Romeijn, Jan-Willem, Williamson, Jon (2018) Intervention and Identifiability in Latent Variable Modelling. Minds and Machines, 28 (2). pp. 243-264. ISSN 0924-6495. (doi:10.1007/s11023-018-9460-y) (KAR id:66599)
PDF
Publisher pdf
Language: English ![]() This work is licensed under a Creative Commons Attribution 4.0 International License. |
||
Download (527kB)
Preview
|
|
|
This file may not be suitable for user of assistive technology. Request an accessible format. |
||
PDF
Pre-print
Language: English |
||
Download (272kB)
Preview
|
|
|
This file may not be suitable for user of assistive technology. Request an accessible format. |
||
Official URL https://doi.org/10.1007/s11023-018-9460-y |
Abstract
We consider the use of interventions for resolving a problem of unidentified statistical models. The leading examples are from latent variable modelling, an influential statistical tool in the social sciences. We first explain the problem of statistical identifiability and contrast it with the identifiability of causal models. We then draw a parallel between the latent variable models and Bayesian networks with hidden nodes. This allows us to clarify the use of interventions for dealing with unidentified statistical models. We end by discussing the philosophical and methodological import of our result.
Item Type: | Article |
---|---|
DOI/Identification number: | 10.1007/s11023-018-9460-y |
Uncontrolled keywords: | Interventions, Statistical inference, Identifiability, Latent variable modelling |
Subjects: |
B Philosophy. Psychology. Religion > B Philosophy (General) B Philosophy. Psychology. Religion > BC Logic |
Divisions: | Divisions > Division of Arts and Humanities > School of European Culture and Languages |
Depositing User: | Jon Williamson |
Date Deposited: | 30 Mar 2018 14:18 UTC |
Last Modified: | 16 Feb 2021 13:53 UTC |
Resource URI: | https://kar.kent.ac.uk/id/eprint/66599 (The current URI for this page, for reference purposes) |
Williamson, Jon: | ![]() |
- Export to:
- RefWorks
- EPrints3 XML
- BibTeX
- CSV
- Depositors only (login required):