{"id":47851,"date":"2026-10-22T08:59:37","date_gmt":"2026-10-22T08:59:37","guid":{"rendered":"https:\/\/dpc-software.de\/?page_id=47851"},"modified":"2026-11-02T09:14:49","modified_gmt":"2026-11-02T09:14:49","slug":"german-stata-conference-2024","status":"publish","type":"page","link":"https:\/\/dpc-software.de\/en\/german-stata-conference-2024\/","title":{"rendered":"German Stata Conference 2024"},"content":{"rendered":"\t\t
\n\t\t\t\t
\n\t\t
\n\t\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t

German Stata Conference 2024<\/h1>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t
\n\t\t\t
<\/div>\n\t\t<\/div>\n\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t
\n\t\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t

Report German Stata Conference 2024<\/span><\/p><\/h2>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t

\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t

This year\u2019s German Stata Conference was held on June 7th\/8th at GESIS Leibniz Institute for the Social Sciences in Mannheim.\u00a0<\/p>

Prof. Reinhard Pollack<\/strong>, head of department data and research on society hosted this conference<\/strong>, while the DPC team<\/strong> was responsible for the organization<\/strong>. GESIS is one of the largest research institutions in Europe providing data infrastructure in the Social Sciences and therefore a perfect partner for Stata. Prof. Johannes Giesecke<\/strong> (Humboldt University Berlin) and Prof. Ulrich Kohler<\/strong> (University of Potsdam) served as the scientific committee<\/strong>. As representatives of Stata Corp. we had the pleasure to have Kristin MacDonald<\/strong> and Chinh Nguyen<\/strong> as guests. The DPC team and the conference participants have used the chance to discuss a couple of issues around Stata with its developers in person.\u00a0<\/p>

On Thursday, we had a very good pre-conference workshop on difference-in-difference (DiD) estimation<\/strong>, in which Felix Knauf<\/strong> from Science Po in Paris presented newest and most recent developments of this methods. As an expert in the field he developed together with colleagues a couple of user-written ados, which he explained in detail in his 7-hour workshop. Participants had the chance to conduct hands-on sessions and discuss specific issues with Felix in detail. He covered the theoretical basis of the method, different DiD-approaches and -designs as well as recent papers that further develops the method. Felix also provided a couple of web-based resources to DiD estimation.\u00a0<\/p>

The conference started on Friday with a presentation of GESIS by Prof. Pollack<\/strong>, followed by the keynote of Prof. Paul Lambert<\/strong> (University of Leicester and Karolinska Institutet Stockholm) on survival analysis<\/strong> (aka event history models). He argued for flexible parametric survival models as an alternative to the \u201agold standard\u2018 of the Cox proportional hazards model. In that context, he presented his user-written ado \u2019stpm3\u2032 and a couple of postestimation tools. Also related to survival analysis was the second presentation by Prof. Harald Tauchmann<\/strong> (Friedrich-Alexander-University of Erlangen-Nurenberg), where he presented his user-written ados \u201acfbinout\u2018<\/strong> and \u201axtdhazard\u2018<\/strong>. Both presentations represent most recent discussions in the field of survival analysis.\u00a0<\/p>

The last two presentations before the lunch break were given by Peter Krause<\/strong> (DIW Berlin) on multidimensional well-being, deprivation, and inequality<\/strong> and by Wolfgang Langer<\/strong> (Martin-Luther-University Halle-Wittenberg) on model fit statistics of logistic regression models<\/strong>. He developed an own program that is able to calculate different pseudo-r-squared statistics in order to asses the goodness of fit of these non-linear models. This presentation was followed by a fruitful and controversial discussion on the validity of the different measures, because a community-wide accepted standard has not agreed upon yet. The user-written ado \u201afit_cmlogit\u2018 contributes to the ongoing debate on how to assess the goodness of fit for logistic regression models.\u00a0<\/p>

