Statalist The Stata Listserver


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: Graphing change over time of a categorical variable


From   "Stevenson, Brendan" <B.S.Stevenson@massey.ac.nz>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: Graphing change over time of a categorical variable
Date   Tue, 20 Mar 2007 14:47:42 +1200

Hi all,
I was hoping that somebody out there could help me:
I began doing this in Visual Basic (after summarising the data in STATA and copying the table to EXCEL), but thought perhaps it would be possible to do this in STATA. I will describe the problem using one of the variables in our longitudinal study...
Household Type:
Sole person
Sole parent
Couple no children
Couple with children
Shared House/Flat
Other

Over time individuals will move between these different household types (usually in fairly predictable ways), unfortunately by the fifth sampling wave the possible combinations (i.e. the number of different possible household types and individual has been part of over time) becomes unwieldy and difficult to summarise.
Given that the groups we often disseminate results tend to be statistically na´ve (government and community agencies/groups), presenting complex information graphically becomes essential.

A simple 2 wave example would be
Wave 1			Wave 2
Sole Person			Sole Person
Sole Person			Sole Parent
Sole Person			Couple no children
Sole Person			Couple with children
Sole Person			Shared House/Flat
Sole Person			Other
Sole Parent			Sole Person
Sole Parent			Sole Parent
Sole Parent			Couple no children
Sole Parent			Couple with children
Sole Parent			Shared House/Flat
Sole Parent			Other
Couple no children	Sole Person
Couple no children	Sole Parent
Couple no children	Couple no children
Couple no children	Couple with children
Couple no children	Shared House/Flat
Couple no children	Other
...

My idea was to have arrows between the categories (like a path diagram), where the arrow will represent a summary variable (e.g. %, mean, or mean of a second variable) in its thickness. Over successive sampling waves, by filtering out all but those who began in say a couple with children household, you could map which households individual were more likely to have moved to over time (and conversely least likely).

Any ideas?
Have I described the problem adequately?

Cheers,
Brendan Stevenson
Massey University
New Zealand

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



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