Asahi-Todai Elite Survey

Description

The Asahi-Todai Elite survey was conducted by the University of Tokyo in collaboration with a major national newspaper, the Asahi Shimbun, covering all candidates (both incumbents and challengers) for the eight Japanese Upper and Lower House elections that occurred between 2003 and 2013. In six out of eight waves, the survey was also administered to a nationally representative sample of voters with the sample size ranging from approximately 1,100 to about 2,000. The novel feature of the data is that there are a set of common policy questions, which can be used to scale both politicians and voters over time on the same dimension.

All together, the data set contains a total of N = 19,443 respondents, including 7,734 politicians and 11,709 voters. There are J = 98 unique questions in the survey, most of which consisted of questions asking for responses on a 5-point Likert scale. However, these scales were collapsed into a 3-point Likert scale for estimation with ordIRT(). In the data set, we include estimates obtained via MCMC using both the 3 and 5-point scale data. See Hirano et al. 2011 for more details.

Usage

data(AsahiTodai)

Format

list, containing the following variables:

dat.all

Survey data, formatted for input to ordIRT().

start.values

Start values, formatted for input to ordIRT().

priors

Priors, formatted for input to ordIRT().

ideal3

Ideal point estimates with data via MCMC, using collapsed 3-category data.

ideal5

Ideal point estimates with data via MCMC, using original 5-category data.

obs.attri

Attribute data of the respondents.

References

Shigeo Hirano, Kosuke Imai, Yuki Shiraito and Masaaki Taniguchi. 2011. “Policy Positions in Mixed Member Electoral Systems:Evidence from Japan.” Working Paper.

Kosuke Imai, James Lo, and Jonathan Olmsted (2016). “Fast Estimation of Ideal Points with Massive Data.” American Political Science Review, Vol. 110, No. 4 (December), pp. 631-656.

See Also

'ordIRT'.

Examples

## Not run: 
### Real data example: Asahi-Todai survey (not run)
## Collapses 5-category ordinal survey items into 3 categories for estimation
data(AsahiTodai)
out.varinf <- ordIRT(.rc = AsahiTodai$dat.all, .starts = AsahiTodai$start.values,
					.priors = AsahiTodai$priors, .D = 1,
					.control = {list(verbose = TRUE,
                     thresh = 1e-6, maxit = 500)})

## Compare against MCMC estimates using 3 and 5 categories
cor(ideal3, out.varinf$means$x)
cor(ideal5, out.varinf$means$x)  

## End(Not run)

Two-parameter Binary IRT estimation via EM

Description

binaryIRT estimates a binary IRT model with two response categories. Estimation is conducted using the EM algorithm described in the reference paper below. The algorithm will produce point estimates that are comparable to those of ideal, but will do so much more rapidly and also scale better with larger data sets.

Usage

  binIRT(.rc, .starts = NULL, .priors = NULL, .D = 1L, .control = NULL,
  .anchor_subject = NULL, .anchor_outcomes = FALSE)

Arguments

Value

An object of class binIRT.

iters

integer, number of iterations run.

conv

integer, convergence flag. Will return 1 if threshold reached, and 0 if maximum number of iterations reached.

threads

integer, number of threads used to estimated model.

tolerance

numeric, tolerance threshold for convergence. Identical to thresh argument in input to .control list.

Author(s)

Kosuke Imai imai@harvard.edu

James Lo lojames@usc.edu

Jonathan Olmsted jpolmsted@gmail.com

References

Kosuke Imai, James Lo, and Jonathan Olmsted “Fast Estimation of Ideal Points with Massive Data.” Working Paper. American Political Science Review, Vol. 110, No. 4 (December), pp. 631-656.

See Also

'convertRC', 'makePriors', 'getStarts'.

