Survey Data Analysis in Stata

Jeff Pitblado
Associate Director, Statistical Software
StataCorp LP
2009 Canadian Stata Users Group Meeting

Outline
1 Types of data 2

2 Survey data characteristics 4

2.1 Single stage designs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2 Multistage designs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.3 Poststratification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.4 Strata with a single sampling unit . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.5 Certainty units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

3 Variance estimation 15
3.1 Linearization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.1.1 Total estimator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.1.2 Regression models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.2 Balanced repeated replication (BRR) . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.3 Jackknife . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

4 Estimation for subpopulations 25

5 Summary 28
Why survey data?

Collecting data can be expensive and time consuming.

Consider how you would collect the following data:

Smoking habits of teenagers

Birth weights for expectant mothers with high blood pressure

Using stages of clustered sampling can help cut down on the expense and time.

1 Types of data
Simple random sample (SRS ) data
Observations are "independently" sampled from a data generating process.

Typical assumption: independent and identically distributed (iid)

Make inferences about the data generating process

Sample variability is explained by the statistical model attributed to the data generating pro-
cess

Standard data
Well use this term to distinguish this data from survey data.

Correlated data
Individuals are assumed not independent.
Cause:

Observations are taken over time

Random effects assumptions

Cluster sampling

Treatment:

Time-series models

Longitudinal/panel data models

cluster() option

2
Survey data
Individuals are sampled from a fixed population according to a survey design.
Distinguishing characteristics:

Complex nature under which individuals are sampled

Make inferences about the fixed population

Sample variability is attributed to the survey design

Standard data

Estimation commands for standard data:

proportion
regress

Well refer to these as standard estimation commands.

Survey data

Survey estimation commands are governed by the svy prefix.

svy: proportion
svy: regress

svy requires that the data is svyset.

3
2 Survey data characteristics
2.1 Single stage designs
Single-stage syntax
   
svyset psu weight , strata(varname) fpc(varname)

Primary sampling units (PSU)

Sampling weights pweight

Strata

Finite population correction (FPC)

Sampling unit
An individual or collection of individuals from the population that can be selected for observation.

Sampling groups of individuals is synonymous with cluster sampling.

Cluster sampling usually results in inflated variance estimates compared to SRS.

Example

High schools for sampling from the population of 12th graders.

Hospitals for sampling from the population of newborns.

Sampling weight
The reciprocal of the probability for an individual to be sampled.

Probabilities are derived from the survey design.

Sampling units
Strata

Typically considered to be the number of individuals in the population that a sampled indi-
vidual represents.

Reduces bias induced by the sampling design.

4
 Example
If there are 100 hospitals in our population, and we choose 5 of them, the sampling weight is
20 = 100/5. Thus a sampled hospital represents 20 hospitals in the population.
Sampling weights correct for over/under sampling of sections in the population. Many times
this over/under sampling is on purpose.
Strata
In stratified designs, the population is partitioned into well-defined groups, called strata.

Sampling units are independently sampled from within each stratum.

Stratification usually results in smaller variance estimates compared to SRS.

Example

States of the union are typically used as strata in national surveys in the US.

Demographic information like age group, gender, and ethnicity.

Although there is potential for improving efficiency by reducing sampling variability, it is usu-
ally not very practical to stratify on demographic information.
Finite population correction (FPC)
An adjustment applied to the variance due to sampling without replacement.

Sampling without replacement from a finite population reduces sampling variability.

q Note

The FPC affects the number of components in the linearized variance estimator for multi-
stage designs.

