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: how to handle rank ordered responses

From   ramesh <>
Subject   st: how to handle rank ordered responses
Date   Fri, 18 Oct 2013 09:14:05 -0700 (PDT)

Hello Statalist,
I am working to analyze how site specific attributes (water quality,
closeness, wildlife, and size) affect people’s participation on freshwater
recreation activities (boating, swimming, fishing, viewing, etc.). I have
survey data and people expressed their views regarding the importance of
site specific attributes in a scale from 0 to 3, with 0 - not important at
all and 3 -most important. My data structure looks like:
id  boating   fishing   swimming  viewing   water-quality   closeness  
wildlife   size   controls(x)
--  -------    -----      ------     -----      -----------    --------      
----     ----   ------
1     yes       no             yes       yes          3                  1              
2        0         yes
2     no         yes           no         yes         1                  2              
3        0          yes
3     yes        yes          yes        no          1                  2              
0         3          yes
 .     ..             ...              ....

I want to estimate which one attribute is most important and important for
each of these activities. I am thinking to use logit/probit type model and
control for site specific attributes. Do you think it is right way to
estimate the model? I am bit confused since the responses on these site
specific attributes are not binary, rather they are ordinal. I would great
appreciate your response.

Ramesh Ghimire
University of Georgia

View this message in context:
Sent from the Statalist mailing list archive at

*   For searches and help try:

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