Examples

## Data from 109th US Senate
data(s109)

## Convert data and make starts/priors for estimation
rc <- convertRC(s109)
p <- makePriors(rc$n, rc$m, 1)
s <- getStarts(rc$n, rc$m, 1)

## Conduct estimates
lout <- binIRT(.rc = rc,
                .starts = s,
                .priors = p,
                .control = {
                    list(threads = 1,
                         verbose = FALSE,
                         thresh = 1e-6
                         )
                }
                )

## Look at first 10 ideal point estimates
lout$means$x[1:10]


lout2 <- binIRT(.rc = rc,
                .starts = s,
                .priors = p,
                .control = {
                    list(threads = 1,
                         verbose = FALSE,
                         thresh = 1e-6
                         )
                },
                .anchor_subject = 2
                )
                                        # Rotates so that Sen. Sessions (R AL)
                                        # has more of the estimated trait

lout3 <- binIRT(.rc = rc,
                .starts = s,
                .priors = p,
                .control = {
                    list(threads = 1,
                         verbose = FALSE,
                         thresh = 1e-6
                         )
                },
                .anchor_subject = 10
                )
                                        # Rotates so that Sen. Boxer (D CA)
                                        # has more of the estimated trait

cor(lout2$means$x[, 1],
    lout3$means$x[, 1]
    )
                                        # = -1 --> same numbers, flipped
                                        # orientation

Parametric bootstrap of EM Standard Errirs

Description

boot_emIRT take an emIRT() object (from binary, ordinal, dynamic, or hierarchical models) and implements a parametric bootstrap of the standard errors for the ideal points. It assumes you have already run the model successfully on one of those functions, and takes the output from that estimate, along with the original arguments, as the arguments for the bootstrap function.

Usage

  boot_emIRT(emIRT.out, .data, .starts, .priors, .control, Ntrials=50, verbose=10)

Arguments

Value

An object of class emIRT. The output takes the original emIRT.out object and appends the following:

Author(s)

Kosuke Imai kimai@princeton.edu

James Lo jameslo@princeton.edu

Jonathan Olmsted jpolmsted@gmail.com

References

Kosuke Imai, James Lo, and Jonathan Olmsted (2016). “Fast Estimation of Ideal Points with Massive Data.” American Political Science Review, Vol. 110, No. 4 (December), pp. 631-656.

See Also

'binIRT', 'ordIRT', 'hierIRT', 'dynIRT'.

Examples

## Not run: 

### Binary IRT example
example(binIRT)
boot.bin <- boot_emIRT(lout, .data = rc, .starts = s, .priors = p,
  .control = list(threads = 1, verbose = FALSE, thresh = 1e-06), Ntrials=10, verbose=2)
boot.bin$bse$x

### Dynamic IRT example
example(dynIRT)
boot.dyn <- boot_emIRT(lout, .data = mq_data$data.mq, .starts = mq_data$cur.mq,
  .priors = mq_data$priors.mq, .control = list(threads = 1, verbose = FALSE,
  thresh = 1e-06), Ntrials=10, verbose=2)
boot.dyn$bse$x

### Ordinal IRT example
example(ordIRT)
boot.ord <- boot_emIRT(lout, .data=newrc, .starts=cur, .priors=priors,
  .control = list(threads = 1, verbose = TRUE, thresh = 1e-6, maxit=300,
  checkfreq=50), Ntrials=5, verbose=1)
boot.ord$bse$x

### Hierarhical IRT example
example(hierIRT, run.dontrun=TRUE)
boot.hier <- boot_emIRT((lout, .data=dwnom$data.in, .starts=dwnom$cur, .priors=dwnom$priors,
	.control=list(threads = 8, verbose = TRUE, thresh = 1e-4, maxit=200, checkfreq=1),
	Ntrials=5, verbose=1)
boot.hier$bse$x_implied


## End(Not run)

Convert Roll Call Matrix Format

Description

convertRC takes a rollcall object and converts it into a format suitable for estimation with 'binIRT'.

Usage

  convertRC(.rc, type = "binIRT") 

Arguments

Value

An object of class rollcall, with votes recoded such that yea=1, nay=-1, missing data = 0.

Author(s)

Kosuke Imai kimai@princeton.edu

James Lo jameslo@princeton.edu

Jonathan Olmsted jpolmsted@gmail.com

References

Kosuke Imai, James Lo, and Jonathan Olmsted “Fast Estimation of Ideal Points with Massive Data.” Working Paper. Available at http://imai.princeton.edu/research/fastideal.html.

