491468 1.969531
sindh | 1.150385 .2243874 0.72 0.473 .7839022 1.688202
kpk | 1.237139 .2541046 1.04 0.301 .8260464 1.852819
My Population size was 980 as I have mentioned above, however in this univariate cox model my population size is Population size = 3.961e-08
my stset command stset dayofneonataldeath [pweight = v005], failure(neonataldeath==1)
Question2: What should I do for the correct Model Building in stepwise forward manner ????? How I can correct my population size appearing as 3.961e-08 ???
Best Regards
Dr. Hussain
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
