Planned maintenance
A system upgrade is planned for 10/12-2024, at 12:00-13:00. During this time DiVA will be unavailable.
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Model averaging and variable selection methods for causal models
Jönköping University, Jönköping International Business School, JIBS, Statistics.ORCID iD: 0000-0003-0279-5305
2024 (English)Doctoral thesis, comprehensive summary (Other academic)
Abstract [en]

This doctoral dissertation comprises four papers. The aims of this dissertation are twofold: first, it introduces model averaging methods that use weights inversely proportional to the Hausman test statistic; second, it explores variable selection methods for estimating causal effects by identifying both valid and invalid instruments. Each paper assesses the performance of these novel methods by developing theoretical asymptotic properties, conducting Monte Carlo simulations, and applying them empirically.

The first paper focuses on the use of instrumental variable estimation methods regression with many instruments in the linear model. A weighted average approach is introduced by combining least squares and limited information maximum likelihood estimators.

The second paper addresses the issue of endogeneity in the instrumental variable probit model by developing a Stein weighted average control function maximum likelihood estimator. The asymptotic distribution and asymptotic risk of the proposed estimator are derived.

The third paper focuses on the use of Lasso instrumental variables estimation methods. In addition, the jackknife instrumental variable approach is introduced using the Lasso procedure. The proposed methods are robust for estimating causal effects in the presence of both invalid and valid instruments. Additionally, for convenience, we created an R package for implementing the proposed methods.

The fourth paper introduces the best subset instrumental estimator via mixed integer optimization to estimate causal effects and select invalid instruments. It is shown that the best subset instrumental variable estimator outperforms the two-stage least squares, Lasso-type instrumental variables methods, and two-sample analysis methods.

Abstract [sv]

Denna doktorsavhandling består av fyra artiklar. Avhandlingens syften är tvåfaldiga: för det första introduceras metoder som tar ett genomsnitt baserat på vikter som är proportionella till inversen av ett Hausman-test; för det andra undersöks metoder för val av variabler för att skatta kausala effekter genom att identifiera både giltiga och ogiltiga instrumentvariabler. Varje artikel utvärderar prestationsförmågan hos de nya metoderna genom att härleda teoretiska asymptotiska egenskaper, genom Monte Carlo-simuleringar samt genom empiriska tillämpningar.

Den första artikeln fokuserar på skattningsmetoder för linjära regressionsmodeller med instrumentvariabler som inkluderar många instrument. En metod baserat på ett viktat genomsnitt introduceras genom att kombinera minsta kvadratmetoden och s.k. limited information maximum likelihood-skattning.

Den andra artikeln adresserar problemet med endogenitet i probitmodeller med instrumentvariabler genom att utveckla en viktad Stein-estimator som inkluderar en kontrollfunktionsskattning genom maximum likelihood. Den asymptotiska fördelningen och asymptotiska risken för den föreslagna skattningsmetoden härleds.

Den tredje artikeln fokuserar på användningen av Lasso-metoder för att skatta instrumentvariabelmodeller. Dessutom introduceras en jackknife-baserad metod för att skatta instrumentvariabelmodeller med hjälp av Lasso. De föreslagna metoderna är robusta när man ska skatta kausala effekter genom modeller som innehåller många ogiltiga och giltiga instrumentvariabler.

Den fjärde artikeln introducerar val av instrumentvariabler genom modellvalssmetoden best subset via mixed integer-optimering för att skatta kausala effekter samt för att välja ogiltiga instrumentvariabler. Resultaten visar att den nya metoden är överlägsen flera tidigare utvecklade metoder.

Place, publisher, year, edition, pages
Jönköping: Jönköping University, Jönköping International Business School , 2024. , p. 29
Series
JIBS Dissertation Series, ISSN 1403-0470 ; 167
National Category
Probability Theory and Statistics
Identifiers
URN: urn:nbn:se:hj:diva-66497ISBN: 978-91-7914-048-9 (print)ISBN: 978-91-7914-049-6 (electronic)OAI: oai:DiVA.org:hj-66497DiVA, id: diva2:1909240
Public defence
2024-12-02, B1014, Jönköping International Business School, Jönköping, 13:15 (English)
Opponent
Supervisors
Available from: 2024-10-30 Created: 2024-10-30 Last updated: 2024-10-30Bibliographically approved
List of papers
1. A weighted average limited information maximum likelihood estimator
Open this publication in new window or tab >>A weighted average limited information maximum likelihood estimator
2024 (English)In: Statistical papers, ISSN 0932-5026, E-ISSN 1613-9798, Vol. 65, p. 2641-2666Article in journal (Refereed) Published
Abstract [en]

In this article, a Stein-type weighted limited information maximum likelihood (LIML) estimator is proposed. It is based on a weighted average of the ordinary least squares (OLS) and LIML estimators, with weights inversely proportional to the Hausman test statistic. The asymptotic distribution of the proposed estimator is derived by means of local-to-exogenous asymptotic theory. In addition, the asymptotic risk of the Stein-type LIML estimator is calculated, and it is shown that the risk is strictly smaller than the risk of the LIML under certain conditions. A Monte Carlo simulation and an empirical application of a green patent dataset from Nordic countries are used to demonstrate the superiority of the Stein-type LIML estimator to the OLS, two-stage least squares, LIML and combined estimators when the number of instruments is large.

