Methoden

methoden@mailman.uni-konstanz.de

3 participants
1901 discussions

Wissenschaftliche*n Mitarbeiter*in gesucht – Quantitative Methoden – Uni Bielefeld
by Kühne, Simon 06 Feb '20

06 Feb '20

Sehr geehrte Damen und Herren, Liebe Kolleginnen und Kollegen, zum nächstmöglichen Zeitpunkt suchen wir an der Fakultät für Soziologie, Universität Bielefeld eine/n Wissenschaftliche*n Mitarbeiter*in (m/w/d, TV-L 13, 100%, 3 Jahre) im Bereich Quantitative Methoden. Die Stelle soll bei der Entwicklung eines Digitalisierungskonzepts für die Methodenausbildung mitarbeiten, Lehre im Umfang von 4 SWS in den Quantitativen Methoden und Sozialstrukturanalyse anbieten und sich in gemeinsamen Forschungsschwerpunkten der Quantitativen Methoden an der Fakultät für Soziologie engagieren. Mehr Informationen zur Stelle finden Sie in der PDF im Anhang bzw. unter https://www.uni-bielefeld.de/Universitaet/Aktuelles/Stellenausschreibungen/… Bewerbungsfrist ist der 04.03.2020. Wir danken für die Weiterleitung der Ausschreibung an geeignete Personen. Mit freundlichen Grüßen Simon Kühne -- Dr. Simon Kühne Akademischer Rat - Universität Bielefeld AG Methoden der empirischen Sozialforschung Prof. Dr. Martin Kroh Raum B3-201, Gebäude X, Universitätsstraße 25 Tel.: +49 521 106 4681 E-Mail: simon.kuehne(a)uni-bielefeld.de<mailto:simon.kuehne@uni-bielefeld.de>

1 0

3rd annual Women in Data Science Mannheim conference on March 4, 2020
by Florian Keusch 06 Feb '20

06 Feb '20

We’re excited to launch the 3rd annual Women in Data Science Mannheim conference on March 4, 2020 in cooperation with SAP Next-Gen and Stanford University! We will unite local Data Science industry and academic leaders, as well as students and post-doc researchers for an exceptional one-day technical conference. The Women in Data Science conference aims to inspire and educate data scientists worldwide, regardless of gender, and support women in the field. This annual conference is held at Stanford University and 150+ locations worldwide. We’re proud to announce this amazing speaker line-up: * Julia Krönung, Professor of Information Systems, EBS Business School * Katrin Lehmann, Senior Vice President S/4HANA Cloud Data Management, SAP * Iris Heckmann, Principal, Camelot ITLab * Cosima Meyer, PhD candidate, Graduate School of Economics and Social Sciences & research associate, Chair of Political Science, International Relations, University of Mannheim * Simona Marincei, Head of Intelligent Processes – SME Product Management, SAP * Tatjana Schröder, AI Manager, trans-o-flex & Master Student, International Program in Survey and Data Science, University of Mannheim When? March 4, 2020, 3-7.30 pm Where? SAP Experience Room Walldorf, Building 49, Dietmar-Hopp-Allee 17, 69190 Walldorf Join us, share your experience and learn about cutting edge data science-related innovation and impact! Register now: https://events.sap.com/de/wids-2020-walldorf/en/home -- Florian Keusch Professor of Statistics and Methodology (interim) University of Mannheim A5, 6 Room B218, 2nd floor 68159 Mannheim, Germany +49 (0)621 - 181 3214 f.keusch(a)uni-mannheim.de floriankeusch.weebly.com

1 0

New issue of "methods, data, analyses (mda)"
by Jan Karem Höhne 05 Feb '20

05 Feb '20

***Apologies for cross-posting Dear colleagues, I am happy to announce the new issue of the open-access journal "methods, data, analyses (mda)", which is edited by Melanie Revilla: https://mda.gesis.org/index.php/mda mda publishes research on all questions important to quantitative methods, with a special emphasis on survey methodology. mda is released online as open-access journal. All content is freely available and can be distributed without any restrictions, ensuring the free flow of information that is crucial for scientific progress. Of course, mda does not raise any publication fees and welcomes submissions from all over the world. mda is indexed and abstracted in the Emerging Sources Citation Index. mda is also listed in the Directory of Open Access Journals and SCOPUS. Best regards Jan Karem Höhne (associate editor) -- Jan Karem Höhne Postdoctoral Researcher University of Mannheim B6, 30-32 (Room: 346) 68131 Mannheim, Germany +49 (0)621/181-3483 hoehne(a)uni-mannheim.de

