We are pleased to welcome Johanna Einhorn to the Chair of Statistics and Econometrics.

Frink, N.; Schmid, T. 

Abstract: Small area estimation methods are proposed that use generalized tree-based machine learning techniques to improve the estimation of disaggregated means in small areas using discrete survey data. Specifically, two existing approaches based on random forests - the Generalized Mixed Effects Random Forest (GMERF) and a Mixed Effects Random Forest (MERF) - are extended to accommodate count outcomes, addressing key challenges such as overdispersion. Additionally, three bootstrap methodologies designed to assess the reliability of point estimators for area-level means are evaluated. The numerical analysis shows that the MERF, which does not assume a Poisson distribution to model the mean behavior of count data, excels in scenarios of severe overdispersion. Conversely, the GMERF performs best under conditions where Poisson distribution assumptions are moderately met. In a case study using real-world data from the state of Guerrero, Mexico, the proposed methods effectively estimate area-level means while capturing the uncertainty inherent in overdispersed count data. These findings highlight their practical applicability for small area estimation.

Nicolas Frink & Timo Schmid (2025) Small area prediction of counts under machine learning-type mixed models, Computational Statistics & Data Analysis, DOI: https://doi.org/10.1016/j.csda.2025.108218

Krennmair, P.; Würz, N.; Schmid, T. 

Abstract: Evidence-based policy-making requires reliable, spatially disaggregated indicators. The framework of mixed effects random forests leverages the advantages of random forests and hierarchical data in small area estimation. These methods require typically access to auxiliary information on population level, which is a strong limitation for practitioners. In contrast, our proposed method—for point and uncertainty estimation—abstains from access to unit-level population data but adaptively incorporates aggregated auxiliary information through calibration weights. We demonstrate its usage for estimating opportunity cost of care work for Germany from the Socio-Economic Panel and census aggregates. Simulation studies evaluate our proposed method.

Patrick Krennmair, Nora Würz & Timo Schmid (2025) Analysing opportunity cost of care work using mixed effects random forests under aggregated auxiliary data, Journal of the Royal Statistical Society Series C, DOI: https://doi.org/10.1093/jrsssc/qlaf031

https://academic.oup.com/jrsssc

Edochie, I.; Newhouse, D.; Tzavidis, N.; Schmid, T.; Foster, E.; Hernandez, A. L.; Ouedraogo, A.; Sanoh, A.; Savadogo, A.

Abstract: The paper presents methodology to generate experimental small area estimates (SAE) of poverty in four West African countries: Chad, Guinea, Mali, and Niger. Due to the absence of recent census data in the four countries, household level survey data are integrated with grid-level geospatial data, which are used as covariates in model-based estimation. Leveraging geospatial data enables reporting of poverty estimates more frequently at disaggregated administrative levels and makes estimation feasible in areas for which survey data are not available. The paper leverages the availability of a recent census in Burkina Faso for evaluation purposes. Estimates obtained with the same survey instruments and candidate geospatial covariates as the other four countries are compared against estimates obtained using recent census data and an empirical best predictor under a unit level model. For Burkina Faso, estimates obtained using geospatial data are highly correlated with the census-based ones in sampled areas but moderately correlated in non-sampled areas. The results demonstrate that in the absence of recent census data, small area estimation with publicly available geospatial covariates is feasible, can lead to large efficiency improvements compared to direct estimation, and improve the timeliness of small area estimates. 

Ifeanyi Edochie, David Newhouse, Nikos Tzavidis, Timo Schmid, Elizabeth Foster, Angela Luna Hernandez,  Aissatou Ouedraogo, Aly Sanoh & Aboudrahyme Savadogo (2024) Small Area Estimation of Poverty in Four West African Countries by Integrating Survey and Geospatial Data, Journal of Official Statistics, forthcoming, DOI: https://doi.org/10.1177/0282423X241284890

Runge, M.; Schmid, T.

Abstract: In this article, we propose a framework for small area estimation with multiply imputed survey data. Many statistical surveys suffer from (a) high nonresponse rates due to sensitive questions and response burden and (b) too small sample sizes to allow for reliable estimates on (unplanned) disaggregated levels due to budget constraints. One way to deal with missing values is to replace them by several plausible/imputed values based on a model. Small area estimation, such as the model by Fay and Herriot, is applied to estimate regionally disaggregated indicators when direct estimates are imprecise. The framework presented tackles simultaneously multiply imputed values and imprecise direct estimates. In particular, we extend the general class of transformed Fay-Herriot models to account for the additional uncertainty from multiple imputation. We derive three special cases of the Fay-Herriot model with particular transformations and provide point and mean squared error estimators. Depending on the case, the mean squared error is estimated by analytic solutions or resampling methods. Comprehensive simulations in a controlled environment show that the proposed methodology leads to reliable and precise results in terms of bias and mean squared error. The methodology is illustrated by a real data example using European wealth data. 

