How can I calculate the standard deviation with weight??

[Stephen Wee]

Dear All,
I have survey data. It's collected through stratified sampling method. I set a weight which means the inverse of the probability of the observation is included. Therefore,when I calculate the mean or run regression, I should use "pweight". But pweight can't be used to calculate sd, then what should I do to calculate the standard deviation? (I use "collapse" to calculate mean\median\sd)
Thank you!

[Phil Bromiley]

You didn't get a quick answer. You will increase your chances of a useful answer by following the FAQ on asking questions – provide Stata code in code delimiters, readable Stata output, and sample data using dataex.

The conventional way to calculate summary statistics is the summarize command. It does allow weights. I don't do survey data so I can't comment on whether this is appropriate for survey data.

[Rich Goldstein]

my guess is that what you are calling the "standard deviation" is what Stata calls the standard error for survey data; first, of course, you need to -svyset- your data; then

help mean

within the help for -mean-, click on "Methods and Formulas" under "links to pdf documentation"; there you will see a formula for the variance followed by the following statement: "The standard error of the mean is the square root of the variance."

since you provided no real information on the survey, I have no idea whether -collapse- is giving you anything meaningful

if this is not what you mean, please clarify

[Richard Williams]

Unfortunately, sum does not work with pweights. Instead, svyset your data, and check out -help svy_estat-. Here is an example:

. webuse nhanes2f, clear

. svyset

      pweight: finalwgt
          VCE: linearized
  Single unit: missing
     Strata 1: stratid
         SU 1: psuid
        FPC 1: <zero>

. svy: mean health weight height
(running mean on estimation sample)

Survey: Mean estimation

Number of strata =      31      Number of obs   =       10,335
Number of PSUs   =      62      Population size =  116,997,257
                                Design df       =           31

--------------------------------------------------------------
             |             Linearized
             |       Mean   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
      health |   3.607662   .0228896      3.560979    3.654346
      weight |   71.91131   .1670327      71.57065    72.25198
      height |   168.4647   .1471856      168.1645    168.7649
--------------------------------------------------------------

. estat sd

-------------------------------------
             |       Mean   Std. Dev.
-------------+-----------------------
      health |   3.607662    1.148297
      weight |   71.91131    15.43409
      height |   168.4647    9.702569
-------------------------------------

.

[Stephen Wee]

Unfortunately, sum does not work with pweights. Instead, svyset your data, and check out -help svy_estat-. Here is an example:

. webuse nhanes2f, clear

. svyset

pweight: finalwgt
VCE: linearized
Single unit: missing
Strata 1: stratid
SU 1: psuid
FPC 1: <zero>

. svy: mean health weight height
(running mean on estimation sample)

Survey: Mean estimation

Number of strata = 31 Number of obs = 10,335
Number of PSUs = 62 Population size = 116,997,257
Design df = 31

--------------------------------------------------------------
| Linearized
| Mean Std. Err. [95% Conf. Interval]
-------------+------------------------------------------------
health | 3.607662 .0228896 3.560979 3.654346
weight | 71.91131 .1670327 71.57065 72.25198
height | 168.4647 .1471856 168.1645 168.7649
--------------------------------------------------------------

. estat sd

-------------------------------------
| Mean Std. Dev.
-------------+-----------------------
health | 3.607662 1.148297
weight | 71.91131 15.43409
height | 168.4647 9.702569
-------------------------------------

.

Thank you. I will try this.