Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: Re: st: Calculate response rate (proportio​​​n of non-missin​​​g value)

From   Maarten buis <>
Subject   st: Re: st: Calculate response rate (proportio​​​n of non-missin​​​g value)
Date   Mon, 14 Feb 2011 08:40:40 +0000 (GMT)

--- On Mon, 14/2/11, Guo Chen wrote:
> I have a survey data right now and need to calculate the
> response rate to each question by  whether the interviewee
> is in a group, and then do a ttest of the response rates
> between the two groups of people. 

To get the response rate you need a dumy variable that is 1
when the respondent answered to that question and 0 otherwise.
The -missing()- function gives you exactly the opposite. If 
you precede that with a -!- you reverse the logical statement.
Without thinking about it I literally read -!missing()- as 
"not missing". For more on this see: <> and
an application of this logic on this text on one of my T-shirts:

Once you have that dummy, then the rate is just the mean of 
that dummy. For convenience you can us -ds- to get a list of
variables you want to test, and use a -foreach- loop to loop
over that list, but that is just icing on the cake; it saves
you typing (and thus diminishes the chance of a typo...). 

*-------------------------- begin example -------------------
sysuse nlsw88, clear

ds race, not
local varl "`r(varlist)'"

// other (3) and missing values will be turned into missing
// so 1 is black, 0 is white
gen black = race == 2 if race <=2
gen byte miss = .

foreach var of local varl {
	qui replace miss = !missing(`var')
	di "================= testing response rate for `var'"
	ttest miss, by(black)
*-------------------------- end example -----------------------
(For more on examples I sent to the Statalist see: )

Hope this helps,

Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen


*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index