After the lunch break Kristin Macdonald<\/strong> from Stata Corp. presented Stata\u2019s recent commands for customizing tables, namely the renewed table-command, etable<\/strong> and dtable<\/strong>. Those are wrappers to the new and very elaborated collect suite<\/strong> that allows to pile up a collection of statistical tables, which can be modified according to individual needs and exported to MS Word, Excel, HTML, LaTeX or pdf format. The export of tables from Stata into other formats was an obstacle for many years and has led to a couple of user-written ados, such as estab or estout. The implementation of the rather complex collect suite and the more accessible wrappers table, dtable and etable has filled a gap in Stata\u2019s command universe and will clearly facilitate the construction and export of tables from Stata to Word processing software.\u00a0<\/p>

After that, Prof. Ben Jann<\/strong> (University of Bern, Switzerland) presented his very large ado \u201ageoplot\u2018, which is able to draw maps in Stata based on standardized shape-files<\/strong> and other formats. Shape-files contain vector-data information and are widely available for any geographic entity. In combination with datafiles, the geoplot-command is able to generate map graphs with any available spatial data. Prof. Jann\u2019s presentation addressed an audience beyond statistical experts and offers a lot of application possibilities across different disciplines.\u00a0<\/p>

A topic that has its origin in the replication crisis within Psychology is the question how to ensure reproducibility.<\/strong> This question has gained attention in all scientific disciplines, because increasing complexity of both, datasets and statistical methods, has led to a situation, which makes the reproduction of scientific results increasingly difficult. This issue challenges the scientific system as a whole, because of its central epistemological importance. Daniel Kr\u00e4hmer<\/strong> (Ludwig-Maximilian-University of Munich) presented his command \u201arepreport\u2018, which \u2013 similar to a classical log-file \u2013 automatically compiles information during a running analysis in order to collect only the information necessary for reproduction of a study. This presentation started a pathway that MUST be continued and developed further, because it is of major importance for the integrity of any discipline\u2019s research.\u00a0<\/p>

A similar direction is taken by the presentation of Maarten Buis<\/strong>, who developed a set of commands (\u201amkproject\u2018 and \u201aboilerplate\u2018) which aims at standardizing the process of analysis<\/strong>. This also relates to the reproduction problem, because increasing individualization of research processes do challenge the reproductivity of research, too.\u00a0<\/p>

The last presentation of the conference also touches a very general topic that is important for common researchers, namely the data structures in general, and the transfer between them. Daniel Schneider <\/strong>(Max-Planck-Institute for Demographic Research, Rostock) aimed at starting a discussion on how on could use Stata\u2019s potential in this respect.\u00a0<\/p>

To sum up the topics of this year\u2019s Stata Conference, it becomes apparent that there were many general than specific issues. Especially the reproduction issue is important for ANY research discipline and should be continued.\u00a0<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t

\n\t\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t
Keynote Speaker<\/div>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t
\n\t\t\t\t\t
\n\t\t
\n\t\t
\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t
\"\"<\/div><\/div><\/div>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t
\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t
Professor Paul C. Lambert\n\n<\/h6>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t
Professor of Biostatistics \n\n<\/h6>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t

Paul Lambert is Professor of Biostatistics at the University of Leicester, UK and Karolinska Institute, Sweden. His main research interests is survival analysis methods in epidemiology, specializing in methods in population-based cancer studies. He has developed various Stata commands, mainly in the area of survival analysis. (https:\/\/pclambert.net\/#software<\/a>)\u00a0<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t

\n\t\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t
Workshop: DID Estimation Using Stata<\/div>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t
\n\t\t
\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t
Presenter\n<\/h6>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t
Felix Knauf\n\n<\/h6>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t
SciencesPo, Paris \u2013 Department of Economics\n\n<\/h6>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t

Felix Knauf studied both a bachelor\u2019s and a master\u2019s degree in economics at the University of Mannheim and successfully completed them last summer. Meanwhile, he also worked there as a research assistant and at the IAB in Nuremberg as an intern and later research assistant. Since September 2023 he has been working as a full-time research assistant for Professor Clement de Chaisemartin at Sciences Po. There he mainly works on the implementation of various (DID) estimation methods for causal treatment effects as Stata packages as well as the \u201euser support\u201c for the corresponding commands.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t

\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t
Topics<\/h6>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t