Re: st: Re: Problem with combining sibling data

[Nick Cox <njcoxstata@gmail.com>]

You would need to exclude codes of -1 and -2 e.g. by

... if min(nolder, nyounger) >= 0
Nick
njcoxstata@gmail.com


On 13 May 2013 10:51, Nick Cox <njcoxstata@gmail.com> wrote:
> What is "sibling shape"? More generally, what precisely do the data
> look like and precisely what code did you try and precisely why did
> not that code do what you want? Otherwise this is only marginally
> different from "my code doesn't do what I want".
>
> On the face of it,
>
> total number of siblings = number older + number younger
>
> and so forth, so "+" appears to be all the Stata you need here.
>
> generate nsiblings = nyounger + nolder
>
> and variations on the same theme.
>
> Nick
> njcoxstata@gmail.com
>
>
> On 13 May 2013 10:36, Annabel Mullin <ajcmullin@yahoo.com> wrote:
>>>> Hi,
>>>>
>>>> I am keen to try and create a total sibling variable for data that I have that is currently separated out into four vars:
>>>> no.of brother-sister in same house older,
>>>> no.of brother-sister in same house younger
>>>> no.of brother-sister outside house older
>>>> no.of brother-sister outside house younger
>>>
>>>> for each 0-4 = number siblings and -1 is missing questionnaire and -2 is not stated
>>>>
>>>> This is from a longitudinal cohort study and I'm being dense but can't seem to create a variable that gives me the sibling shape of each cohort member - I have used sum, count, total but they're not an accurate portrayal of the siblings that each cohort member has.
>>>>
>>>> What can you recommend?
>>>>
>>>> Thank you
>>>> Annabel
>>>
>>
>>
>> *
>> *   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/