Marina Runge & Timo Schmid (2023) Small Area with Multiply Imputed Survey Data, Journal of Official Statistics, 39, pp. 507-533, DOI: https://doi.org/10.2478/jos-2023-0024

Harmening, S.; Kreutzmann, A.-K..; Schmidt, S.; Salvati, N.; Schmid, T.

Abstract: The R package emdi facilitates the estimation of regionally disaggregated indicators using small area estimation methods and provides tools for model building, diagnostics, presenting, and exporting the results. The package version 1.1.7 includes unit-level small area models that rely on access to micro data. The area-level model by Fay and Herriot (1979) and various extensions have been added to the package since the release of version 2.0.0. These extensions include (a) area-level models with back-transformations, (b) spatial and robust extensions, (c) adjusted variance estimation methods, and (d) area-level models that account for measurement errors. Corresponding mean squared error estimators are implemented for assessing the uncertainty. User-friendly tools like a stepwise variable selection, model diagnostics, benchmarking options, high quality maps and results exportation options enable a complete analysis procedure. The functionality of the package is illustrated by examples based on synthetic data for Austrian districts.

Sylvia Harmening, Ann-Kristin Kreutzmann, Sören Schmidt, Nicola Salvati & Timo Schmid (2023) A Framework for Producing Small Area Estimates Based on Area-Level Models in R, The R Journal, DOI: https://doi.org/10.32614/RJ-2023-039

Rendtel, U. ; Lee, Y. ; Gerks H.

Abstract: In a survey among newly enrolled graduate students, we asked about social background, students’ financial situation, and their motivation for choosing their major. By connecting administrative data with our survey data, we can analyze students’ academic performance and overall success in relation to individual characteristics. Through this approach, we avoid common pitfalls of student surveys, such as high non-response rates and recall errors. We use this approach for a comparison of five master programs at the economic department of the Freie Universität Berlin. After the initial survey, the students were followed over the first 6 semesters with respect to the earning of credit points, the finalizing of the master program and the students’ final grades.

In the linked data, we noticed a response bias towards more successful students in the entire cohort. However, this bias can be controlled by using a suitable weight according to the response rate. We compare the master programs in different phases: the starting phase, the reaching of the planned number of semesters to complete the program and the finalizing of the program until the 6th semester. Furthermore we study the impact of the background variables on the final grades. The starting phase reveals clear differences between the master programs. However, if we condition on the study success in the starting period differences vanish in the later phases and for the gained final grade. The impact of the bachelor score on the study performance and the gained final score is surprisingly low. Our results indicate the possibility to predict relatively sure a breaking off of the study on the basis of the gained credit points in the starting phase of the program.

Ulrich Rendtel, Yeonjoo Lee & Harmut Gerks (2023) Eine Analyse des Studienerfolgs im Masterstudium auf der Basis von Umfrage- und administrativen Prüfungsdaten: Ein Vergleich von fünf Masterstudiengängen am Fachbereich Wirtschaftswissenschaft der Freien Universität Berlin, AStA Wirtschafts- und Sozialstatistisches Archiv, DOI: https://doi.org/10.1007/s11943-023-00325-x

Lee, Y.; Rojas-Perilla, N.; Runge, M.; Schmid, T.

Abstract: When data analysts use linear mixed models, they usually encounter two practical problems: (a) the true model is unknown and (b) the Gaussian assumptions of the errors do not hold. While these problems commonly appear together, researchers tend to treat them individually by (a) finding an optimal model based on the conditional Akaike information criterion (cAIC) and (b) applying transformations on the dependent variable. However, the optimal model depends on the transformation and vice versa. In this paper, we aim to solve both problems simultaneously. In particular, we propose an adjusted cAICby using the Jacobian of the particular transformation such that various model candidates with differently transformed data can be compared. From a computational perspective, we propose a step-wise selection approach based on the introduced adjusted cAIC. Model-based simulations are used to compare the proposed selection approach to alternative approaches. Finally, the introduced approach is applied to Mexican data to estimate poverty and inequality indicators for 81 municipalities.

Yeonjoo Lee, Natalia Rojas-Perilla, Marina Runge & Timo Schmid(2023) Variable selection using conditional AIC for linear mixed models with data-driven transformations, Statistics and Computing, DOI: https://doi.org/10.1007/s11222-022-10198-9