We can use svyset to specify an SRS design.

q

5
Example: svyset for single-stage designs

1. auto specifying an SRS design

2. nmihs the National Maternal and Infant Health Survey (1988) dataset came from a strati-
fied design

3. fpc a simulated dataset with variables that identify the characteristics from a stratified and
without-replacement clustered design

*** The auto data that ships with Stata

. sysuse auto
(1978 Automobile Data)
. svyset _n
pweight: <none>
VCE: linearized
Single unit: missing
Strata 1: <one>
SU 1: <observations>
FPC 1: <zero>
*** National Maternal and Infant Health Survey
. webuse nmihs
. svyset [pw=finwgt], strata(stratan)
pweight: finwgt
VCE: linearized
Single unit: missing
Strata 1: stratan
SU 1: <observations>
FPC 1: <zero>
*** Simulated data
. webuse fpc
. svyset psuid [pw=weight], strata(stratid) fpc(Nh)
pweight: weight
VCE: linearized
Single unit: missing
Strata 1: stratid
SU 1: psuid
FPC 1: Nh

6
 Below is a visual representation of a hypothetical population. Suppose each blue dot represents
an individual.

Population 1000

The following shows a 20% simple-random-sample. The solid symbols identify sampled indi-
viduals.

SRS sample 200

7
 Here we partition the population into small blocks, then sample 20% of the blocks. Not all
blocks contain the same number of individuals, so the sample size is a random quantity.

Cluster sample 20 (208 obs)

Here we partition the population into four big regions, then perform a 20% sample within each
region. The sample size is not exactly 20% of the population size due to unbalanced regions and
rounding.

Stratified sample 198

8
 Here we re-establish the smaller blocks within the four regions, then sample 20% of the blocks
within each region.

Stratified-cluster sample 20 (215 obs)

2.2 Multistage designs

Multistage syntax
  
svyset psu weight , strata(varname) fpc(varname)
  
|| ssu , strata(varname) fpc(varname)
  
|| ssu , strata(varname) fpc(varname) ...

Stages are delimited by ||

SSU secondary/subsequent sampling units

FPC is required at stage s for stage s + 1 to play a role in the linearized variance estimator

q Note
svyset will note that it is disregarding subsequent stages when an FPC is not specified for a
given stage.
q

9
 2.3 Poststratification
Poststratification
A method for adjusting sampling weights, usually to account for underrepresented groups in the
population.

Adjusts weights to sum to the poststratum sizes in the population

Reduces bias due to nonresponse and underrepresented groups

Can result in smaller variance estimates

Syntax

svyset ... poststrata(varname) postweight(varname)

q Note
Recall that I said it is usually not vey practical to stratify on demographic information such as
age group, gender, and ethnicity. However we can usually poststratify on these variables using the
frequency distribution information available from census data.
q

Example: svyset for poststratification

A veterinarian has 1300 clients, 450 cats and 850 dogs. He would like to estimate the average
annual expenses of his clientele but only has enough time to gather information on 50 randomly
selected clients. Thus we have an SRS design, the sampling weight is 26 = 1300/50.
Notice that the dog clients are (on average) twice as expensive as cat clients. We can use the
above frequency distribution of dogs and cats to poststratify on animal type.

*** Cat and dog data from Levy and Lemeshow (1999)
. webuse poststrata
. bysort type: sum totexp

-> type = dog

Variable Obs Mean Std. Dev. Min Max

totexp 32 49.85844 8.376695 32.78 66.2

-> type = cat

Variable Obs Mean Std. Dev. Min Max

totexp 18 21.71111 8.660666 7.14 39.88

10
Here are the mean estimates with postratification:

. svyset [pw=weight], poststrata(type) postweight(postwgt) fpc(fpc)

pweight: weight
VCE: linearized
Poststrata: type
Postweight: postwgt
Single unit: missing
Strata 1: <one>
SU 1: <observations>
FPC 1: fpc
. svy: mean totexp
(running mean on estimation sample)
Survey: Mean estimation
Number of strata = 1 Number of obs = 50
Number of PSUs = 50 Population size = 1300
N. of poststrata = 2 Design df = 49

Linearized
Mean Std. Err. [95% Conf. Interval]
totexp 40.11513 1.163498 37.77699 42.45327