See Also

'binIRT', 'makePriors', 'getStarts'.

Examples

## Data from 109th US Senate
data(s109)

## Convert data and make starts/priors for estimation
rc <- convertRC(s109)
p <- makePriors(rc$n, rc$m, 1)
s <- getStarts(rc$n, rc$m, 1)

## Conduct estimates
lout <- binIRT(.rc = rc,
                .starts = s,
                .priors = p,
                .control = {
                    list(threads = 1,
                         verbose = FALSE,
                         thresh = 1e-6
                         )
                }
                )

## Look at first 10 ideal point estimates
lout$means$x[1:10]

Poole-Rosenthal DW-NOMINATE data and scores, 80-110 U.S. Senate

Description

This data set contains materials related to the Poole-Rosenthal DW-NOMINATE measure of senator ideology. The software (and other materials) is available at https://voteview.com/, which includes a simpler example of an application to the 80-110 U.S. Senate. The data set here is derived from the data and estimates from that example, but are formatted to be run in hierIRT(). In particular, start values for estimation are identical to those provided by the example.

Usage

data(dwnom)

Format

list, containing the following elements:

data.in

Legislator voting data, formatted for input to hierIRT().

cur

Start values, formatted for input to hierIRT().

priors

Priors, formatted for input to hierIRT().

legis

data frame, containing contextual information about the legislators estimated.

nomres

data frame, containing estimates from DW-NOMINATE on the same data. These are read from the file SL80110C21.DAT.

References

DW-NOMINATE is described in Keith T. Poole and Howard Rosenthal. 1997. Congress: A Political Economic History of Roll Call Voting. Oxford University Press. See also https://voteview.com/.

Variational model is described in Kosuke Imai, James Lo, and Jonathan Olmsted (2016). “Fast Estimation of Ideal Points with Massive Data.” American Political Science Review, Vol. 110, No. 4 (December), pp. 631-656.

See Also

'hierIRT'.

Examples

### Real data example of US Senate 80-110 (not run)
### Based on voteview.com example of DW-NOMINATE 
### We estimate a hierarchical model without noise and a linear time covariate
### This model corresponds very closely to the DW-NOMINATE model

## Not run: 
data(dwnom)

## This takes about 10 minutes to run on 8 threads
## You may need to reduce threads depending on what your machine can support
lout <- hierIRT(.data = dwnom$data.in,
                    .starts = dwnom$cur,
                    .priors = dwnom$priors,
                    .control = {list(
                    threads = 8,
                    verbose = TRUE,
                    thresh = 1e-4,
				    maxit=200,
				    checkfreq=1
                        )})

## Bind ideal point estimates back to legislator data
final <- cbind(dwnom$legis, idealpt.hier=lout$means$x_implied)

## These are estimates from DW-NOMINATE as given on the Voteview example
## From file "SL80110C21.DAT"
nomres <- dwnom$nomres

## Merge the DW-NOMINATE estimates to model results by legislator ID
## Check correlation between hierIRT() and DW-NOMINATE scores
res <- merge(final, nomres, by=c("senate","id"),all.x=TRUE,all.y=FALSE)
cor(res$idealpt.hier, res$dwnom1d)


## End(Not run)

Dynamic IRT estimation via Variational Inference

Description

ordIRT estimates an dynamic IRT model with two response categories per item, over several sessions. Ideal points over time follow a random walk prior, and the model originates from the work of Martin and Quinn (2002). Estimation is conducted using the variational EM algorithm described in the reference paper below. The algorithm will produce point estimates that are comparable to those of MCMCdynamicIRT1d, but will do so much more rapidly and also scale better with larger data sets.

Usage

  dynIRT(.data, .starts = NULL, .priors = NULL, .control = NULL) 

Arguments

Value

An object of class dynIRT.

iters

integer, number of iterations run.

conv

integer, convergence flag. Will return 1 if threshold reached, and 0 if maximum number of iterations reached.

threads

integer, number of threads used to estimated model.

tolerance

numeric, tolerance threshold for convergence. Identical to thresh argument in input to .control list.