1 0

Frühjahrstagung Methodensektion der DGS
by Tobias Wolbring 30 Jan '20

30 Jan '20

━━━━━━━━━━━━━━━━━━━━━━━━━ GRAPHICAL CAUSAL MODELS ━━━━━━━━━━━━━━━━━━━━━━━━━ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ Datum 6. März 2020, 14:00–18:00 7. März 2020, 9:00–12:30 Ort Universität Potsdam Campus Griebnitzsee August-Bebel-Straße 89 14482 Potsdam Raum Gebäude 6 Raum S21 Workshop 6. März 2020, 12:00–14:00 Universität Potsdam Campus Griebnitzsee Gebäude 6 Raum S21 Anmeldung bis 24.2.2020 auf [https://terminplaner4.dfn.de/C6UtOk2q6F1Lcbg9] ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ Die Sektion [Methoden der empirischen Sozialforschung] der [Deutschen Gesellschaft für Soziolgie] läd alle Kolleginnen und Kollegen aus der Soziolgie und den angrenzenden Fächern zu ihrer Frühjahrstagung zum Thema *Graphical Causal Models* ein. Aus Anlass der Frühjahrstagung veranstaltet das [Potsdam Center for Quantiative Research] gemeinsam mit der Methodensektion den einführenden Workshop *Causal Graphs Basics* von [Julian Schüssler] (Universität Konstanz). Eine genauere Beschreibung des Workshops finden Sie am Ende der Kurzzusammenfassungen. [Methoden der empirischen Sozialforschung] https://dgs-methoden.uni-konstanz.de/ [Deutschen Gesellschaft für Soziolgie] https://soziologie.de/aktuell [Potsdam Center for Quantiative Research] https://www.uni-potsdam.de/pcqr [Julian Schüssler] https://www.polver.uni-konstanz.de/cdm/people/students/schuessler/ Freitag 6. März 2020 ════════════════════ Workshop ──────── ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 12:00–14:00 Causal Graphs Basics Workshop von Julian Schüssler ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ Tagung ────── ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 13:30 –14:00 Registrierung (/Kaffee/) ──────────────────────────────────────────────────────────────────────────────────────── 14:00–14:10 Willkommen 14:10–14:50 Kausale Mediationsanalyse mit Directed Acyclic Graphs* Michael Kühhirt 14:50–15:30 Does positive affect mediate the effect of multimorbidity on depression? Ibrahim Demirer ──────────────────────────────────────────────────────────────────────────────────────── 15:30–16:00 /Kaffee, Kuchen/ ──────────────────────────────────────────────────────────────────────────────────────── 16:00–16:40 Defining and Identifying Discrimination Using DAGs* Sebastian Wenz 16:40–17:20 Bringing Research Design Back In* Fabian Class/Ulrich Kohler/Tim Sawert 17:20-18:00 The Art of Programming Questionnaires* Claudia Saalbach ──────────────────────────────────────────────────────────────────────────────────────── 18:00–20:00 Mitgliederversammlung ──────────────────────────────────────────────────────────────────────────────────────── 20:00 Gemeinsames Abendessen ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ Samstag 7. März 2020 ════════════════════ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 9:00–10:00 External Validity and Transportability* Julian Schüssler 10:00–10:40 Multilevel Analysis with Few Clusters* Jan Paul Heisig ─────────────────────────────────────────────────────────────────────────────────────────────────────────── 10:40–11:10 /Kaffee, Gebäck/ ─────────────────────────────────────────────────────────────────────────────────────────────────────────── 11:10–11:50 Graph Theory and Macroeconomic Regimes Miguel Carrión Álvarez/Dirk Ehnts 11:50–12:30 Analyse ökonomischer Selektivität im politischen Engagement unter Anwendung von RMC und DAGs Juliana Witkowski/Simon Ress ─────────────────────────────────────────────────────────────────────────────────────────────────────────── 12:30 Ende der Tagung ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ Anmeldung ═════════ Um Anmeldung unter [https://terminplaner4.dfn.de/C6UtOk2q6F1Lcbg9] wird gebeten. Anreise ═══════ Der Campus Griebnitzsee der Universität Potsdam liegt unmittelbar am S- und Regionalbahnhof Griebnitzsee. Der Bahnhof ist im 10-Mintuten-Takt mit der S-Bahn-Linie 7 von Potsdam-Hauptbahnhof bzw. Berlin-City (Charlottenburg, Savigny-Platz, Zoologischer Garten, Hauptbahnhof, Friedrichstraße, etc.) sowie mit der Regionalbahn RB 21 zu erreichen. Der Bahnhof Griebnitzsee gehört zum Berliner Tarifbereich C. Bei Anfahrt aus Berlin ist ein Ticket für /ABC/ zu lösen. DB-City-Fahrscheine für Berlin haben keine Gültigkeit. • [Fahrinfo der BVG] • [OSM-Karte] • [Lageplan Campus Griebnitzsee] [Fahrinfo der BVG] https://www.bvg.de/de/Willkommen [OSM-Karte] https://www.openstreetmap.de/karte.html?zoom=17&lat=52.39332&lon=13.12927&l… [Lageplan Campus Griebnitzsee] https://www.uni-potsdam.de/db/zeik-portal/gm/lageplan-up.php?komplex=3 Unterkünfte ═══════════ Die Teilnehmer werden gebeten, sich selbst um Unterkünfte zu bemühen. Neben dem unmittelbar in der Nachbarschaft des Universitätscampus gelegenen Seminaris Avendi Hotels (s.u.) bieten sich Hotels in der Umgebung von S-Bahnhöfen entlang der S7 an. Nachfolgend einige Vorschläge (Fahrtzeit in Klammern) • Fußläufig • [Seminaris Avendi Hotel Potsdam] • Bereich Potsdam Hbf. (6 min) • [Mercure Hotel Potsdam City] • S-Bahn Charlottenburg (20 min) • [ART-Hotel Charlottenburger Hof ] • [HappyGoLucky Hotel] • [City Pension Berlin] • S-Bahn Zoologischer Garten (24 min) • [Hotel Motel One] • [Aletto Hotel Kudamm] [Seminaris Avendi Hotel Potsdam] https://www.seminaris.de/hotels/potsdam/seminaris-avendi-hotel-potsdam [Mercure Hotel Potsdam City] https://www.mercure-potsdam.com/ [ART-Hotel Charlottenburger Hof ] https://charlottenburger-hof.de/ [HappyGoLucky