Here are the mean estimates without postratification:

. svyset _n [pw=weight]
pweight: weight
VCE: linearized
Single unit: missing
Strata 1: <one>
SU 1: <observations>
FPC 1: <zero>
. svy: mean totexp
(running mean on estimation sample)
Survey: Mean estimation
Number of strata = 1 Number of obs = 50
Number of PSUs = 50 Population size = 1300
Design df = 49

Linearized
Mean Std. Err. [95% Conf. Interval]

totexp 39.7254 2.265746 35.17221 44.27859

11
2.4 Strata with a single sampling unit
How do we get stuck with strata that have only one sampling unit?

Missing data can cause entire sampling units to be dropped from the analysis, possibly leav-
ing a single sampling unit in the estimation sample.

Certainty units

Bad design

Big problem for variance estimation

Consider a sample with only 1 observation

svy reports missing standard error estimates by default

Finding these lonely sampling units

Use svydes:

Describes the strata and sampling units

Helps find strata with a single sampling unit

12
Example: svydes
The NHANES2 data has 31 strata, each containing 2 PSUs.

*** Second National Health and Nutrition Examination Survey

. webuse nhanes2
. svydes
Survey: Describing stage 1 sampling units
pweight: finalwgt
VCE: linearized
Single unit: missing
Strata 1: strata
SU 1: psu
FPC 1: <zero>
#Obs per Unit

Stratum #Units #Obs min mean max

1 2 380 165 190.0 215
2 2 185 67 92.5 118
3 2 348 149 174.0 199
4 2 460 229 230.0 231
5 2 252 105 126.0 147
6 2 298 131 149.0 167
7 2 476 206 238.0 270
8 2 338 158 169.0 180
9 2 244 100 122.0 144
10 2 262 119 131.0 143
11 2 275 120 137.5 155
12 2 314 144 157.0 170
13 2 342 154 171.0 188
14 2 405 200 202.5 205
15 2 380 189 190.0 191
16 2 336 159 168.0 177
17 2 393 180 196.5 213
18 2 359 144 179.5 215
20 2 285 125 142.5 160
21 2 214 102 107.0 112
22 2 301 128 150.5 173
23 2 341 159 170.5 182
24 2 438 205 219.0 233
25 2 256 116 128.0 140
26 2 261 129 130.5 132
27 2 283 139 141.5 144
28 2 299 136 149.5 163
29 2 503 215 251.5 288
30 2 365 166 182.5 199
31 2 308 143 154.0 165
32 2 450 211 225.0 239

31 62 10351 67 167.0 288

13
Some variables in this dataset have enough missing values to cause us the lonely PSU problem.

*** Mean high density lipids (mg/dL)

. svy: mean hdresult
(running mean on estimation sample)
Survey: Mean estimation
Number of strata = 31 Number of obs = 8720
Number of PSUs = 60 Population size = 98725345
Design df = 29

Linearized
Mean Std. Err. [95% Conf. Interval]

hdresult 49.67141 . . .

Note: missing standard error because of stratum with single

sampling unit.

Use if e(sample) after estimation commands to restrict svydess focus on the estimation sam-
ple. The single option will further restrict output to strata with one sampling unit.

*** Restrict to the estimation sample

. svydes if e(sample), single
Survey: Describing strata with a single sampling unit in stage 1
pweight: finalwgt
VCE: linearized
Single unit: missing
Strata 1: strata
SU 1: psu
FPC 1: <zero>
#Obs per Unit

Stratum #Units #Obs min mean max

1 1* 114 114 114.0 114

2 1* 98 98 98.0 98

Specifying variable names with svydes will result in more information about missing values.