Author(s)

Kosuke Imai imai@harvard.edu

James Lo lojames@usc.edu

Jonathan Olmsted jpolmsted@gmail.com

References

Original model and the example is based off of Andrew Martin and Kevin Quinn, “Dynamic Ideal Point Estimation via Markov Chain Monte Carlo for the U.S. Supreme Court, 1953-1999.” Political Analysis 10(2) 134-153.

Variational model is described in Kosuke Imai, James Lo, and Jonathan Olmsted (2016). “Fast Estimation of Ideal Points with Massive Data.” American Political Science Review, Vol. 110, No. 4 (December), pp. 631-656.

See Also

'mq_data'.

Examples

### Replication of Martin-Quinn Judicial Ideology Scores
### Based on July 23, 2014 (2014 Release 01) release of the Supreme Court Database
### Start values and priors based on replication code provided by Kevin Quinn

data(mq_data)

## Estimate dynamic variational model using dynIRT()
lout <- dynIRT(.data = mq_data$data.mq,
                    .starts = mq_data$cur.mq,
                    .priors = mq_data$priors.mq,
                    .control = {list(
                    threads = 1,
                    verbose = TRUE,
                    thresh = 1e-6,
		    maxit=500
                    )})

## Extract estimate from variational model
## Delete point estimates of 0, which are justices missing from that session
vi.out <- c(t(lout$means$x))
vi.out[vi.out==0] <- NA
vi.out <- na.omit(vi.out)

## Compare correlation against MCMC-estimated result
## Correlates at r=0.93 overall, and 0.96 when excluding Douglas
cor(vi.out, mq_data$mq_mcmc)	
cor(vi.out[mq_data$justiceName != "Douglas"],
 mq_data$mq_mcmc[mq_data$justiceName != "Douglas"])	

Generate Starts for binIRT

Description

getStarts generates starting values for binIRT.

Usage

  getStarts(.N, .J, .D, .type = "zeros")

Arguments

Value

Author(s)

Kosuke Imai kimai@princeton.edu

James Lo jameslo@princeton.edu

Jonathan Olmsted jpolmsted@gmail.com

References

Kosuke Imai, James Lo, and Jonathan Olmsted “Fast Estimation of Ideal Points with Massive Data.” Working Paper. Available at http://imai.princeton.edu/research/fastideal.html.

See Also

'binIRT', 'makePriors', 'convertRC'.

Examples

## Data from 109th US Senate
data(s109)

## Convert data and make starts/priors for estimation
rc <- convertRC(s109)
p <- makePriors(rc$n, rc$m, 1)
s <- getStarts(rc$n, rc$m, 1)

## Conduct estimates
lout <- binIRT(.rc = rc,
                .starts = s,
                .priors = p,
                .control = {
                    list(threads = 1,
                         verbose = FALSE,
                         thresh = 1e-6
                         )
                }
                )

## Look at first 10 ideal point estimates
lout$means$x[1:10]

Hierarchichal IRT estimation via Variational Inference

Description

hierIRT estimates an hierarchical IRT model with two response categories, allowing the use of covariates to help determine ideal point estimates. Estimation is conducted using the variational EM algorithm described in the reference paper below. A special case of this model occurs when time/session is used as the covariate — this allows legislator ideal points to vary over time with a parametric time trend. Notably, the popular DW-NOMINATE model (Poole and Rosenthal, 1997) is one such example, in which legislator ideal points shift by a constant amount each period, and the error term in the hierarchical model is set to 0. In contrast to other functions in this package, this model does not assume a ‘rectangular’ roll call matrix, and all data are stored in vector form.

Usage

  hierIRT(.data, .starts = NULL, .priors = NULL, .control = NULL) 

Arguments

Value

An object of class hierIRT.

iters

integer, number of iterations run.

conv

integer, convergence flag. Will return 1 if threshold reached, and 0 if maximum number of iterations reached.

threads

integer, number of threads used to estimated model.

tolerance

numeric, tolerance threshold for convergence. Identical to thresh argument in input to .control list.

