A5089959851- Statistical Methods and Bayesian Inference
- Statistical Methods and Inference
- Spatial and Panel Data Analysis
- Social Policy and Reform Studies
- Advanced Statistical Methods and Models
- Statistical Methods in Clinical Trials
- Economic and Environmental Valuation
- Bayesian Methods and Mixture Models
- Efficiency Analysis Using DEA
- Monetary Policy and Economic Impact
- Survey Methodology and Nonresponse
- Optimal Experimental Design Methods
- Vehicle emissions and performance
- Health disparities and outcomes
- Advanced Causal Inference Techniques
- Energy, Environment, and Transportation Policies
- Psychometric Methodologies and Testing
- Income, Poverty, and Inequality
- Advanced Statistical Modeling Techniques
- Environmental Impact and Sustainability
- Fiscal Policy and Economic Growth
- Health and Medical Studies
- Electoral Systems and Political Participation
- Labor market dynamics and wage inequality
- Gender, Labor, and Family Dynamics
Universität Hamburg
2011-2024
German Institute for Economic Research
1999-2009
Europa-Universität Flensburg
2006-2007
University of Regensburg
1996-1998
Factor loadings and item discrimination parameters play a key role in scale construction. A multitude of heuristics regarding their interpretation are hardwired into practice-for example, neglecting low assigning items to exactly one scale. We challenge the common sense these by providing counterexamples general results which altogether cast doubt on our understanding parameters. In particular, we highlight counterintuitive way best prediction test taker's latent ability depends factor...
Die Daten des Sozio-oekeonomischen Panels (SOEP) werden seit 1984 auf Grundlage eines mit dem Befragungsinstitut Infratest abgestimmten Erhebungskonzepts jahrlich erhoben. Erfasst die zentral in Munchen bei TNS Infratest, aufbereitet und beim DIW Berlin, um dann moglichst zeitnah durch SOEP-Gruppe am Berlin an externe Nutzer weitergegeben zu werden. Das komplexe Zusammenspiel von Erhebung, Aufbereitung, Speicherung Weitergabe der SOEP-Daten wird laufend aktualisiert immer wieder jeweiligen...
The sensitivity of multiple imputation methods to deviations from their distributional assumptions is investigated using simulations, where the parameters scientific interest are coefficients a linear regression model, and values in predictor variables missing at random. performance newly proposed method based on generalized additive models for location, scale, shape (GAMLSS) investigated. Although predictive mean matching virtually unbiased, they suffer mild moderate under-coverage, even...
By modeling the effects of predictor variables as a multiplicative function regression parameters being invariant over categories, and category-specific scalar effects, ordered stereotype logit model is flexible for ordinal response variables. In this article, we propose generalized estimating equations (GEE) approach to estimate panel data based on working covariance matrices, which are not required be correctly specified. A simulation study compares performance GEE estimators various...
Summary The paper proposes an estimation approach for panel models with mixed continuous and ordered categorical outcomes based on generalized estimating equations the mean pseudoscore covariance parameters. A numerical study suggests that efficiency can be gained in parameter estimators by using individual matrices is applied to estimate returns occupational qualification terms of income perceived job security a 9-year period German Socio-Economic Panel. To compensate missing data, combined...
Handedness is one of the main issues in laterality research and known to be related a large number morphological asymmetries central nervous system. However, focus previous studies were cerebral structures, which ignored spinal cord as most distal neural entity innervating muscles extremities. We analyzed morphometrically motoneurons from segments arms hands compared them with that innervated upper trunk. found an asymmetry larger motoneuron perikaryas on right side limbs. To our knowledge...
Empirical research usually takes place in a space of available external information, like results from single studies, meta-analyses, official statistics or subjective (expert) knowledge. The information ranges simple means and proportions to known relations between multitude variables estimated distributions. In psychological research, derived the named sources may be used build theory derive hypotheses. addition, techniques do exist that use estimation process, for example prior...
In this paper we discuss the rationale for tracing non-original sample members (Non-OSMs) in household panel studies, and particular SOEP, implications weighting. We present results on incidence, survival rates, thus relevance of Non-OSMs SOEP.
Im Januar des Jahres 2008 begann die Feldarbeit der 25. Welle SOEP. Jahr 2009 werden damit etwa 2500 anonymisierte Datensatze von Personen zur Auswertung Verfugung stehen, seit 1984 ununterbrochen an 25 Befragungen in den westdeutschen SOEP-Teilstichproben A und B teilgenommen haben. wird auserdem bereits 20. Teilstichprobe C Ostdeutschland erhoben, mit nicht nur Lebenslaufe Erwachsenen im Transformationsprozess umfassend dokumentiert sind, sondern auch Entwicklung Kindern Wiedervereinigung...
Data analysis, common to all empirical sciences, often requires complete data sets.Unfortunately, real world collection will usually result in values not being observed.We present a package for robust multiple imputation (the ImputeRobust package) that allows the use of generalized additive models location, scale, and shape context chained equations.The paper describes basics technique which builds on semi-parametric regression model (GAMLSS) algorithms functions provided with corresponding...
Dieser Aufsatz ist der Frage Weiterentwicklung des SOEP insgesamt gewidmet. Daher nehmen wir in einem ersten Teil eine Bestandsaufnahme Entscheidungsstrukturen zum Erhebungsprogramm (Governance) vor. Im zweiten Abschnitt beschaftigen uns mit den Inhalten SOEP-Erhebung - also dem SOEPsurvey. Danach gehen auf die Daten-Services fur wachsende Gruppe externen Forscherinnen und Forscher im In- Ausland (SOEPservice) ein. Und schlieslich diskutieren einige forschungspolitische Probleme, bei...
Abstract Since Liang and Zeger (1986) proposed the ‘generalized estimating equations’ approach for estimation of regression parameters in models with correlated discrete responses, a lot work has been devoted to investigation properties corresponding GEE estimators. However, effects different kinds covariates have often overlooked. In this paper it is shown that use non‐singular block invariant matrices covariates, as e.g. design matrix an analysis variance model, leads estimators which are...
Household panel studies usually provide data on parents, children, and siblings, but not, e.g., all ex-partners after a divorce or split in cohabiting unit. This kind of information, however, can be an interesting source for intergenerational research or, more generally, the dynamics life course. To best our knowledge, among diverse household studies, only German Socio-Economic Panel (SOEP) contains individuals entering sample wave two also leaving that again. allows members to traced once...