*** Specifying variables for more information

. svydes hdresult, single
Survey: Describing strata with a single sampling unit in stage 1
pweight: finalwgt
VCE: linearized
Single unit: missing
Strata 1: strata
SU 1: psu
FPC 1: <zero>
#Obs with #Obs with #Obs per included Unit
#Units #Units complete missing
Stratum included omitted data data min mean max

1 1* 1 114 266 114 114.0 114

2 1* 1 98 87 98 98.0 98

14
 Handling lonely sampling units

1. Drop them from the estimation sample.

2. svyset one of the ad-hoc adjustments in the singleunit() option.
3. Somehow combine them with other strata.

2.5 Certainty units

Sampling units that are guaranteed to be chosen by the design.
Certainty units are handled by treating each one as its own stratum with an FPC of 1.

3 Variance estimation
Stata has three variance estimation methods for survey data:

Linearization
Balanced repeated replication
The jackknife

q Note

Linearization

Statas robust for complex data

The default variance estimation method for svy.

Replication methods

Motivation
Linearization can have poor performance in datasets with a small number of sam-
pling units.
Due to privacy concerns, data providers are reluctant to release strata and sampling
unit information in public-use data. Thus some datasets now come packaged with
weight variables for use with replication methods.
Concept
Think of a replicate as a copy of the point estimates.
The idea is to resample the data, computing replicates from each resample, then
using the replicates to estimate the variance.
q

15
3.1 Linearization
Linearization
A method for deriving a variance estimator using a first order Taylor approximation of the point
estimator of interest.

Foundation: Variance of the total estimator

Syntax
 
svyset ... vce(linearized)

Delta method

Huber/White/robust/sandwich estimator

3.1.1 Total estimator

Total estimator Stratified two-stage design

yhijk observed value from a sampled individual

Strata: h = 1, . . . , L

PSU: i = 1, . . . , nh

SSU: j = 1, . . . , mhi

Individual: k = 1, . . . , mhij

X
Yb = whijk yhijk
X nh X
Vb (Yb ) = (1 fh ) (yhi y h )2 +
n h1 i
h
X X mhi X
fh (1 fhi ) (yhij y hi )2
h i
mhi 1 j

fh is the sampling fraction for stratum h in the first stage.

fhi denotes a sampling fraction in the second stage.

Remember that the design degrees of freedom is

df = NPSU Nstrata

16
Example: svy: total
Lets use our (imaginary) survey data on high school seniors to estimate the number of smokers in
the population.
. webuse seniors
. svyset
pweight: sampwgt
VCE: linearized
Single unit: missing
Strata 1: state
SU 1: county
FPC 1: ncounties
Strata 2: <one>
SU 2: school
FPC 2: nschools
Strata 3: gender
SU 3: <observations>
FPC 3: nseniors
*** Estimate number of seniors who have smoked
. svy: total smoked
(running total on estimation sample)
Survey: Total estimation
Number of strata = 50 Number of obs = 10559
Number of PSUs = 100 Population size = 20992929
Design df = 50

Linearized
Total Std. Err. [95% Conf. Interval]

smoked 8347260 331155.1 7682115 9012404

*** Use first stage without FPC

. svyset county [pw=sampwgt], strata(state)
pweight: sampwgt
VCE: linearized
Single unit: missing
Strata 1: state
SU 1: county
FPC 1: <zero>
. svy: total smoked
(running total on estimation sample)
Survey: Total estimation
Number of strata = 50 Number of obs = 10559
Number of PSUs = 100 Population size = 20992929
Design df = 50

Linearized
Total Std. Err. [95% Conf. Interval]

smoked 8347260 346853.4 7650584 9043935

17
3.1.2 Regression models

Linearized variance for regression models

Model is fit using estimating equations.

G()
b is a total estimator, use Taylor expansion to get Vb ().
b

X
G()
b = w j sj x j = 0
j

b = DVb {G()}| 0
Vb () b bD
=

ML models

G()
b is the gradient

sj is an equation-level score

D is the inverse negative Hessian matrix at the solution

Least squares regression

G()
b is the normal equations

sj is a residual

D is the inverse of the weighted outer product of the predictorsincluding the intercept

D = (X0 WX)1

18
Example: svy: logit
Here is an example of a logistic regression, modeling the incidence of high blood pressure as a
function of some demographic variables.