D

integer, number of dimensions (i.e. number of covariates, including intercept).

G

integer, number of groups.

I

integer, number of ideal points.

J

integer, number of items/bill parameters.

L

integer, number of observed votes.

Author(s)

Kosuke Imai imai@harvard.edu

James Lo lojames@usc.edu

Jonathan Olmsted jpolmsted@gmail.com

References

Variational model is described in Kosuke Imai, James Lo, and Jonathan Olmsted “Fast Estimation of Ideal Points with Massive Data.” American Political Science Review, Volume 110, Issue 4, November 2016, pp. 631-656. <DOI:10.1017/S000305541600037X>.

See Also

'dwnom'.

Examples

### Real data example of US Senate 80-110 (not run)
### Based on voteview.com example of DW-NOMINATE (\url{https://voteview.com/})
### We estimate a hierarchical model without noise and a linear time covariate
### This model corresponds very closely to the DW-NOMINATE model

## Not run: 
data(dwnom)

## This takes about 10 minutes to run on 8 threads
## You may need to reduce threads depending on what your machine can support
lout <- hierIRT(.data = dwnom$data.in,
                    .starts = dwnom$cur,
                    .priors = dwnom$priors,
                    .control = {list(
                    threads = 8,
                    verbose = TRUE,
                    thresh = 1e-4,
				    maxit=200,
				    checkfreq=1
                        )})

## Bind ideal point estimates back to legislator data
final <- cbind(dwnom$legis, idealpt.hier=lout$means$x_implied)

## These are estimates from DW-NOMINATE as given on the Voteview example
## From file "SL80110C21.DAT"
nomres <- dwnom$nomres

## Merge the DW-NOMINATE estimates to model results by legislator ID
## Check correlation between hierIRT() and DW-NOMINATE scores
res <- merge(final, nomres, by=c("senate","id"),all.x=TRUE,all.y=FALSE)
cor(res$idealpt.hier, res$dwnom1d)


## End(Not run)

Generate Priors for binIRT

Description

makePriors generates diffuse priors for binIRT.

Usage

  makePriors(.N = 20, .J = 100, .D = 1)

Arguments

Value

x$mu

A (D x D) prior means matrix for respondent ideal points x_i.

x$sigma

A (D x D) prior covariance matrix for respondent ideal points x_i.

beta$mu

A ( D+1 x 1) prior means matrix for \alpha_j and \beta_j.

beta$sigma

A ( D+1 x D+1 ) prior covariance matrix for \alpha_j and \beta_j.

Author(s)

Kosuke Imai imai@harvard.edu

James Lo jameslo@princeton.edu

Jonathan Olmsted jpolmsted@gmail.com

References

Kosuke Imai, James Lo, and Jonathan Olmsted. (2016). “Fast Estimation of Ideal Points with Massive Data.” Working Paper. American Political Science Review, Vol. 110, No. 4 (December), pp. 631-656.

See Also

'binIRT', 'getStarts', 'convertRC'.

Examples

## Data from 109th US Senate
data(s109)

## Convert data and make starts/priors for estimation
rc <- convertRC(s109)
p <- makePriors(rc$n, rc$m, 1)
s <- getStarts(rc$n, rc$m, 1)

## Conduct estimates
lout <- binIRT(.rc = rc,
                .starts = s,
                .priors = p,
                .control = {
                    list(threads = 1,
                         verbose = FALSE,
                         thresh = 1e-6
                         )
                }
                )

## Look at first 10 ideal point estimates
lout$means$x[1:10]

German Manifesto Data

Description

A word frequency matrix containing word frequencies from 25 German party manifestos between 1990-2005. Obtained from Slapin and Proksch AJPS paper, also used in Lo, Slapin and Proksch.

Usage

data(manifesto)

Format

list, containing the following elements:

data.manif

Term-document matrix, formatted for input to ordIRT().

starts.manif

Start values, formatted for input to ordIRT().

priors.manif

Priors, formatted for input to ordIRT().

References

Jonathan Slapin and Sven-Oliver Proksch. 2009. “A Scaling Model for Estimating Time-Series Party Positions from Texts.” American Journal of Political Science 52(3), 705-722

James Lo, Jonathan Slapin, and Sven-Oliver Proksch. 2016. “Ideological Clarify in Multiparty Competition: A New Measure and Test Using Election Manifestos.” British Journal of Political Science, 1-20

Kosuke Imai, James Lo, and Jonathan Olmsted. (2016). “Fast Estimation of Ideal Points with Massive Data.” American Political Science Review, Vol. 110, No. 4 (December), pp. 631-656.

See Also

'poisIRT'.

Examples

## Not run: 
## Load German Manifesto data
data(manifesto)

## Estimate variational Wordfish model
lout <- poisIRT(.rc = manifesto$data.manif,
		    i = 0:(ncol(manifesto$data.manif)-1),
		    NI=ncol(manifesto$data.manif),
                    .starts = manifesto$starts.manif,
                    .priors = manifesto$priors.manif,
                    .control = {list(
                        threads = 1,
                        verbose = TRUE,
                        thresh = 1e-6,
						maxit=1000
                        )})

## Positional Estimates for Parties
lout$means$x  

## End(Not run)

Martin-Quinn Judicial Ideology Scores

Description

This data set contains materials related to the Martin-Quinn measures of judicial ideology, estimated for every justice serving from October 1937 to October 2013. The materials are based on the July 23, 2014 (2014 Release 01) release of the Supreme Court Database, which contain the votes of each Supreme Court justice on each case heard in the court. The data is set up for input to dynIRT(), and also includes point estimates of the same model obtained using standard MCMC techniques. Start values and priors input to this model are identical to those used in the MCMC estimates, and were provided by Kevin Quinn.

Usage

data(mq_data)

Format

list, containing the following elements:

data.mq

Justice voting data, formatted for input to dynIRT().

cur.mq

Start values, formatted for input to dynIRT().

priors.mq

Priors, formatted for input to dynIRT().

mq_mcmc

Ideal point estimates with data via MCMC.

justiceName

A vector of names identifying the justice that goes with each estimated ideal point.

References

Original model and the example is based off of Andrew Martin and Kevin Quinn, “Dynamic Ideal Point Estimation via Markov Chain Monte Carlo for the U.S. Supreme Court, 1953-1999.” Political Analysis 10(2) 134-153.

Variational model is described in Kosuke Imai, James Lo, and Jonathan Olmsted (2016). “Fast Estimation of Ideal Points with Massive Data.” American Political Science Review, Vol. 110, No. 4 (December), pp. 631-656.

See Also

'dynIRT'.

Examples

### Replication of Martin-Quinn Judicial Ideology Scores
### Based on July 23, 2014 (2014 Release 01) release of the Supreme Court Database
### Start values and priors based on replication code provided by Kevin Quinn

data(mq_data)

## Estimate dynamic variational model using dynIRT()
lout <- dynIRT(.data = mq_data$data.mq,
                    .starts = mq_data$cur.mq,
                    .priors = mq_data$priors.mq,
                    .control = {list(
                    threads = 1,
                    verbose = TRUE,
                    thresh = 1e-6,
		    maxit=500
                    )})

## Extract estimate from variational model
## Delete point estimates of 0, which are justices missing from that session
vi.out <- c(t(lout$means$x))
vi.out[vi.out==0] <- NA
vi.out <- na.omit(vi.out)

## Compare correlation against MCMC-estimated result
## Correlates at r=0.93 overall, and 0.96 when excluding Douglas
cor(vi.out, mq_data$mq_mcmc)	
cor(vi.out[mq_data$justiceName != "Douglas"],
 mq_data$mq_mcmc[mq_data$justiceName != "Douglas"])	

Network IRT estimation via EM

Description

networkIRT estimates an IRT model with network in cells. Estimation is conducted using the EM algorithm described in the reference paper below. The algorithm generalizes a model by Slapin and Proksch (2009) that is commonly applied to manifesto data.

