Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


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

Re: st: Statistical significance of standardized rates


From   Steve Samuels <sjsamuels@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Statistical significance of standardized rates
Date   Fri, 5 Apr 2013 18:15:52 -0400

This usage is completely inappropriate. I doubt that you need direct
standardization, but to know that you should learn something about the
technique, including what a "standard population" is and what "stratum"
means. The manual is a good place to start and the topic is covered in
every epidemiology text. Some kind of regression program--logistic or
Poisson, may be better suited to your needs. For us to advise, you'd
have to provide a description of your study. Also, when you present
results, follow the FAQ dictate to show us exactly what you typed (not a
"summary") and all the results.


Steve



On Apr 4, 2013, at 12:03 PM, Karman Tandon wrote:

Hi Nick,
I also noticed that the confidence intervals have zero length. Here is
a summary of the dstdize I ran:

dstdize x pop a b c d, by(quartiles)

- x is mortality, a dichotomous 1/0
- pop is the entire population in my data set
- a, b, c, d are a mix of continuous and dichotomous variables
- quartiles is the quartile group that each member of the population
falls into based on a continuous variable "e" that is not included in
the a,b,c,d being standardized by.

Is this an appropriate use of the command?

Thank you,
Karman

On Thu, Apr 4, 2013 at 11:57 AM, Nick Cox <njcoxstata@gmail.com> wrote:
> -dstdize- is a command, not a function.
> 
> Before you proceed further, note that the confidence intervals appear
> to be of essentially zero length. That seems implausible, but if it's
> true, any difference you like is significant at conventional levels.
> Getting a P-value is moot. However, I'd be suspicious without being
> sure that the instructions were correct.
> 
> Nick
> njcoxstata@gmail.com
> 
> On 4 April 2013 16:44, Karman Tandon <karmantandon@gmail.com> wrote:
>> Hi,
>> 
>> I've used the dstdize function to standardize/adjust across a number
>> of variables. My data looks like this:
>> 
>> Summary of Study Populations:
>>     quartile        N      Crude     Adj_Rate       Confidence Interval
>> --------------------------------------------------------------------------
>>        1           290   0.196552     0.060000    [  0.060000,    0.060000]
>>        2           233   0.124464     0.030526    [  0.030526,    0.030526]
>>        3           216   0.087963     0.020000    [  0.020000,    0.020000]
>>        4           210   0.066667     0.014737    [  0.014737,    0.014737]
>> 
>> How do I determine p-values for the significance of the adjusted rates
>> compared to one another?
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/faqs/resources/statalist-faq/
> *   http://www.ats.ucla.edu/stat/stata/
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/


*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index