Today, Mixed Effects Models (MEM) are extremely popular in many statistical domains for their success in prediction, data adaptiveness and experimental designs. As therefore they have been studied by a broadly composed scientific community, to the mentioned three adds the advantage of ‘being well studied and understood’. The first mentioned advantage (prediction power) is the main reason why they have become standard in Small Area Estimation (SEA). However, along with the increasing use in practice come new challenges like when using SAE predictors for comparative statistics or when combining standard approaches with model selection and highly nonlinear methods, often summarized as machine learning procedures. It is quite easy to show and see that the classical and therefore commonly used methods fail to work or get misinterpreted in what they actually offer to its user. As a simple example, none of the so far existing methods for constructing prediction intervals for SAE parameters was made for post-selection inference although model selection is typically applied in practice. They are neither made for doing comparative analysis between small areas, but SAE-based resource-allocation implies exactly this, what partly explains the existence of rank-estimates for small area parameters. Focusing on particularly poor or polluted small areas requires conditional inference, a topic that so far has largely been ignored. More complex data structures require highly nonlinear methods like random forests that also allow for the inclusion of qualitative data. Etc. This session presents recent advances to attack the statistical problems that arise from exactly those examples, namely uniform, conditional, post-selection and random forest inference for SAE predictors.

The already confirmed 3 speakers/authors will be

Prof. Tatyana Krivobokova, University of Vienna, Austria, https://sme.univie.ac.at/tatyana-krivobokova will talk about her research on conditional and post-LASSO inference in mixed models with examples in SAE, partly already publishes e.g. in JASA

Prof. Nikos Tzavidis, University of Southampton, UK, https://www.southampton.ac.uk/people/5wzwrv/professor-nikos-tzavidis will present his research on doing inference with Mixed Effects Random Forests

Prof. Stefan Sperlich, University of Geneva, Switzerland, https://sites.google.com/view/stefansperlich will talk about uniform inference in SAE and model selection for mixed parameters.

As discussant we invited (and he already confirmed): Dr. David Newhouse, Senior Economist at the World Bank, https://www.worldbank.org/en/about/people/d/david-newhouse will discuss these presentations from the practitioners’ point(s) of view.

The session is jointly proposed and organized by Nikos Tzavidis and Stefan Sperlich.

As outlined in the description, we focus on recent advances in complex inference for Mixed Effects Models (MEM) as these are extremely popular in many statistical domains. In this session we concentrate on new challenges for using MEM for small area estimation (SAE, where "small areas" may refer to any kind of clusters, not necessarily geographical ones), a statistics domain in which MEMs are particularly popular due to their prediction power.  
More specifically, we consider inference problems that arise when using MEM for small area predictors (a) to do comparative or joint statistics for various or all areas simultaneously, (b) when looking at estimates resulting from model selection, or (c) predictors resulting from highly nonlinear methods like random forests with mixed effects. It is easy to see that the existing classic, commonly used methods fail to work then, or at least don't offer what the practitioner expects them to do. Examples are that e.g. classical prediction intervals for SAE parameters were not made for any of the above problems: not for uniform or simultaneous infernce, not for post-selection inference, not for highly non-linear predictors. They are neither made for doing comparative analysis between small areas, but SAE-based resource-allocation implies exactly this. Morever, focusing on particularly poor or polluted small areas requires joint conditional inference, a topic that so far has largely been ignored. Likelwise, more complex data structures require highly nonlinear methods like random forests that would in addition allow for the inclusion of qualitative data.
Over the last four to five years, the invited speakers have been working exactly on these particular inference problems, i.e. uniform, post-selection, conditional, and highly non-linear inference for MEM. This session is intended to give an overview of the recent advances of such complex inference problems that practitioners are more and more frequently facing in small area estimation (in particular in official and environmental statistics, and especially for the new requirements of providing SDG-indicators on highly disaggregated levels).
Prof Tatyana Krivobokova will talk about her findings on conditional and post-LASSO inference in mixed models, Prof. Nikos Tzavidis will present his research on doing inference with mixed effects Random Forests, and Prof. Stefan Sperlich will talk about uniform and simultaneous inference for SAE. Dr. David Newhouse, Senior Economist at the World Bank, will discuss these advances from the practitioners’ point(s) of view, and Prof. Monica Pratesi, an internationally expert for small area estimation in practice and theory, acts as Session Chair.

Organiser: PROF. DR. stefan sperlich 

Speaker: PROF. DR. stefan sperlich 

Speaker: PROF. DR. Nikolaos Tzavidis 

Speaker: Prof. Tatyana Krivobokova 

Discussant:  David Newhouse