*** Second National Health and Nutrition Examination Survey

. webuse nhanes2
. svyset
pweight: finalwgt
VCE: linearized
Single unit: missing
Strata 1: strata
SU 1: psu
FPC 1: <zero>
*** Model high blood pressure on some demographics
. describe highbp height weight age female
storage display value
variable name type format label variable label

highbp byte %8.0g 1 if BP > 140/90, 0 otherwise

height float %9.0g height (cm)
weight float %9.0g weight (kg)
age byte %9.0g age in years
female byte %8.0g 1=female, 0=male
. svy: logit highbp height weight age female
(running logit on estimation sample)
Survey: Logistic regression
Number of strata = 31 Number of obs = 10351
Number of PSUs = 62 Population size = 117157513
Design df = 31
F( 4, 28) = 178.69
Prob > F = 0.0000

Linearized
highbp Coef. Std. Err. t P>|t| [95% Conf. Interval]
height -.0316386 .0058648 -5.39 0.000 -.0435999 -.0196772
weight .0511574 .0031191 16.40 0.000 .0447959 .057519
age .0492406 .0023624 20.84 0.000 .0444224 .0540587
female -.3215716 .0884387 -3.64 0.001 -.5019435 -.1411998
_cons -2.858968 1.049395 -2.72 0.010 -4.999224 -.7187117

19
 3.2 Balanced repeated replication (BRR)
Balanced repeated replication
For designs with two PSUs in each of L strata.

Compute replicates by dropping a PSU from each stratum.

Find a balanced subset of the 2L replicates. L r < L + 4

The replicates are used to estimate the variance.

Syntax
 
svyset ... vce(brr) mse

q Note

The idea is to resample the data, compute replicates from each resample, then use the repli-
cates to estimate the variance.

Balance here means that stratum specific contributions to the variance cancel out. In other
words, no stratum contributes more to the variance than any other.

We can find a balanced subset by finding a Hadamard matrix of order r.

When the dataset contains replicate weight variables, you do not need to worry about Hadamard
matrices.
q
q Note

These replicate weights are used to produce a copy of the point estimates (replicate). The
replicates are then used to estimate the variance.

svy brr can employ replicate weight variables in the dataset, if you svyset them. Oth-
erwise, svy brr will automatically adjust the sampling weights to produce the replicates;
however, a Hadamard matrix must be specified.
q

20
 BRR variance formulas

b point estimates

b(i) ith replicate of the point estimates

(.) average of the replicates

Default variance formula:

r
1 X b b(i) (.) }0
V () =
b b { (i) (.) }{
r i=1
Mean squared error (MSE) formula:
r
1 X b b 0
V () =
b b { (i) }{
b b(i) }
r i=1
q Note

The default variance formula uses deviations of the replicates from their mean.

The MSE formula uses deviations of the replicates from the point estimates.

BRR * is clickable, taking you to a short help file informing you that you used the MSE
formula for BRR variance estimation. q

21
Example: svy brr: logit
Lets revisit the previous logistic model fit, but use BRR for variance estimation.

*** Second National Health and Nutrition Examination Survey

. webuse nhanes2brr
. svyset [pw=finalwgt], vce(brr) mse brrweight(brr_*)
pweight: finalwgt
VCE: brr
MSE: on
brrweight: brr_1 brr_2 brr_3 brr_4 brr_5 brr_6 brr_7 brr_8 brr_9 brr_10
brr_11 brr_12 brr_13 brr_14 brr_15 brr_16 brr_17 brr_18 brr_19
brr_20 brr_21 brr_22 brr_23 brr_24 brr_25 brr_26 brr_27 brr_28
brr_29 brr_30 brr_31 brr_32
Single unit: missing
Strata 1: <one>
SU 1: <observations>
FPC 1: <zero>
. svy: logit highbp height weight age female
(running logit on estimation sample)
BRR replications (32)
1 2 3 4 5
................................
Survey: Logistic regression Number of obs = 10351
Population size = 117157513
Replications = 32
Design df = 31
F( 4, 28) = 173.94
Prob > F = 0.0000

