Skip to content

Research at St Andrews

Parameter redundancy and the existence of maximum likelihood estimates in log-linear models

Research output: Contribution to journalArticlepeer-review

DOI

Open Access permissions

Open

Abstract

Log-linear models are typically fitted to contingency table data to describe and identify the relationship between different categorical variables. However, the data may include observed zero cell entries. The presence of zero cell entries can have an adverse effect on the estimability of parameters, due to parameter redundancy. We describe a general approach for determining whether a given log-linear model is parameter redundant for a pattern of observed zeros inthe table, prior to fitting the model to the data. We derive the estimable parameters or functions of parameters and also explain how to reduce the unidentifiable model to an identifiable one. Parameter redundant models have a flat ridge in their likelihood function. We further explain when this ridge imposes some additional parameter constraints on the model, which can lead to obtaining unique maximum likelihood estimates for parameters that otherwise would not have been estimable. In contrast to other frameworks, the proposed novel approach informs on those constraints, elucidating the model that is actually being fitted.
Close

Details

Original languageEnglish
JournalStatistica Sinica
DOIs
Publication statusAccepted/In press - 10 Sep 2019

    Research areas

  • Contingency table, Extended maximum likelihood estimate, Parameter redundancy, Sampling zero

Discover related content
Find related publications, people, projects and more using interactive charts.

View graph of relations

Related by author

  1. Common methodological challenges encountered with multiple systems estimation studies

    Vincent, K., Sharifi Far, S. & Papathomas, M., 22 Dec 2020, In: Crime and Delinquency. OnlineFirst, 13 p.

    Research output: Contribution to journalArticlepeer-review

  2. Multiple system estimation of victims of human trafficking: model assessment and selection

    Cruyff, M., Overstall, A., Papathomas, M. & McCrea, R., 22 Dec 2020, In: Crime and Delinquency. Online First, 17 p.

    Research output: Contribution to journalArticlepeer-review

  3. On the correspondence of deviances and maximum-likelihood and interval estimates from log-linear to logistic regression modelling

    Jing, W. & Papathomas, M., 15 Jan 2020, In: Royal Society Open Science. 7, 1, 13 p., 191483.

    Research output: Contribution to journalArticlepeer-review

  4. On the correspondence from Bayesian log-linear modelling to logistic regression modelling with g-priors

    Papathomas, M., Mar 2018, In: TEST. 27, 1, p. 197-220 24 p.

    Research output: Contribution to journalArticlepeer-review

  5. Primary education in Vietnam and pupil online engagement

    Nguyen, Q., Naguib, R., Das, A., Papathomas, M., Vallar, E., Wickramasinghe, N., Santos, G. N., Galvez, M. C. & Nguyen, V., 8 Jan 2018, In: International Journal of Educational Management. 32, 1, p. 71-83

    Research output: Contribution to journalArticlepeer-review

Related by journal

  1. On optimality and construction of circular repeated-measurements designs

    Bailey, R. A., Cameron, P. J., Filipiak, K., Kunert, J. & Markiewicz, A., Jan 2017, In: Statistica Sinica. 27, 1, p. 1-22 22 p.

    Research output: Contribution to journalArticlepeer-review

  2. Construction of trend-resistant factorial designs

    Bailey, R. A., Cheng, C-S. & Kipnis, P., 1992, In: Statistica Sinica. 2, 2, p. 393-411 19 p.

    Research output: Contribution to journalArticlepeer-review

  3. Efficient semi-Latin squares

    Bailey, R. A., 1992, In: Statistica Sinica. 2, 2, p. 413-437 25 p.

    Research output: Contribution to journalArticlepeer-review

ID: 251876319

Top