Usage

   networkIRT(.y, .starts = NULL, .priors = NULL, .control = NULL,
    .anchor_subject = NULL, .anchor_item = NULL)

Arguments

Value

An object of class networkIRT.

iters

integer, number of iterations run.

conv

integer, convergence flag. Will return 1 if threshold reached, and 0 if maximum number of iterations reached.

threads

integer, number of threads used to estimated model.

tolerance

numeric, tolerance threshold for convergence. Identical to thresh argument in input to .control list.

Author(s)

Kosuke Imai imai@Harvard.edu

James Lo jameslo@princeton.edu

Jonathan Olmsted jpolmsted@gmail.com

References

Kosuke Imai, James Lo, and Jonathan Olmsted (2016). “Fast Estimation of Ideal Points with Massive Data.” American Political Science Review, Vol. 110, No. 4 (December), pp. 631-656.

See Also

'ustweet'

Examples

## Not run: 
data(ustweet)

## A ridiculously short run to pass CRAN
## For a real test, set maxit to a more reasonable number to reach convergence
lout <- networkIRT(.y = ustweet$data,
                   .starts = ustweet$starts,
                   .priors = ustweet$priors,
                   .control = {list(verbose = TRUE,
                                    maxit = 3,
                                    convtype = 2,
                                    thresh = 1e-6,
                                    threads = 1
                                    )
                           },
                   .anchor_item = 43
                   )


## End(Not run)

Two-parameter Ordinal IRT estimation via EM

Description

ordIRT estimates an ordinal IRT model with three ordered response categories. Estimation is conducted using the EM algorithm described in the reference paper below. The algorithm will produce point estimates that are comparable to those of MCMCordfactanal, but will do so much more rapidly and also scale better with larger data sets.

Usage

  ordIRT(.rc, .starts = NULL, .priors = NULL, .D = 1L, .control = NULL) 

Arguments

Value

An object of class ordIRT.

iters

integer, number of iterations run.

conv

integer, convergence flag. Will return 1 if threshold reached, and 0 if maximum number of iterations reached.

threads

integer, number of threads used to estimated model.

tolerance

numeric, tolerance threshold for convergence. Identical to thresh argument in input to .control list.

Author(s)

Kosuke Imai imai@Harvard.edu

James Lo jameslo@princeton.edu

Jonathan Olmsted jpolmsted@gmail.com

References

Kosuke Imai, James Lo, and Jonathan Olmsted (2016). “Fast Estimation of Ideal Points with Massive Data.” American Political Science Review, Vol. 110, No. 4 (December), pp. 631-656.

See Also

'AsahiTodai'.

Examples

## Not run: 
### Real data example: Asahi-Todai survey (not run)
## Collapses 5-category ordinal survey items into 3 categories for estimation
data(AsahiTodai)
out.varinf <- ordIRT(.rc = AsahiTodai$dat.all, .starts = AsahiTodai$start.values,
					.priors = AsahiTodai$priors, .D = 1,
					.control = {list(verbose = TRUE,
                     thresh = 1e-6, maxit = 500)})

## Compare against MCMC estimates using 3 and 5 categories
cor(ideal3, out.varinf$means$x)
cor(ideal5, out.varinf$means$x)  

## End(Not run)


### Monte Carlo simulation of ordIRT() model vs. known parameters
## Set number of legislators and items
set.seed(2)
NN <- 500
JJ <- 100

## Simulate true parameters from original model
x.true <- runif(NN, -2, 2)
beta.true <- runif(JJ, -1, 1)
tau1 <- runif(JJ, -1.5, -0.5)
tau2 <- runif(JJ, 0.5, 1.5)
ystar <- x.true %o% beta.true + rnorm(NN *JJ)

## These parameters are not needed, but correspond to reparameterized model
#d.true <- tau2 - tau1
#dd.true <- d.true^2
#tau_star <- -tau1/d.true
#beta_star <- beta.true/d.true

