17 Jun
2008
17 Jun
'08
3:31 p.m.
Hi.
I am trying to figure out two issues on Amelia II:
1) How to determine the q-order polynomials: The software allows one to specify
between 0-3. Since I am mostly interested in cross-section shifts, setting it
to 0 is not an option. But how would I determine to which order I'd set them
to? Also: in the Honacker and King (2006) article on missing values LOESS or
spline basis functions are recommended. Does Amelia II offer insights about
these two techniques as well?
2) How to set the case prior efficiently: I have 135 developing countries in my
data and would like to use a case prior to identify existing
similarities/differences between countries. Theoretically it makes sense to
group the countries by region/sub-region in terms of their similarity. Is it
possible to identify countries' similarities by regions/sub-regions rather than
making pairwise assessments? Also: the above-mentioned Honacker and King (2006)
article talks about observational priors but does not discuss the
advantages/drawbacks of case priors. Could you help me out here?
I look forward to your response.
Thanks very much!
Simone Dietrich
Ph.D. Candidate
The Pennsylvania State University
17 Jun
17 Jun
8:02 p.m.
SIMONE S DIETRICH wrote:
Hi.
I am trying to figure out two issues on Amelia II:
1) How to determine the q-order polynomials: The software allows one
to specify between 0-3. Since I am mostly interested in cross-section
shifts, setting it to 0 is not an option. But how would I determine to
which order I'd set them to? Also: in the Honacker and King (2006)
article on missing values LOESS or spline basis functions are
recommended. Does Amelia II offer insights about these two techniques
as well?
i'd start with the lower order polynomials and see whether they fit.
there are some examples of fit in the paper and some diagnostics in the
program.
2) How to set the case prior efficiently: I have 135 developing
countries in my data and would like to use a case prior to identify
existing similarities/differences between countries. Theoretically it
makes sense to group the countries by region/sub-region in terms of
their similarity. Is it possible to identify countries' similarities
by regions/sub-regions rather than making pairwise assessments? Also:
the above-mentioned Honacker and King (2006) article talks about
observational priors but does not discuss the advantages/drawbacks of
case priors. Could you help me out here?
not sure what you mean by 'case'
priors. but you might want to consider
adding in fixed effects for regions, which might get you what you want.
otherwise you can set the priors on individual cell values if you like.
Best of luck with your research,
Gary
---
http://gking.harvard.edu
I look forward to your response.
Thanks very much!
Simone Dietrich
Ph.D. Candidate
The Pennsylvania State University
-
Amelia mailing list served by Harvard-MIT Data Center
[Un]Subscribe/View Archive: http://lists.gking.harvard.edu/?info=amelia
9:02 p.m.
Hi Simone,
Just to be clear, the case priors are not implemented in Amelia II.
It's a feature that we planned on putting in, but it kept getting
pushed back. We're planning on disabling the "case priors" button.
Having said that, Gary's advice is right on. Adding a dummies for
groupings of cases that makes sense for your application. That might
be regions or trade partners or strategic allies.
good luck,
matt blackwell.
On Tue, Jun 17, 2008 at 8:02 PM, Gary King <king(a)harvard.edu> wrote:
SIMONE S DIETRICH wrote:
-
Amelia mailing list served by Harvard-MIT Data Center
[Un]Subscribe/View Archive: http://lists.gking.harvard.edu/?info=amelia
Hi.
I am trying to figure out two issues on Amelia II:
1) How to determine the q-order polynomials: The software allows one to
specify between 0-3. Since I am mostly interested in cross-section shifts,
setting it to 0 is not an option. But how would I determine to which order
I'd set them to? Also: in the Honacker and King (2006) article on missing
values LOESS or spline basis functions are recommended. Does Amelia II offer
insights about these two techniques as well?
i'd start with the lower order polynomials and see whether they fit. there
are some examples of fit in the paper and some diagnostics in the program.
2) How to set the case prior efficiently: I have 135 developing countries
in my data and would like to use a case prior to identify existing
similarities/differences between countries. Theoretically it makes sense to
group the countries by region/sub-region in terms of their similarity. Is it
possible to identify countries' similarities by regions/sub-regions rather
than making pairwise assessments? Also: the above-mentioned Honacker and
King (2006) article talks about observational priors but does not discuss
the advantages/drawbacks of case priors. Could you help me out here?
not sure what you mean by 'case' priors. but you might want to consider
adding in fixed effects for regions, which might get you what you want.
otherwise you can set the priors on individual cell values if you like.
Best of luck with your research,
Gary
---
http://gking.harvard.edu
I look forward to your response.
Thanks very much!
Simone Dietrich
Ph.D. Candidate
The Pennsylvania State University
-
Amelia mailing list served by Harvard-MIT Data Center
[Un]Subscribe/View Archive: http://lists.gking.harvard.edu/?info=amelia
5799
days inactive
5800
days old
amelia@lists.gking.harvard.edu
2 comments
3 participants
participants (3)
-
Gary King -
Matt Blackwell -
SIMONE S DIETRICH