Hotel] http://www.happygoluckyhotel.com/ [City Pension Berlin] https://www.city-pension.de/ [Hotel Motel One] https://www.motel-one.com/de/hotels/berlin/hotel-berlin-upper-west/ [Aletto Hotel Kudamm] https://www.aletto.de/de/ Kurzzusammenfassungen ═════════════════════ Kausale Mediationsanalyse mit Directed Acyclic Graphs ───────────────────────────────────────────────────── Referent ╌╌╌╌╌╌╌╌ • Michael Kühhirt, Universität zu Köln, Institut für Soziologie und Sozialpsychologie Zusammenfassung ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌ Die Untersuchung von Kausalzusammenhängen geht häufig einher mit der Frage nach den diesen Zusammenhängen zugrundeliegenden Prozessen und Mechanismen. Ein zentraler Aspekt dieser Frage ist, ob und zu welchem Maße der Effekt einer Variablen über eine oder mehrere weitere Variablen (i.e., potentielle Mediatoren) vermittelt wird. In der Praxis wird diese Frage häufig anhand von Veränderungen in Regressionskoeffizienten nach zusätzlicher Kontrolle der potentiellen Mechanismen untersucht. Die neuere Literatur zu kausaler Inferenz zeigt jedoch, dass dieser Ansatz mit Annahmen behaftet ist, die in der angewandten Forschung nur selten thematisiert und damit auch häufig keiner Plausibilitätsprüfung unterzogen werden. Am Beispiel des Effekts der sozialen Herkunft auf den Arbeitsmarkterfolg nutzt dieser Beitrag Directed Acyclic Graphs und die kontrafaktische Definition direkter und indirekter Kausaleffekte, um die Bedingungen transparent zu machen, unter denen Mediationsanalyse valide Schlussfolgerungen über direkte und indirekte Effekte zulässt. Im Anschluss daran werden verschiedene neuere Methoden der Schätzung direkter und indirekter Effekte mit den traditionellen Verfahren der Differenz- und Produktmethode verglichen. Does positive affect mediate the effect of multimorbidity on depression? ──────────────────────────────────────────────────────────────────────── Referent ╌╌╌╌╌╌╌╌ • Ibrahim Demirer, Institut für Medizinsoziologie, Versorgungsforschung und Rehabilitationswissenschaft Zusammenfassung ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌ Although the association between multimorbidity (MM) and depression (DP) is well established, the pathways through which MM increases the risk of DP are often left obscure. Consequently, the identified associations suffer from unobserved heterogeneity. Moreover, the association is described as bidirectional and heavily influenced by intermediate factors. According to Beck (2008), DP is rooted in the individuals' negative interpretations of experiences. A major determinant of this interpretation is positive affect (PA). Exposition to MM tends to influence PA negatively. Thus, PA is an intermediate factor on the pathway between MM and DP. Mediation analysis is a common practice for the inspection of intermediate factors. Yet, for causal analysis, non-parametrical assumptions are required, mostly defined as sequential ignorability (SI). Translating longitudinal mediational settings to graphical causal models (DAGs) shows that SI is often violated due to exposure-induced mediator-outcome confoundment. In addition, classical approaches fail in parametric accountment for this type of confounding. As a solution, DAGs are derived that try to capture the longitudinal process between MM, PA and DP. The empirical testing is applied by marginal structural modelling. As a Database, the German Ageing Survey (DEAS) is utilized. Beck, AT (2008). The evolution of the cognitive model of depression and its neurobiological correlates. American Journal of Psychiatry 165(8):969–977. Defining and Identifying Discrimination Using DAGs ────────────────────────────────────────────────── Referent ╌╌╌╌╌╌╌╌ • Sebatian Wenz, GESIS – Leibniz Institute for the Social Sciences. Kurzzusammenfassung ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌ While some broader definitions virtually equate discrimination with unconditional inequality, some very narrow definitions limit discrimination to intentional treatment that harms individuals and results in inequality between groups. My contribution is based on both the substantive literature on discrimination as causal effect and the methodological literature on causality and Directed Acyclic Graphs (DAGs) in particular. I show that DAGs can help in both defining discrimination as causal effect in a substantively meaningful and empirically useful way and – once defined – identifying discrimination using different research designs. Insights from this endeavor include: • Defining and identifying discrimination are intertwined processes. • Virtually all – substantive and methodological – ideas that went into my definition and the corresponding DAGs are rather old (e.g., Rubin 1986); DAGs alone didn't help. • Discrimination is virtually indistinguishable from unconditional inequality if conceptualized as total causal effect (e.g., Blank et al. 2004) of characteristics that are assigned very early in life and, afterwards, immutable. • Solutions that use direct effect conceptualizations of discrimination (e.g., Pearl 2001, 2014) are themselves problematic; their substantive meaning changes when mediators are changed or added. • Ethnic discrimination and social class discrimination are typically confounded in all research designs unless explicitly addressed. • Depending on the definition, identifying discrimination becomes harder or easier using different research designs. Bringing Research Desing Back In ──────────────────────────────── Referent ╌╌╌╌╌╌╌╌ • Fabian Class, Universität Potsdam, PCQR Kurzzusammenfassung ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌ The choice of variables that are adjusted for in a statistical analysis is the single most important aspect in x-centered research designs. The choice of this "`adjustment set"' first and foremost depends the parameter of interest itself. It is thus impossible to make an informed choice without clearly defining the parameter of interest. Secondly, the choice of the adjustment set depends on assumptions about the causal relations between the covariates (as opposed to the associations between the covariates and the outcome). These assumptions should be thus made standout. Thirdly, the choice of the adjustment set depend on the covariate of interest. It is thus usually not sensible to interpret more than one coefficient in a statistical model. A review of all x-centered research papers published in the European Sociological Review in 2016 and 2017 (N = 118) shows that the vast majority of papers use insufficient reasoning for the adjustment set. Particularly, only a minority of papers clearly define the parameter of interest and disclose their assumptions about the causal relations between the covariates. In consequence, a huge number of papers interpret several coefficients of their models as if they had the same meaning. We conclude that parts of the ritualized research design in sociological papers can be described as an immunization strategy. The Art of Programming Questionnaires: Zur Umsetzung von Gestaltungsprinzipien für Fragebögen mit surveyAMC ─────────────────────────────────────────────────────────────────────────────────────────────────────────── Referentin ╌╌╌╌╌╌╌╌╌╌ • Claudia Saalbach, Universität Potsdam, PCQR Zusammenfassung ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌ Der Vortrag stellt das Programmpaket SurveyAMC vor, mit dem sich selbst-administrierte, papierbasierte, maschinenlesbare Fragebögen in hoher typografischer Qualität nach den wissenschaftlichen Standards der Fragebogengestaltung (Jenkins & Dillman, 1995) erstellen lassen. Zudem enthält SurveyAMC Begleitprogramme mit Hilfe derer sowohl Papier- als auch Online Fragebögen, Codebücher und Dateien zur Beschriftungvon Datensätzen, Ergebnistabellen und -grafiken automatisch aus der gleichen Quelldatei generiert werden können. Der Vortrag zeigt, warum ein hochwertiges visuelles Fragebogendesign für die Qualität von selbst-administrierten Umfragedaten wichtig ist, wie dieses mit surveyAMC umgesetzt wird und welche Vorteile die Arbeit mit einer zentralen Quelldatei haben kann. Graphical Causal Models for Survey Inference, External Validity, and Transportability ───────────────────────────────────────────────────────────────────────────────────── Referent ╌╌╌╌╌╌╌╌ • Julian Schüssler, Universität Konstanz Zusammenfassung ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌ We demonstrate the usefulness of graphical causal models to communicate theoretical assumptions about the collection of survey data, determine whether typical population parameters of interest to survey researchers can be recovered from a survey sample, and support the choice of suitable adjustment strategies. Starting from graphical representations of prototypical selection scenarios, we provide an explicit justification for the use of standard weighted regression estimators, which is missing in the literature. We then further discuss the semantics of selection bias versus transportability problems, summarize the recent methodological literature, and discuss areas for future social science research that can benefit from recent advances in graph literature in computer science and epidemiology. Multilevel Analysis with Few Clusters: Improving Likelihood-based Methods to Provide Unbiased Estimates and Accurate Inference | ──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────── Referent ╌╌╌╌╌╌╌╌ • Jan Paul Heisig, Wissenschaftszentrum Berlin für Sozialforschung (WZB) Zusammenfassung ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌ Quantitative comparative social scientists have long worried about the performance of multilevel models when the number of upper-level units is small. Adding to these concerns, an influential Monte Carlo study by Stegmueller (2013) suggests that standard maximum likelihood methods yield biased point estimates and severely anti-conservative inference with few upper-level units. In this paper, we seek to rectify this negative assessment. First, we show that maximum likelihood estimators of coefficients are unbiased in linear multilevel models. The apparent bias in coefficient estimates found by Stegmueller can be attributed to Monte Carlo Error and a flaw in the design of his simulation study. Second, we show how inferential problems can be overcome by using restricted maximum likelihood estimators for variance parameters and a t-distribution with appropriate degrees of freedom for statistical inference. Thus, accurate multilevel analysis is possible within the framework that most practitioners are familiar with, even if there are only a few upper-level units The Roads Not Taken: Graph Theory and Macroeconomic Regimes in Stock-flow Consistent Modeling ───────────────────────────────────────────────────────────────────────────────────────────── Referenten ╌╌╌╌╌╌╌╌╌╌ • Miguel Carrión Álvarez, Grupo Santander • Dirk Ehnts, Bard College Berlin Zusammenfassung ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌ Standard presentations of stock-flow consistent modeling use specific Post Keynesian closures, even though a given stock-flow accounting structure supports various different economic dynamics. In this paper we separate the dynamic closure from the accounting constraints and cast the latter in the language of graph theory. The graph formulation provides (1) a representation of an economy as a collection of cash flows on a network and (2) a collection of algebraic techniques to identify independent versus dependent cash-flow variables and solve the accounting constraints. The separation into independent and dependent variables is not unique, and we argue that each such separation can be interpreted as an institutional structure or policy regime. Questions about macroeconomic regime change can thus be addressed within this framework. We illustrate the graph tools through application of the simple stock-flow consistent model, or “SIM model,” found in Godley and Lavoie (2007). In this model there are eight different possible dynamic closures of the same underlying accounting structure. We classify the possible closures and discuss three of them in detail: the “standard” Godley–Lavoie closure, where government spending is the key policy lever; an “austerity” regime, where government spending adjusts to taxes that depend on private sector decisions; and a “colonial” regime, which is driven by taxation. Ökonomische Selektivität im politischen Engagement ────────────────────────────────────────────────── Referenten ╌╌╌╌╌╌╌╌╌╌ • Juliana Witkowski, Ruhr-Universität Bochum, Fakultät für Sozialwissenschaft, Sektion sozialwissenschaftliche Methoden und Statistik, • Simon Ress, Ruhr-Universität Bochum, Fakultät für Sozialwissenschaft, Sektion sozialwissenschaftliche Methoden und Statistik, Zusammenfassung ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌ In dem geplanten Vortrag werden die Analysemethode und -ergebnisse des Einflusses ökonomischer Faktoren auf die Bereitschaft, zum politischen Engagement, vorgestellt. Mit bürgerschaftlichem Engagement im Allgemeinen, zu dem auch politische Partizipation zählt, wird die Hoffnung verbunden, verschiedene gesellschaftliche Gruppen zu integrieren. Studien wiesen in der Vergangenheit darauf hin, dass sich die Partizipation in der Zivilgesellschaft zwar im Durchschnitt erhöht hat, bestimmte Formen jedoch rückläufig sind und sich auf bestimmte soziale Gruppen konzentrieren (Tesch-Römer et al. 2017; Bödeker 2012). Insbesondere im Bereich von politischem Engagement ist zu berücksichtigen, dass neben der sozialen Integration auch die Repräsentation einer breiten Masse der Bevölkerung in der Politik leidet, wenn einige soziale Gruppen unterproportional vertreten sind. Anhand der Daten des SOEPs wird geschätzt, wie das Einkommen auf die Bereitschaft und das Ausmaß, sich in Bürgerinitiativen, politischen Organisationen oder auf Ebene der Kommunalpolitik zu engagieren, wirkt. Dabei wird sowohl temporale als auch gruppenspezifische Heterogenität des Effekts zugelassen. Mit Hilfe von Directed Acycling Graphs wird gemäß des Potential Outcome Models (Rubin 2005) eine geeignete Auswahl von Confounder-Variablen aus dem bisherigen Forschungsstand abgeleitet, sodass Gültigkeit der Conditional Independence Assumption angenommen und ein kausaler Effekt geschätzt werden kann. Modelliert wird die Schätzung mittels Difference-in-Differences Propensity Score Matching. Bödeker, Sebastian (2012). Das uneingelöste Versprechen der Demokratie. Zum Verhältnis von sozialer Ungleichheit und politischer Partizipation in der repräsentativen Demokratie. in: Vorgänge. Zeitschrift für Bürgerrechte und Gesellschaftspolitik, 51(3)43-52. Rubin, Donald B. (2005). Causal Inference Using Potential Outcomes. Design, Modeling, Decisions. Journal of the American Statistical Association, (469) Tesch-Römer, Clemens; Simonson, Julia; Vogel, Claudia; Ziegelmann, Jochen P. (2017). Ergebnisse des Deutschen Freiwilligensurveys 2014: Implikationen für die Engagementpolitik, in: Simonson, Julia; Vogel, Claudia; Tesch-Römer, Clemens (Hrsg.), Freiwilliges Engagement in Deutschland. Empirische Studien zum bürgerschaftlichen Engagement. Wiesbaden: Springer. Workshop: Causal Graphs Basics ══════════════════════════════ Ort und Zeit ──────────── ━━━━━━━━━━━━━━━━━━━━━━━━━━━ 6. März 2020, 12:00–14:00 Universität Potsdam Campus Griebnitzsee Gebäude 6 Raum S21 ━━━━━━━━━━━━━━━━━━━━━━━━━━━ Beschreibung ──────────── The workshop discusses causal graphs as a fundamental modelling framework for empirical researchers in the social sciences. Questions addressed in interaction with participants include drawing and interpreting a graph, the connection to path diagrams and structural equations, understanding /d/-separation (with an application to instrumental variables), the nature of post-treatment bias, and simulating data from graphs. Referent ──────── Julian Schüssler ist PhD Student an der [Graduate School of Decision Sciences] und Mitglied im [Center for Data and Methods] der Universität Konstanz. Seine Dissertation befasst sich mit neuen Anwendungen von kausalen Graphen und /potential outcomes/ auf Instrumentalvariablen, Surveyforschung und kausalen Mechanismen. Julian Schüsslers Seminar zu [Causal Graphs] erhielt den /Causality in Statistics Education Award 2018/ der American Statistical Association. [Graduate School of Decision Sciences] https://www.gsds.uni-konstanz.de/ [Center for Data and Methods] https://www.polver.uni-konstanz.de/en/cdm/ [Causal Graphs] http://www.julianschuessler.net/graphs2018.html

1 0

Post-doc position in computational social sciences in Bremen (until Dec 2023) @EXC-Chair Mario L. Small
by Betina Hollstein 27 Jan '20

27 Jan '20

Liebe Kolleginnen und Kollegen, Colleagues, für eine Weiterleitung an ggf. infrage kommende Kolleg/innen wären wir dankbar! Thanks! Betina Hollstein & Mario L. Small The University of Bremen, U Bremen Excellence Chair Prof. Mario L. Small PhD/ Research Group "Large-Scale Data and Field Research in the Study of Social Networks" at the SOCIUM Research Center on Inequality and Social Policy, invites applications for the following position Post-Doctoral Researcher in Computational Social Sciences German pay-scale EG 13 TV-L, full time and limited until December 31, 2023 Start of contract: April 1, 2020 or earlier. The University of Bremen, a mid-sized university with 320 professors, 19.500 students and a full-spectrum of academic disciplines is one of Europes leading research universities and maintains close cooperation with international universities and non-university research institutions in the region. The promotion of young researchers is a core element of Bremen's research strategy. The 'U Bremen Excellence Chairs' program is a new initiative at University of Bremen since 2019. It enables internationally outstanding researchers to establish their own working groups at the university and to integrate them into a network of excellent research institutions worldwide in their respective disciplines. Mario L. Small's research group will study Large-Scale Data and Field Research in the Study of Social Networks. One major promise of big data was the ability to understand how social networks emerge, operate, and shape human behavior on a much larger scale than previously possible. However, as wide-eyed enthusiasm has given way to sober analysis, researchers and the public have become increasingly aware of the limitations of such data. This project will examine the extent to which field-based research (interview and survey research) can help address the limits of computationally intensive analysis of large-scale administrative data. Focusing on problems such as boundary specification, locality of interaction, algorithmic confounding, and misinterpretation of meaning, we will examine the extent to which bringing field research to bear on the analysis of large-scale administrative data can help improve our understanding of the relationship between networks and social inequality. The postdoctoral researcher will help design, execute, and write the results of research bringing field methods to bear on large-scale, administrative data analysis. Requirements: * expertise in social science computational methods and the management of large datasets, either expertise or demonstrated familiarity with research in social networks, and a completed Ph.D. degree in either sociology or a related social science. * demonstrated evidence of initiative, independence, drive, efficiency, productivity, the ability to work in groups, and a commitment to high quality social science research. * high level of proficiency in English (including demonstrated academic writing skills), proficiency in German is welcome, but not required. * experience with working in international projects, organizing academic events, and developing project proposals, preferably shown through employment, fellowship, or extended research stay at an international research institution within the past five years, is preferred but not required. * strong interest in developing connections with other researchers; experience gained at other national/ international research institutions is welcome. The researcher will also oversee a Ph.D. student at U Bremen who will conduct qualitative research. The postdoctoral researcher need not have direct expertise in qualitative interview methods or in survey methods, but some exposure to both methods is preferred. Salary and benefits are linked to the German employee scale TVL13 (100%). International candidates are highly encouraged to apply. The review process will begin on February 25, 2020. Applications and inquiries should be sent electronically as pdf-attachment by February 24, 2020 with the reference number A386/19 to <mailto:socium-bewerbungen@uni-bremen.de> socium-bewerbungen(a)uni-bremen.de Applications will be reviewed until the position is filled. Please note that the application should be submitted in English and that we can only accept it if it includes all required documents as a single PDF file: CV (with publication record), academic certificates, a sample chapter of your PhD thesis or other written paper, and a cover letter describing your background, interests, and motivation for applying. The application letter should also contain information about the data set(s) you have worked with. The University of Bremen has received a number of awards for its gender and diversity policies and is particularly aiming to increase the number of female researchers. Applications from female candidates, international applications and applications of academics with a migration background are explicitly welcome. Disabled persons with the same professional and personal qualifications will be given preference. For further information please contact Prof. Mario L. Small, PhD ( <mailto:mario_small@harvard.edu> mario_small(a)harvard.edu) or Prof. Dr. Betina Hollstein ( <mailto:betina.hollstein@uni-bremen.de> betina.hollstein(a)uni-bremen.de) Prof. Dr. Betina Hollstein University of Bremen SOCIUM - Research Center on Inequality and Social Policy, Head Mary-Somerville-Str. 9, R. 9.3090 D - 28359 Bremen tel +49 (0)421-218-58512 / 218-58638 (secr. Ms. Neumann) e-mail: betina.hollstein(a)uni-bremen.de <mailto:betina.hollstein@uni-bremen.de> http://www.socium.uni-bremen.de/about-the-socium/members/betina-hollstein/en /? QUALISERVICE Research Data Center for Qualitative Social Science Research Data, Head https://www.qualiservice.org/de/ - relaunch! Recent publications: <https://www.socium.uni-bremen.de/ueber-das-socium/mitglieder/betina-hollste in/publikationen/?publ=8749> What autobiographical narratives tell us about the life course. Contributions of qualitative sequential analytical methods, in: Advances in Life Course Research, online-first (18.12.2018), <http://dx.doi.org/10.1016/j.alcr.2018.10.001> doi:10.1016/j.alcr.2018.10.001 Collecting egocentric network data with visual tools. A comparative study, in: Network Science, forthcoming (with T. Töpfer & J. Pfeffer)

1 0

CfA RC33: "Contemporary issues in longitudinal and panel data analysis"
by Daniel Seddig 26 Jan '20

26 Jan '20

Dear Colleagues, ***We apologize for cross-posting*** We would like to invite researchers to submit abstracts to a session devoted to Contemporary issues in longitudinal and panel data analysis at The International Conference on Social Science Methodology (RC33) September 8th 11th, 2020, Nicosia, Cyprus. Deadline for the submission of abstracts is 31st January 2020. An abstract should contain max. 300 words. Session details: Contemporary issues in longitudinal and panel data analysis Longitudinal and panel data are used to study change within the units of interest (e.g., persons, groups, organizations, or countries). However, researchers may face several difficulties when working with panel data. For example, the identification of within-unit effects and its separation from stable variation between units as well as the correct specification of the direction of effects is a central issue for causal modeling. However, in many cases the issue of reversed effects is neglected. Furthermore, a prerequisite for valid analyses and comparisons of substantive constructs across time is that the measurements are actually comparable (i.e., equivalent) across occasions. However, often this prerequisite is only assumed but not tested. The methodological and statistical literature offers powerful tools to specify panel models that separate within- and between-unit variability and account for reciprocal causation (e.g., structural equation models with fixed effects) and to test the equivalence of measurements across time (longitudinal confirmatory factor analysis). This session aims at presenting studies that address particular analytical and conceptual problems associated with the broad issues outlined above. For example: · Causal modeling of panel data § Sequential exogeneity and reciprocal causality § Lagged dependent and lagged independent variables § Flexible specification and comparison/test of alternative models § ML/WLS estimation and handling of missing data · Measurement equivalence in panel data § How to deal with nonequivalence? Are small deviations tolerable? o Bayesian approximate methods, alignment method o Partial measurement equivalence § Sources of longitudinal measurement nonequivalence? o Systematic panel attrition o Life-course events o Developmental processes We welcome presentations that apply one of the methods mentioned above (or associated ones) and approach one of the issues related to panel data either using empirical data (1), and/or taking a methodological approach for example by using Monte-Carlo simulations. Best regards Daniel Seddig & Heinz Leitgöb -- PD Dr. Daniel Seddig University of Cologne Institute of Sociology and Social Psychology (ISS) dseddig(a)uni-koeln.de <mailto:dseddig@uni-koeln.de> <http://www.iss-wiso.uni-koeln.de/en/institute/staff/s/dr-daniel-seddig/> http://www.iss-wiso.uni-koeln.de/en/institute/staff/s/dr-daniel-seddig/

1 0

Deadline approaching (RC33): CfA "What About Causal Mechanisms? Analytical Approaches Using Longitudinal Data in Life Course Research"
by Katharina Loter 25 Jan '20

25 Jan '20

Dear Colleagues, we would like to remind you of the approaching deadline (one week left) for abstract submission for the RC33 Conference 2020 in Nicosia, Cyprus and draw your attention to our session: What About Causal Mechanisms? Analytical Approaches Using Longitudinal Data in Life Course Research The International Conference on Social Science Methodology (RC33): 8th-11th September 2020 in Nicosia, Cyprus http://cyprusconferences.org/rc33/ Session: What About Causal Mechanisms? Analytical Approaches Using Longitudinal Data in Life Course Research http://cyprusconferences.org/rc33/submission/ Deadline for Abstracts: 31st January 2020, abstract (max. 300 words) Session details: Research questions that motivate the great majority of life course studies are longitudinal and causal in nature. After several decades of research that produced numerous cross-sectional findings, the need for studies aiming at disentangling causal effects from non-causal associations, as well as mechanisms from confounding factors is becoming increasingly recognized. Following recent developments in causal theory, new opportunities for innovative life course research are being generated by the increasing availability of complex panel data with individuals observed for several years, belonging to different birth cohorts and socio-economic contexts, affected by period effects, and being nested in multiple structures (e.g., couples, households, families, social networks, school classes, schools, countries). Going beyond the estimation of the gross (causal) effect, researchers are increasingly interested in identifying and examining the intervening mechanisms. However, given the complexity of panel data, this could be very challenging. The purpose of our session is to meet challenges with regard to (causal) mediation analysis using panel data (e.g., within the framework of fixed-effects models, growth curves, DAGs, multilevel longitudinal models, SEM, event history models). We invite submissions examining and testing mediating mechanisms, moderated mediation or applying effect decomposition in life-course related longitudinal studies based on either regression or SEM framework. This could be studies using the age-period-cohort specification, multiple-cohort design or multilevel design. Substantively, we are particularly interested in research on social inequalities and/or similar topics. Best wishes and apologies for cross-posting, Katharina Loter & Oliver Arránz Becker Martin-Luther-University Halle-Wittenberg

1 0

WG: DAGStat-Symposium "KI in der Medizin" am 26.03.2020
by Tobias Wolbring 24 Jan '20

24 Jan '20

11. Symposium der Deutschen Arbeitsgemeinschaft Statistik (DAGStat) Donnerstag, 26. März 2020, 18.00 Uhr, Urania Berlin, Raum Edison An der Urania 17, 10787 Berlin Künstliche Intelligenz in der Medizin: Aufbruch in eine neue Ära oder leeres Versprechen? https://www.dagstat.de/aktivitaeten/symposium/ka14nstliche-intelligenz/ Wir laden Sie ein, sich über das Thema zu informieren und mit Fachleuten aus Statistik, Medizin, Wirtschaft und Politik den aktuellen Stand zu diskutieren!

1 0

Stellenausschreibung E14 (100%), BIBB
by Mergener, Dr. Alexandra 15 Jan '20

15 Jan '20

Liebe Kolleginnen und Kollegen, ich möchte Sie gerne auf die folgende Stellenausschreibung im Bundesinstitut für Berufsbildung (BIBB) in Bonn aufmerksam machen. Wir haben eine Wissenschaftliche Mitarbeiter/innen-Stelle (E14, 100 %) in einem BMBF-geförderten Drittmittelprojekt zum Thema „Polarisierung 4.0 – Fachkräftequalifikationen und Fachkräftebedarf in der digitalisierten Arbeit von morgen“ zu besetzen. Die Stelle richtet sich an promovierte Wissenschaftler/innen bzw. Wissenschaftler/innen kurz vor Abschluss der Promotion oder mit vergleichbarer Berufserfahrung in der Forschung. Der Ausschreibungstext findet sich hier: https://www.bibb.de/de/106154.php Die Bewerbungsfrist endet am 07.02.2020. Weitere Informationen zum Projektkontext sind hier zu finden: https://www.bibb.de/de/94793.php Wir freuen uns auf aussagekräftigte Bewerbungen und wären dankbar, wenn Sie/ ihr die Ausschreibung auch an geeignete Kandidatinnen und Kandidaten weiterleiten würden/ würdet. Mit bestem Dank und besten Grüßen Alexandra Mergener Dr. Alexandra Mergener ____________________________________________________________ Bundesinstitut für Berufsbildung (BIBB) Wissenschaftliche Mitarbeiterin Arbeitsbereich 1.2 “ Qualifikation, berufliche Integration und Erwerbstätigkeit“ Federal Institute for Vocational Education and Training (BIBB) Research Associate Section 1.2 “Qualifications, Occupational Integration and Employment” Robert‐Schuman‐Platz 3 D‐53175 Bonn Fon: +49 228 107 1414 www.bibb.de<http://www.bibb.de/>

1 0

Call for Paper | Special Issue, mda, Vignette Analysis: Methodology and Recent Developments
by Lena Verneuer 14 Jan '20

14 Jan '20

Liebe Kolleginnen und Kollegen, hiermit möchten wir gerne auf den Call für das Special Issue zum Thema 'Vignette Analysis: Methodology and Recent Developments' in der Zeitschrift /methods, data, analyses/ aufmerksam machen. Alle Informationen sind weiter unten in der Mail oder unter https://mda.gesis.org/index.php/mda/announcement/view/6 zu finden! Wir freuen uns über Einreichungen und die Weiterleitung an interessierte Kolleg*innen. Beste Grüße, Stefanie Eifler, Hermann Dülmer und Lena Verneuer ********************************************************************************************* *_Call for Papers: Special Issue of /methods, data, analyses/_* *Vignette Analysis: Methodology and Recent Developments* *Guest Editors: * Dr. Lena M. Verneuer (RWTH Aachen University) Prof. Dr. Stefanie Eifler (Catholic University of Eichstätt-Ingolstadt) PD Dr. Hermann Dülmer (University of Cologne) ** Vignettes have become an important tool in social science research to measure attitudes, normative judgments, and behavioural intentions. In general, vignettes are fictitious descriptions of a person or situation, which have to be judged or answered by the respondents. Drawing upon different disciplinary traditions, the term ‘vignette analysis’ embraces many different techniques that can be applied within the framework of non-experimental, quasi-experimental and experimental designs. As indirect measures used with non-experimental designs, vignettes allow for emphasizing situational aspects and are seen to solve problems of obtrusiveness and validity of direct measurements. Within the framework of the factorial survey approach, vignettes are combined with experimental designs (full-factorial or fractional). An important strength of factorial surveys is that by implementing them in surveys they allow for combining the high internal validity of experiments with the high external validity of surveys. In contrast to survey designs as they are typical for simple cross-sectional work, factorial surveys are full-fledged experiments and for this reason allow a strict test of causal relationships. Recently, several methodological challenges related to vignette analyses were addressed. Depending on theoretical and methodological objectives, the applied techniques vary in a broad range and lead to different and sometimes inconsistent results. Due to this diversity, findings on methodological and substantial issues can have different meanings and impacts for further research. Among these are for example the susceptibility of vignettes to social desirability, learning and fatigue effects and effects of order, variation, wording and presentation mode. Accordingly, the methodological challenges related to vignette analyses present a diverse field of research. This special issue chooses one way of anticipating this and aims at shedding light on the state of affairs by discussing recent developments and pooling new findings of projects that try to enrich the discussion. The focus of the call for papers is explicitly broad, and all contributions dealing with different analytical strategies or empirical designs that make use of factorial surveys or other vignette analyses are welcome. Papers matching one of the following aspects are cordially invited to be part of this special issue: -- comparison and discussion of design-related questions regarding methodological or substantive aspects, -- recent developments in measuring intentions with vignettes, - - theoretical ideas for modelling the relationship between intentions and behaviour for further empirical analyses, - - cross validation strategies (new approaches, replications), -- discussion of (dis-)advantages of vignette-designs, validations strategies and/or measures, - - issues of data-collection, - - substantive applications of factorial surveys ** *Submission procedure and timeline* If you are interested in submitting a manuscript to the mda special issue on “Vignette Analysis: Methodology and Recent Developments”, please upload your manuscript at https://mda.gesis.org/index.php/mda/about/submissions#onlineSubmissions following the guidelines of the mda until September 14, 2020. Guidelines for manuscript preparation can be found in the information for authors: http://mda.gesis.org/index.php/mda/about/submissions#authorGuidelines <https://mda.gesis.org/index.php/mda/about/submissions#authorGuidelines>. When submitting the manuscript, please indicate in the field “Comments for the Editor” that this is a manuscript for the Special Issue on “Vignette Analysis: Methodology and Recent Developments”. Upon submission, manuscripts will be peer-reviewed in accordance with standard journal practice. Accepted papers will be published Online First before print. The printed special issue will appear in October 2021. Queries about this special issue should be directed to Lena Verneuer at _lverneuer(a)soziologie.rwth-aachen.de_ *The detailed time schedule is as follows:* Deadline for paper submission: September 14, 2020 Decision on the full papers:December, 2020 Deadline for the submission of revised manuscripts: February, 2021 Final decision: April, 2021 Online first publications: May – August 2021 Publication of the Full Paper Special Issue: October 2021 ******************************************************************************************************** ** ** -- Dr. phil. Lena M. Verneuer Research Associate RWTH Aachen Institute of Sociology Eilfschornsteinstraße 7 52062 Aachen Tel: +49 241 80-90369

1 0

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Methoden