BRR *
highbp Coef. Std. Err. t P>|t| [95% Conf. Interval]

height -.0316386 .0058774 -5.38 0.000 -.0436255 -.0196516

weight .0511574 .0031267 16.36 0.000 .0447806 .0575343
age .0492406 .0023449 21.00 0.000 .0444581 .054023
female -.3215716 .0897343 -3.58 0.001 -.5045859 -.1385574
_cons -2.858968 1.044318 -2.74 0.010 -4.988868 -.7290671

22
 3.3 Jackknife
The jackknife
A replication method for variance estimation. Not restricted to a specific survey design.

Delete-1 jackknife: drop 1 PSU

Delete-k jackknife: drop k PSUs within a stratum

Syntax
 
svyset ... vce(jackknife) mse

q Note

svy jackknife can employ replicate weight variables in the dataset, if you svyset them.
Otherwise, svy jackknife will automatically adjust the sampling weights to produce the
replicates using the delete-1 jackknife methodology.

In the delete-1 jackknife, each PSU is represented by a corresponding replicate.

The delete-k jackknife is only supported if you already have the corresponding replicate
weight variables for svyset. q

Jackknife variance formulas

b(h,i) replicate of the point estimates from stratum h, PSU i

h average of the replicates from stratum h

mh = (nh 1)/nh delete-1 multiplier for stratum h

Default variance formula:

L
X nh
X
Vb ()
b = (1 fh ) mh { b(h,i) h }0
b(h,i) h }{
h=1 i=1

Mean squared error (MSE) formula:

L
X nh
X
Vb ()
b = (1 fh ) mh {
b(h,i) }{
b b 0
b(h,i) }
h=1 i=1

23
q Note

The default variance formula uses deviations of the replicates from their mean.

The MSE formula uses deviations of the replicates from the point estimates.

Jknife * is clickable, taking you to a short help file informing you that you used the MSE
formula for jackknife variance estimation.

Make sure to specify the correct multiplier when you svyset jackknife replicate weight
variables. q

Example: svy jackknife: logit

Here we are again with our now familiar logistic model fit, using the delete-1 jackknife variance
estimator.

*** Second National Health and Nutrition Examination Survey

. webuse nhanes2
. svyset
pweight: finalwgt
VCE: linearized
Single unit: missing
Strata 1: strata
SU 1: psu
FPC 1: <zero>
. svy jknife, mse: logit highbp height weight age female
(running logit on estimation sample)
Jackknife replications (62)
1 2 3 4 5
.................................................. 50
............
Survey: Logistic regression
Number of strata = 31 Number of obs = 10351
Number of PSUs = 62 Population size = 117157513
Replications = 62
Design df = 31
F( 4, 28) = 178.53
Prob > F = 0.0000

Jknife *
highbp Coef. Std. Err. t P>|t| [95% Conf. Interval]
height -.0316386 .0058674 -5.39 0.000 -.0436052 -.0196719
weight .0511574 .0031203 16.40 0.000 .0447936 .0575213
age .0492406 .0023634 20.83 0.000 .0444204 .0540607
female -.3215716 .088471 -3.63 0.001 -.5020093 -.1411339
_cons -2.858968 1.049924 -2.72 0.011 -5.000302 -.7176329

24
 Replicate weight variable
A variable in the dataset that contains sampling weight values that were adjusted for resampling
the data using BRR or the jackknife.

Typically used to protect the privacy of the survey participants.

Eliminate the need to svyset the strata and PSU variables.

