Weights Parameter Output

[acarrizosa]

Hello,
I'm writing with a question more than an issue exactly. I'm wondering whether the outcome of your dwnominate procedure stores the estimated weight parameter anywhere. I can't seem to find it. I know that the default starting value for the weight parameter is w = 0.3463, and in the stored OC or common_space objects used as starting points I can also find the "fits" parameters which can be used for the weights as well, but both these values describe the starting points for the weights parameter, I think, and not the final weight parameter estimated by dwnominate.

So my questions are two:
First, is the second dimension of the dwnominate output already weighted? That is, has results$legislators$coord2D already been multiplied by the output weight parameter?

Second, is there a way to find the output weight parameter after dwnominate has been carried out, or should I simply use results$start$fits to create weights after the estimation?

The main reason I am asking this is because I am trying to estimate cutting lines for the dwnominate estimates. To do so both coord2D and certain bill-level parameters have to be multiplied by the weight parameter, and while looking through the code I was unsure whether coord2D was already multiplied by weights, and where the weight parameter can be found to multiply it by certain bill-level parameters.

Thank you and best,
Andy Carrizosa

[wmay]

No, the coordinates aren't weighted. They're the standard DW-NOMINATE coordinates, which are scaled to fit between -1 and 1 (more or less).

Right now the function doesn't return the weight parameter, but I agree that it should and it shouldn't be hard to add that feature. I'll see if I can do this in the next couple of days.

[acarrizosa]

I have been following the instructions laid out in the book "Analyzing Spatial Models of Choice and Judgment with R," written by Armstrong et al. (2014). On pages 203-204 they show a procedure for estimating cutting lines that includes the use of the weight paramenter.

[acarrizosa]

Sorry to bother you again, but would it also be possible to make the function return the beta parameter as well when you modify it?

Thanks and best,
Andy

[wmay]

I'm returning the values in the same format as wnominate's nomObject, so with a bit of finagling the results can be used with wnominate::plot.coords:

class(dwnom_result) <- "nomObject"
wnominate::plot.coords(dwnom_result)