1989
DOI: 10.2307/2111112
|View full text |Cite
|
Sign up to set email alerts
|

Vector Autoregression and the Study of Politics

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
131
0
2

Year Published

2009
2009
2023
2023

Publication Types

Select...
5
2
2

Relationship

1
8

Authors

Journals

citations
Cited by 219 publications
(133 citation statements)
references
References 29 publications
0
131
0
2
Order By: Relevance
“…For that reason, we need to look at multivariate time-series techniques, i.e., vector autogression (VAR). Vector autoregression involves the estimation of lags of different time series in the same equation (Sims 1980), and has been used in political science to study time series that interact with each other for several decades (Freeman, Williams, and Lin 1989). Our main innovation in this paper is to jointly fit a VAR and an IRT model so that the VAR can fully incorporate the measurement error present in the Twitter data.…”
Section: Modeling Ideological Polarization Over Timementioning
confidence: 99%
“…For that reason, we need to look at multivariate time-series techniques, i.e., vector autogression (VAR). Vector autoregression involves the estimation of lags of different time series in the same equation (Sims 1980), and has been used in political science to study time series that interact with each other for several decades (Freeman, Williams, and Lin 1989). Our main innovation in this paper is to jointly fit a VAR and an IRT model so that the VAR can fully incorporate the measurement error present in the Twitter data.…”
Section: Modeling Ideological Polarization Over Timementioning
confidence: 99%
“…It allows a researcher to specify a set of potentially causal variables in a predictive model without knowing the exact nature of the causal processes. A kind of 'everything-causes-everything' model that yields consistent time-series estimates (Freeman, Williams, and Lin 1989). It is particularly useful for forecasting how a change in one variable might lead to changes throughout the system if not feedback on itself.…”
Section: Vector Autoregressionmentioning
confidence: 99%
“…In the econometrics and time series literature there are three generally accepted options for analyzing such a dynamic system: a) simultaneous equation models, b) (vector) error correction models and c) vector autoregressions. Freeman, Williams and min Lin (1989) and Brandt and Williams (2007) outline the relative tradeoffs in the selection of these models. A simultaneous equation specification requires the analyst to make assumptions about the (weak) exogeneity of the variables in the model.…”
Section: Empirical Approachmentioning
confidence: 99%