Syntax

svyset ... brrweight(varlist)

 
svyset ... jkrweight(varlist , ... multiplier(#) )

4 Estimation for subpopulations

Focus on a subset of the population

Subpopulation variance estimation:

Assumes the same survey design for subsequent data collection.

The subpop() option.

Restricted-sample variance estimation:

Assumes the identified subset for subsequent data collection.

Ignores the fact that the sample size is a random quantity.
The if and in restrictions.

q Note

As I mentioned earlier on, variability is governed by the survey design, so our variance
estimates assume the design is fixed. The subpop() option assumes this too.

If we discourage you from using if and in, why does svy allow them?

You might want to restrict your sample because of known defects in some of the vari-
ables.
Researchers can use if and in to conduct simulation sudies by simulating survey sam-
ples from a population dataset without having to use preserve and restore.

25
 We can illustrate the difference between these estimators with an SRS design.
q

Total from SRS data

Data is y1 , . . . , yn and S is the subset of observations.


1, if j S
j (S) =
0, otherwise

Subpopulation (or restricted-sample) total:

n
X
YbS = j (S)wj yj
j=1

Sampling weight and subpopulation size:

n
N X N
wj = , NS = j (S)wj = nS
n j=1
n

Variance of a subpopulation total

Sample n without replacement from a population comprised of the NS subpopulation values with
N NS additional zeroes.
n  2
 n n X 1b
V (YS ) = 1
b b j (S)yj YS
N n 1 j=1 n

Variance of a restricted-sample total

Sample nS without replacement from the subpopulation of NS values.
  n  2
n S nS
X 1
Ve (YbS ) = 1 j (S) yj YbS
NS nS 1 j=1 nS

26
Example: svy, subpop()
Suppose we want to estimate the mean birth weight for mothers with high blood pressure. The
highbp variable (in the nmihs data) is an indicator for mothers with high blood pressure.
In the reported results, the subpopulation information is provided in the header. Notice that
although the restricted sample results reproduce the same mean, the standard errors differ.

*** National Maternal and Infant Health Survey

. webuse nmihs
. svyset [pw=finwgt], strata(stratan)
pweight: finwgt
VCE: linearized
Single unit: missing
Strata 1: stratan
SU 1: <observations>
FPC 1: <zero>
*** Focus: birthweight, mothers with high blood pressure
. describe birthwgt highbp
storage display value
variable name type format label variable label

birthwgt int %8.0g Birthweight in grams

highbp byte %8.0g hibp High blood pressure: 1=yes,0=no
. label list hibp
hibp:
0 norm BP
1 hi BP
*** Subpopulation estimation
. svy, subpop(highbp): mean birthwgt
(running mean on estimation sample)
Survey: Mean estimation
Number of strata = 6 Number of obs = 9953
Number of PSUs = 9953 Population size = 3898922
Subpop. no. obs = 595
Subpop. size = 186196.7
Design df = 9947

Linearized
Mean Std. Err. [95% Conf. Interval]

birthwgt 3202.483 33.29493 3137.218 3267.748

*** Restricted sample estimation

. svy: mean birthwgt if highbp
(running mean on estimation sample)
Survey: Mean estimation
Number of strata = 6 Number of obs = 595
Number of PSUs = 595 Population size = 186197
Design df = 589

Linearized
Mean Std. Err. [95% Conf. Interval]
birthwgt 3202.483 28.7201 3146.077 3258.89

27
5 Summary
1. Use svyset to specify the survey design for your data.

2. Use svydes to find strata with a single PSU.

3. Choose your variance estimation method; you can svyset it.

4. Use the svy prefix with estimation commands.

5. Use subpop() instead of if and in.

References
[1] Levy, P. and S. Lemeshow. 1999. Sampling of Populations. 3rd ed. New York: Wiley.

[2] StataCorp. 2009. Survey Data Reference Manual: Release 11. College Station, TX: StataCorp
LP.

28