Wordfish

Wordfish is a Poisson scaling model of one-dimensional document positions (Slapin and Proksch 2008). Wordfish also allows for scaling documents, but compared to Wordscores reference scores/texts are not required. Wordfish is an unsupervised one-dimensional text scaling method, meaning that it estimates the positions of documents solely based on the observed word frequencies.

require(quanteda)

In this example, we show how to apply Wordfish to the Irish budget speeches from 2010. First, we create dfm, afterwards we run Wordfish.

dfmat_irish <- dfm(data_corpus_irishbudget2010, remove_punct = TRUE)
tmod_wf <- textmodel_wordfish(dfmat_irish, dir = c(6,5))
summary(tmod_wf)

## 
## Call:
## textmodel_wordfish.dfm(x = dfmat_irish, dir = c(6, 5))
## 
## Estimated Document Positions:
##                              theta      se
## Lenihan, Brian (FF)        1.79444 0.02010
## Bruton, Richard (FG)      -0.61774 0.02844
## Burton, Joan (LAB)        -1.14741 0.01561
## Morgan, Arthur (SF)       -0.08380 0.02900
## Cowen, Brian (FF)          1.77416 0.02333
## Kenny, Enda (FG)          -0.75762 0.02642
## ODonnell, Kieran (FG)     -0.48645 0.04310
## Gilmore, Eamon (LAB)      -0.59455 0.02991
## Higgins, Michael (LAB)    -0.99302 0.04021
## Quinn, Ruairi (LAB)       -0.90657 0.04267
## Gormley, John (Green)      1.18326 0.07235
## Ryan, Eamon (Green)        0.17248 0.06336
## Cuffe, Ciaran (Green)      0.72229 0.07269
## OCaolain, Caoimhghin (SF) -0.05949 0.03873
## 
## Estimated Feature Scores:
##         when      i presented    the supplementary  budget     to  this
## beta -0.1558 0.3217    0.3582 0.1945         1.077 0.03563 0.3097 0.249
## psi   1.6246 2.7253   -1.7925 5.3324        -1.128 2.71082 4.5208 3.462
##       house   last   april    said     we   could   work    our    way
## beta 0.1461 0.2416 -0.1554 -0.8301 0.4193 -0.6067 0.5262 0.6918 0.2772
## psi  1.0407 0.9874 -0.5725 -0.4533 3.5140  1.0865 1.1164 2.5301 1.4208
##      through  period     of severe economic distress  today    can  report
## beta  0.6125  0.4989 0.2792  1.229   0.4245    1.804 0.0922 0.3057  0.6257
## psi   1.1636 -0.1747 4.4675 -2.007   1.5741   -4.457 0.8399 1.5663 -0.2466
##        that notwithstanding difficulties   past
## beta 0.0192           1.804        1.176 0.4777
## psi  3.8389          -4.457       -1.352 0.9339

We can plot the results of a fitted scaling model using textplot_scale1d().

textplot_scale1d(tmod_wf)

The function also allows to plot scores by a grouping variable, in this case the party affiliation of the speakers.

textplot_scale1d(tmod_wf, groups = docvars(dfmat_irish, "party"))

Finally, we can plot the estimated word positions and highlight certain features.

textplot_scale1d(tmod_wf, margin = "features", 
                 highlighted = c("government", "global", "children", 
                                 "bank", "economy", "the", "citizenship",
                                 "productivity", "deficit"))

If you want to learn more about Wordfish, see:
Slapin, Jonathan and Sven-Oliver Proksch. 2008. “A Scaling Model for Estimating Time-Series Party Positions from Texts.” American Journal of Political Science 52(3): 705-772.