Place, publisher, year, edition, pages
Springer, 2024
Keywords
2SLS, Endogeneity, Instrumental variables, LIML, Many weak instruments, Shrinkage estimator, Stein estimation
National Category
Probability Theory and Statistics
Identifiers
urn:nbn:se:hj:diva-62757 (URN)10.1007/s00362-023-01485-2 (DOI)001092166200001 ()2-s2.0-85173791738 (Scopus ID)HOA;;911837 (Local ID)HOA;;911837 (Archive number)HOA;;911837 (OAI)
Available from: 2023-10-23 Created: 2023-10-23 Last updated: 2024-10-30Bibliographically approved
2. Stein-type control function maximum likelihood estimator for the probit model in the presence of endogeneity
Open this publication in new window or tab >>Stein-type control function maximum likelihood estimator for the probit model in the presence of endogeneity
2024 (English)In: Econometrics and Statistics, ISSN 2452-3062Article in journal (Refereed) Epub ahead of print
Abstract [en]

A Stein-type control function maximum likelihood (CFML) estimator is suggested for the probit model in the presence of endogeneity. This novel estimator combines the probit maximum likelihood and CFML estimators. The asymptotic distribution and risk function for the new estimator is derived. It is demonstrated that, subject to certain conditions of the shrinkage parameter, the asymptotic risk of the new estimator is strictly smaller than the risk of the CFML. Monte Carlo simulations illustrate the method's superiority in finite samples. The method is also applied to analyze the impact of managerial incentives on the use of foreign-exchange derivatives.

Place, publisher, year, edition, pages
Elsevier, 2024
Keywords
Control function, Endogeneity, Instrumental variable, Model averaging, Probit Stein estimator
National Category
Probability Theory and Statistics
Identifiers
urn:nbn:se:hj:diva-63274 (URN)10.1016/j.ecosta.2023.12.001 (DOI)2-s2.0-85181133971 (Scopus ID)HOA;intsam;926131 (Local ID)HOA;intsam;926131 (Archive number)HOA;intsam;926131 (OAI)
Available from: 2024-01-10 Created: 2024-01-10 Last updated: 2024-10-30
3. LASSO-type instrumental variable selection methods with an application to Mendelian randomization
Open this publication in new window or tab >>LASSO-type instrumental variable selection methods with an application to Mendelian randomization
2024 (English)In: Statistical Methods in Medical Research, ISSN 0962-2802, E-ISSN 1477-0334Article in journal (Refereed) Epub ahead of print
Abstract [en]

Valid instrumental variables (IVs) must not directly impact the outcome variable and must also be uncorrelated with nonmeasured variables. However, in practice, IVs are likely to be invalid. The existing methods can lead to large bias relative to standard errors in situations with many weak and invalid instruments. In this paper, we derive a LASSO procedure for the k-class IV estimation methods in the linear IV model. In addition, we propose the jackknife IV method by using LASSO to address the problem of many weak invalid instruments in the case of heteroscedastic data. The proposed methods are robust for estimating causal effects in the presence of many invalid and valid instruments, with theoretical assurances of their execution. In addition, two-step numerical algorithms are developed for the estimation of causal effects. The performance of the proposed estimators is demonstrated via Monte Carlo simulations as well as an empirical application. We use Mendelian randomization as an application, wherein we estimate the causal effect of body mass index on the health-related quality of life index using single nucleotide polymorphisms as instruments for body mass index.

Place, publisher, year, edition, pages
Sage Publications, 2024
Keywords
Causal inference, instrumental variable, model selection, LASSO, jackknife, heteroscedasticity
National Category
Probability Theory and Statistics
Identifiers
urn:nbn:se:hj:diva-66492 (URN)10.1177/09622802241281035 (DOI)001355136200001 ()39544096 (PubMedID)2-s2.0-85209377064 (Scopus ID)HOA;intsam;66492 (Local ID)HOA;intsam;66492 (Archive number)HOA;intsam;66492 (OAI)
Available from: 2024-10-30 Created: 2024-10-30 Last updated: 2024-11-28
4. Best subset instrumental variables selection method using mixed integer optimization with applications to health-related quality of life and education-wage analyses
Open this publication in new window or tab >>Best subset instrumental variables selection method using mixed integer optimization with applications to health-related quality of life and education-wage analyses
(English)Manuscript (preprint) (Other academic)
National Category
Probability Theory and Statistics
Identifiers
urn:nbn:se:hj:diva-66495 (URN)
Note

Included in doctoral dissertation in manuscript form.

Available from: 2024-10-30 Created: 2024-10-30 Last updated: 2024-10-30

Open Access in DiVA

Kappa(2629 kB)57 downloads
File information
File name FULLTEXT01.pdfFile size 2629 kBChecksum SHA-512
61c63ce537b73aebe5d7cdbb74deba90f9fe1f6d372013f7b04086de253c2da0ee2f2cbf2a599a8a489011e10deddcd88972de001a5eeab97d223d9b4a3930d1
Type fulltextMimetype application/pdf

Authority records

Qasim, Muhammad

Search in DiVA

By author/editor
Qasim, Muhammad
By organisation
JIBS, Statistics
Probability Theory and Statistics

Search outside of DiVA

GoogleGoogle Scholar
Total: 57 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

isbn
urn-nbn

Altmetric score

isbn
urn-nbn
Total: 382 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf