quanteda (version 0.9.7-17)

textmodel_wordscores: Wordscores text model

Description

textmodel_wordscores implements Laver, Benoit and Garry's (2003) wordscores method for scaling of a single dimension. This can be called directly, but the recommended method is through textmodel.

Usage

textmodel_wordscores(data, scores, scale = c("linear", "logit"), smooth = 0)
"predict"(object, newdata = NULL, rescaling = "none", level = 0.95, verbose = TRUE, ...)
"print"(x, n = 30L, digits = 2, ...)
"show"(object)
"show"(object)
"print"(x, ...)

Arguments

data

the dfm on which the model will be fit. Does not need to contain only the training documents, since the index of these will be matched automatically.

scores

vector of training scores associated with each document identified in refData

scale

classic LBG linear posterior weighted word class differences, or logit scale of log posterior differences

smooth

a smoothing parameter for word counts; defaults to zero for the to match the LBG (2003) method.

object

a fitted Wordscores textmodel

newdata

dfm on which prediction should be made

rescaling

none for "raw" scores; lbg for LBG (2003) rescaling; or mv for the rescaling proposed by Martin and Vanberg (2007). (Note to authors: Provide full details here in documentation.)

level

probability level for confidence interval width

verbose

If TRUE, output status messages

...

additional arguments passed to other functions

x

for print method, the object to be printed

n

max rows of dfm to print

digits

number of decimal places to print for print methods

Value

The predict method for a wordscores fitted object returns a data.frame whose rows are the documents fitted and whose columns contain the scored textvalues, with the number of columns depending on the options called (for instance, how many rescaled scores, and whether standard errors were requested.) (Note: We may very well change this soon so that it is a list similar to other existing fitted objects.)

Slots

scale

linear or logit, according to the value of scale

Sw

the scores computed for each word in the training set

x

the dfm on which the wordscores model was called

y

the reference scores

call

the function call that fitted the model

method

takes a value of wordscores for this model

Details

Fitting a textmodel_wordscores results in an object of class textmodel_wordscores_fitted containing the following slots:

References

Laver, Michael, Kenneth R Benoit, and John Garry. 2003. "Extracting Policy Positions From Political Texts Using Words as Data." American Political Science Review 97(02): 311-31

Beauchamp, N. 2012. "Using Text to Scale Legislatures with Uninformative Voting." New York University Mimeo.

Martin, L W, and G Vanberg. 2007. "A Robust Transformation Procedure for Interpreting Political Text." Political Analysis 16(1): 93-100.

Laver, Michael, Kenneth R Benoit, and John Garry. 2003. "Extracting Policy Positions From Political Texts Using Words as Data." American Political Science Review 97(02): 311-31.

Martin, L W, and G Vanberg. 2007. "A Robust Transformation Procedure for Interpreting Political Text." Political Analysis 16(1): 93-100.

Examples

(ws <- textmodel(LBGexample, c(seq(-1.5, 1.5, .75), NA), model="wordscores"))
predict(ws)
predict(ws, rescaling="mv")
predict(ws, rescaling="lbg")

# same as:
(ws2 <- textmodel_wordscores(LBGexample, c(seq(-1.5, 1.5, .75), NA)))
predict(ws2)