Of course. The criteria are the same for everyone, though. I'm sure the test is skewed on way or another, but the point remains that none of these guys are anywhere close to where I am.
As someone who has done a bit of digging into the theoretical and philosophical justifications for NOMINATE let us say that the results are almost guaranteed by the process and don't really tell us things that we don't know. The primary reason for this is that while we may be able to get a two dimensional understanding of individuals and the relation of individuals we can't get a two dimensional understanding of policy.
To think of why the results are almost guaranteed, consider that voting is not a first order result from your preferences. I.E. when congress votes they do not necessarily vote because something is closer to their preferences than not, they vote because doing so is beneficial to them. That is, democrats tend to vote with democrats because even if they are not wholly on board, democrats will be on board with their ideas when its time to implement them. And similarly with Republicans. In fact, similarly with any system of winner take all voting. Taking the same system and analyzing Parliament and splitting the results between opposition and collation party we ought to see the same thing.
That being said, because the results do not take into account this problem when doing the two dimensional ranking it is still a (relatively) accurate predictor of results. Basically what it means is that each congress persons position on the graph depends on everyone else's, and if you moved one person to the left or right the entire spectrum would shift.
But being an accurate predictor of results(supposing that you can define the policy on the two dimensional axis correctly, which is entirely arbitrary and/or requires calibration of the second axis which invalidates the two dimensional construction) doesn't give us information about the candidate and whether we should vote for them in our district because of that last fact.
And similarly it does not give us information about ourselves if we were to be placed on that same scale.
To make it make a bit more sense. If we want to be able to determine what the "cutting line" is before hand we have to do traditional vote counting and look at each individuals preferences towards the bill. This negates the two dimensional analysis because it redefines the second dimension for each bill.
It is for this reason that when we see cutting lines what we are seeing is an optimization process run on the actual votes after they've been cast in order to minimize the error.
This is good in minimizing the error, but it only tells us anything about the bill after it has been voted on, since we don't have a good way of creating a cutting line from a priori information. I.E. we cannot look at a bill and say "ah yes, that obviously occupies this unique point in two dimensions and so we can draw our cutting line through that point on this axis"
That doesn't mean it isn't useful, but it says a lot less about things then a lot of people would like to think. Especially those who want to tell you that you are defined politically on a two dimensional axis and look you're really far away from those candidates you think you support.
![[community profile]](https://www.dreamwidth.org/img/silk/identity/community.png){kind=small}
![[identity profile]](https://www.dreamwidth.org/img/silk/identity/openid.png){kind=small}
![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png)
timeasmymeasure
no subject
Date: 2012-03-05 12:56 am (UTC)no subject
Date: 2012-03-05 12:57 am (UTC)no subject
Date: 2012-03-05 01:14 am (UTC)no subject
Date: 2012-03-05 01:30 am (UTC)http://en.wikipedia.org/wiki/NOMINATE_(scaling_method)
no subject
Date: 2012-03-05 03:12 am (UTC)To think of why the results are almost guaranteed, consider that voting is not a first order result from your preferences. I.E. when congress votes they do not necessarily vote because something is closer to their preferences than not, they vote because doing so is beneficial to them. That is, democrats tend to vote with democrats because even if they are not wholly on board, democrats will be on board with their ideas when its time to implement them. And similarly with Republicans. In fact, similarly with any system of winner take all voting. Taking the same system and analyzing Parliament and splitting the results between opposition and collation party we ought to see the same thing.
That being said, because the results do not take into account this problem when doing the two dimensional ranking it is still a (relatively) accurate predictor of results. Basically what it means is that each congress persons position on the graph depends on everyone else's, and if you moved one person to the left or right the entire spectrum would shift.
But being an accurate predictor of results(supposing that you can define the policy on the two dimensional axis correctly, which is entirely arbitrary and/or requires calibration of the second axis which invalidates the two dimensional construction) doesn't give us information about the candidate and whether we should vote for them in our district because of that last fact.
And similarly it does not give us information about ourselves if we were to be placed on that same scale.
no subject
Date: 2012-03-05 03:20 am (UTC)It is for this reason that when we see cutting lines what we are seeing is an optimization process run on the actual votes after they've been cast in order to minimize the error.
This is good in minimizing the error, but it only tells us anything about the bill after it has been voted on, since we don't have a good way of creating a cutting line from a priori information. I.E. we cannot look at a bill and say "ah yes, that obviously occupies this unique point in two dimensions and so we can draw our cutting line through that point on this axis"
That doesn't mean it isn't useful, but it says a lot less about things then a lot of people would like to think. Especially those who want to tell you that you are defined politically on a two dimensional axis and look you're really far away from those candidates you think you support.