## Generate roll call matrix using simulated parameters
newrc <- matrix(0, NN, JJ)
for(j in 1:JJ) newrc[,j] <- cut(ystar[,j], c(-100, tau1[j], tau2[j],100), labels=FALSE)

## Generate starts and priors
cur <- vector(mode = "list")
cur$DD <- matrix(rep(0.5,JJ), ncol=1)
cur$tau <- matrix(rep(-0.5,JJ), ncol=1)
cur$beta <- matrix(runif(JJ,-1,1), ncol=1) 
cur$x <- matrix(runif(NN,-1,1), ncol=1) 
priors <- vector(mode = "list")
priors$x <- list(mu = matrix(0,1,1), sigma = matrix(1,1,1) )
priors$beta <- list(mu = matrix(0,2,1), sigma = matrix(diag(25,2),2,2))

## Call ordIRT() with inputs
time <- system.time({
    lout <- ordIRT(.rc = newrc,
                    .starts = cur,
                    .priors = priors,
                    .control = {list(
                        threads = 1,
                        verbose = TRUE,
                        thresh = 1e-6,
			maxit=300,
			checkfreq=50
                        )})
})

## Examine runtime and correlation of recovered ideal points vs. truth
time
cor(x.true,lout$means$x)

Poisson IRT estimation via EM

Description

poisIRT estimates an IRT model with count (usually word counts) in cells. Estimation is conducted using the EM algorithm described in the reference paper below. The algorithm generalizes a model by Slapin and Proksch (2009) that is commonly applied to manifesto data.

Usage

    poisIRT(.rc, i = 0:(nrow(.rc)-1), NI = nrow(.rc), .starts = NULL, .priors = NULL,
    .control = NULL)

Arguments

Value

An object of class poisIRT.

Author(s)

Kosuke Imai imai@harvard.edu

James Lo jameslo@princeton.edu

Jonathan Olmsted jpolmsted@gmail.com

References

Kosuke Imai, James Lo, and Jonathan Olmsted (2016). “Fast Estimation of Ideal Points with Massive Data.” American Political Science Review, Vol. 110, No. 4 (December), pp. 631-656.

See Also

'manifesto'

Examples

## Not run: 
## Load German Manifesto data
data(manifesto)

## Estimate variational Wordfish model
lout <- poisIRT(.rc = manifesto$data.manif,
		    i = 0:(ncol(manifesto$data.manif)-1),
		    NI=ncol(manifesto$data.manif),
                    .starts = manifesto$starts.manif,
                    .priors = manifesto$priors.manif,
                    .control = {list(
                        threads = 1,
                        verbose = TRUE,
                        thresh = 1e-6,
						maxit=1000
                        )})

## Positional Estimates for Parties
lout$means$x  

## End(Not run)

U.S. Twitter Following Data

Description

Data from U.S. Twitter follower data, obtained from Barbera's (2015) replication archive.

Usage

data(ustweet)

Format

list, containing the following elements:

data

Term-document matrix, formatted for input to networkIRT().

starts

Start values, formatted for input to networkIRT().

priors

Priors, formatted for input to networkIRT().

References

Pablo Barbera. 2015. “Birds of the Same Feather Tweet Together: Bayesian Ideal Point Estimation Using Twitter Data.” Political Analysis 23(1), 76-91

Kosuke Imai, James Lo, and Jonathan Olmsted. (2016). “Fast Estimation of Ideal Points with Massive Data.” American Political Science Review, Vol. 110, No. 4 (December), pp. 631-656.

See Also

'networkIRT'.

Examples

## Not run: 
data(ustweet)

## A ridiculously short run to pass CRAN
## For a real test, set maxit to a more reasonable number to reach convergence
lout <- networkIRT(.y = ustweet$data,
                   .starts = ustweet$starts,
                   .priors = ustweet$priors,
                   .control = {list(verbose = TRUE,
                                    maxit = 3,
                                    convtype = 2,
                                    thresh = 1e-6,
                                    threads = 1
                                    )
                           },
                   .anchor_item = 43
                   )


## End(Not run)