Skip to content

svytable: How do i create a table informing a third variable?

5 messages · Anthony Damico, Juan Ceccarelli Arias, David Winsemius

#
Hello
Im analyzing a survey and i need to obtain some statistics per groups.
Im able to create a table with sex and age. However, if i want to know how
much income earns the population by sex and age, i can't.
Im loading the dataset as describe the line below
NN <- svydesign(ids = ~1, data = encuesta, weights = fact)
Some simple table i can create
table(svytable(~age+sex,design=NN))
But im not able to handle the same tabulate referencing a income variable,
in this case, wage.
Can you help me?
Thanks for your replies and time.
#
# mean
svymean( ~ income_variable , NN )
svyby( ~ income_variable , ~ age + sex , NN , svymean )

# median
svyquantile( ~ income_variable , NN )
svyby( ~ income_variable , ~ age + sex , NN , svyquantile , 0.5 )




On Fri, Sep 2, 2016 at 3:04 PM, Juan Ceccarelli Arias <jfca283 at gmail.com>
wrote:

  
  
#
Thanks a lot. Your code does the trick.
One last question:
The tabulate produced is showing every cross in just one column.
I mean, it presents the region by order and sex=1, and then again the
region but by sex==2.
Can i list or present as this:
             sex1         sex2
region1  323.      3434..
...
regionN 123..  432..

and ignoring the remaining info (standar errors or se in this case)?
Again, thanks Anthony. Really.
On Fri, Sep 2, 2016 at 8:24 PM, Anthony Damico <ajdamico at gmail.com> wrote:

            

  
  
#
(Anthony's probably asleep.)

This doesn't ignore the se's but that could be easily done by omitting that column from the data argument:
#
Your help was everything i needed it.
Please, declare this topic as solved.
And thanks again.

On Sat, Sep 3, 2016 at 10:06 PM, David Winsemius <dwinsemius at comcast.net>
wrote: