Methoden der Empirischen Sozialforschung und Wissenssoziologie

 

Schwerpunkte

Von 2009 bis 2012 war ich Professor für Empirische Politikforschung und biete immer noch Lehrveranstaltungen in diesem Bereich an. Neben der reinen Anwendung von Methoden und Analyseverfahren in meiner gegenstandsbezogenen Forschung beschäftige ich mich auch nach wie vor mit der Methodenforschung selbst. Mein besonderes Interesse gilt dabei der Umfrageforschung.

Aktuelles

Zusammen mit meinem Kollegen Jocelyn Evans habe ich eine Maßzahl entwickelt, mit deren Hilfe sich die Abweichungen zwischen Wahlumfragen und Wahlergebnissen durch eine einzige Größere ausdrücken lassen. Der zugehörige Stata-Befehl heißt surveybias und läßt sich einfach installieren: ssc install surveybias.

Publikationen

  • Arzheimer, Kai and Jocelyn Evans. “A New Multinomial Accuracy Measure for Polling Bias.” Political Analysis 22.1 (2014): 31-44. doi:10.1093/pan/mpt012
    [BibTeX] [Abstract] [Download PDF] [HTML] [DATA]

    In this article, we propose a polling accuracy measure for multi-party elections based on a generalization of Martin, Traugott, and Kennedy s two-party predictive accuracy index. Treating polls as random samples of a voting population, we first estimate an intercept only multinomial logit model to provide proportionate odds measures of each party s share of the vote, and thereby both unweighted and weighted averages of these values as a summary index for poll accuracy. We then propose measures for significance testing, and run a series of simulations to assess possible bias from the resulting folded normal distribution across different sample sizes, finding that bias is small even for polls with small samples. We apply our measure to the 2012 French presidential election polls to demonstrate its applicability in tracking overall polling performance across time and polling organizations. Finally, we demonstrate the practical value of our measure by using it as a dependent variable in an explanatory model of polling accuracy, testing the different possible sources of bias in the French data.

    @Article{ arzheimer-evans-2013,
    author  = {Arzheimer, Kai and Evans, Jocelyn},
    title = {A New Multinomial Accuracy Measure for Polling Bias },
    journal  = {Political Analysis},
    year = 2014,
    abstract  = {In this article, we propose a polling accuracy measure for
    multi-party elections based on a generalization of Martin,
    Traugott, and Kennedy s two-party predictive accuracy
    index. Treating polls as random samples of a voting
    population, we first estimate an intercept only multinomial
    logit model to provide proportionate odds measures of each
    party s share of the vote, and thereby both unweighted and
    weighted averages of these values as a summary index for
    poll accuracy. We then propose measures for significance
    testing, and run a series of simulations to assess possible
    bias from the resulting folded normal distribution across
    different sample sizes, finding that bias is small even for
    polls with small samples. We apply our measure to the 2012
    French presidential election polls to demonstrate its
    applicability in tracking overall polling performance
    across time and polling organizations. Finally, we
    demonstrate the practical value of our measure by using it
    as a dependent variable in an explanatory model of polling
    accuracy, testing the different possible sources of bias in
    the French data.},
    volume  = {22},
    number  = {1},
    pages = {31--44},
    url = {http://pan.oxfordjournals.org/cgi/reprint/mpt012?ijkey=z9z740VU1fZp331&keytype=ref}
    ,
    doi = {10.1093/pan/mpt012},
    data = {http://hdl.handle.net/1902.1/21603},
    html = {http://www.kai-arzheimer.com/new-multinomial-accuracy-measure-for-polling-bias}
    }

  • Arzheimer, Kai, Thorsten Faas, Ulrich Rosar, and Sigrid Roßteutscher. “Vorwort.” Innovative Methoden in der Politikwissenschaft. Eds. Arzheimer, Kai, Thorsten Faas, Ulrich Rosar, and Sigrid Roßteutscher. Vol. 6. Special Issue: Methoden — Daten — Analysen. Zeitschrift für Empirische Sozialforschung. Gesis, 2013. 69-72.
    [BibTeX]
    @InCollection{ arzheimer-faas-rosar-rossteutscher-2013,
    author  = {Arzheimer, Kai and Faas, Thorsten and Rosar, Ulrich and
    Ro{\ss}teutscher, Sigrid},
    title = {Vorwort},
    booktitle  = {Innovative Methoden in der Politikwissenschaft},
    series  = {Special Issue: Methoden -- Daten -- Analysen. Zeitschrift
    f{\"u}r Empirische Sozialforschung},
    volume  = 6,
    number  = 2,
    publisher  = {Gesis},
    year = 2013,
    editor  = {Arzheimer, Kai and Faas, Thorsten and Rosar, Ulrich and
    Ro{\ss}teutscher, Sigrid},
    pages = {69--72}
    }

  • Arzheimer, Kai, Thorsten Faas, Ulrich Rosar, and Sigrid Roßteutscher. Innovative Methoden in der Politikwissenschaft. Special Issue: Methoden — Daten — Analysen. Zeitschrift für Empirische Sozialforschung. Vol. 6. Gesis, 2013.
    [BibTeX]
    @Book{ arzheimer-faas-rossteutscher-wessels-2013,
    author  = {Arzheimer, Kai and Faas, Thorsten and Rosar, Ulrich and
    Ro{\ss}teutscher, Sigrid},
    title = {Innovative Methoden in der Politikwissenschaft},
    publisher  = {Gesis},
    year = 2013,
    series  = {Special Issue: Methoden -- Daten -- Analysen. Zeitschrift
    f{\"u}r Empirische Sozialforschung},
    volume  = {6},
    number  = {2}
    }

  • Arzheimer, Kai and Harald Schoen. “Ein Blick von außen. Anmerkungen zu Steinbrink et al. "Netzwerk(analys)e in der deutschen Humangeographie".” Berichte zur deutschen Landeskunde 86.4 (2012): 377-381.
    [BibTeX] [Abstract] [Download PDF] [HTML]

    Steinbrink et al.’s analysis of various knowledge networks in German Human geography offers fascinating glimpses into the structure of the subfield. We identify two potential problems, and one point of departure for further research. First, Steinbrink et al. limit their analysis to tenured professors, who are not necessarily representative for the the discipline as a whole. Therefore, their results may be biased. Second, they focus on publications in German journals. Without further information, we cannot know whether people who appear to be outsiders within this subnetwork are in fact well integrated once the global network of international journals in considered. Third, the charms of Social Network Analysis are sometimes too seducive. While snapshot studies such as the one by Steinbrink et al. provide intriguing insights, future work should be complemented by explicitly comparative perspectives and a first principles approach to "optimal" structures for knowledge networks.

    @Article{ arzheimer-schoen-2012,
    author  = {Arzheimer, Kai and Schoen, Harald},
    title = {Ein Blick von au{\ss}en. Anmerkungen zu Steinbrink et al.
    "Netzwerk(analys)e in der deutschen Humangeographie"},
    journal  = {Berichte zur deutschen Landeskunde},
    year = 2012,
    volume  = 86,
    pages = {377--381},
    number  = 4,
    abstract  = {Steinbrink et al.'s analysis of various knowledge networks
    in German Human geography offers fascinating glimpses into
    the structure of the subfield. We identify two potential
    problems, and one point of departure for further research.
    First, Steinbrink et al. limit their analysis to tenured
    professors, who are not necessarily representative for the
    the discipline as a whole. Therefore, their results may be
    biased. Second, they focus on publications in German
    journals. Without further information, we cannot know
    whether people who appear to be outsiders within this
    subnetwork are in fact well integrated once the global
    network of international journals in considered. Third, the
    charms of Social Network Analysis are sometimes too
    seducive. While snapshot studies such as the one by
    Steinbrink et al. provide intriguing insights, future work
    should be complemented by explicitly comparative
    perspectives and a first principles approach to "optimal"
    structures for knowledge networks.},
    html = {http://www.kai-arzheimer.com/netzwerkanalyse-deutsche-humangeographie}
    ,
    url = {http://www.kai-arzheimer.com/netzwerkanalyse-deutsche-humangeographie.pdf}
    }

  • Arzheimer, Kai. “Mehr Nutzen als Schaden? Wirkung von Gewichtungsverfahren.” Vom Interview zur Analyse. Methodische Aspekte der Einstellungs- und Wahlforschung. Eds. Schoen, Harald, Hans Rattinger, and Oscar W. Gabriel. Baden-Baden: Nomos, 2009. 361-388.
    [BibTeX]
    @InCollection{ arzheimer-2009d,
    author  = {Arzheimer, Kai},
    title = {Mehr Nutzen als Schaden? Wirkung von
    Gewichtungsverfahren},
    booktitle  = {Vom Interview zur Analyse. Methodische Aspekte der
    Einstellungs- und Wahlforschung},
    pages = {361--388},
    publisher  = {Nomos},
    year = 2009,
    editor  = {Schoen, Harald and Rattinger, Hans and Gabriel, Oscar W.},
    address  = {Baden-Baden}
    }

  • Arzheimer, Kai and Harald Schoen. “Isoliert oder gut vernetzt? Eine vergleichende Exploration der Publikationspraxis in der PVS.” Politische Vierteljahresschrift 50 (2009): 604-626.
    [BibTeX] [Abstract] [Download PDF] [HTML]

    Citations and co-publications are one important indicator of scientific communication and collaboration. By studying patterns of citation and co-publication in four major European Political Science journals (BJPS, PS, PVS and OEZP), we demonstrate that compared to the conduits of communication in the sciences, these networks are rather sparse. British Political Science, however, is clearly less fragmented than its German speaking counterpart.

    @Article{ arzheimer-schoen-2009,
    author  = {Arzheimer, Kai and Schoen, Harald},
    title = {Isoliert oder gut vernetzt? Eine vergleichende Exploration
    der Publikationspraxis in der PVS},
    journal  = {Politische Vierteljahresschrift},
    year = 2009,
    volume  = 50,
    pages = {604--626},
    url = {http://www.kai-arzheimer.com/arzheimer-schoen-netzwerke-politikwissenschaft.pdf}
    ,
    html = {http://www.kai-arzheimer.com/soziale-netzwerke-politikwissenschaft-publikationen}
    ,
    abstract  = {Citations and co-publications are one important indicator
    of scientific communication and collaboration. By studying
    patterns of citation and co-publication in four major
    European Political Science journals (BJPS, PS, PVS and
    OEZP), we demonstrate that compared to the conduits of
    communication in the sciences, these networks are rather
    sparse. British Political Science, however, is clearly less
    fragmented than its German speaking counterpart.}
    }

  • Arzheimer, Kai and Markus Klein. “Die Conjoint-Analyse als Instrument der empirischen Wahlforschung. Eine Projektbeschreibung.” Die Republik auf dem Weg zur Normalität? Wahlverhalten und politische Einstellungen nach acht Jahren Einheit. Eds. van Deth, Jan, Hans Rattinger, and Edeltraud Roller. Opladen: Leske und Budrich, 2000. 385-410.
    [BibTeX]
    @InCollection{ arzheimer-klein-2000b,
    author  = {Kai Arzheimer and Markus Klein},
    title = {Die Conjoint-Analyse als Instrument der empirischen
    Wahlforschung. Eine Projektbeschreibung},
    pages = {385--410},
    publisher  = {Leske und Budrich},
    editor  = {Jan van Deth and Hans Rattinger and Edeltraud Roller},
    booktitle  = {Die Republik auf dem Weg zur Normalit{\"a}t? Wahlverhalten
    und politische Einstellungen nach acht Jahren Einheit},
    year = 2000,
    address  = {Opladen}
    }

  • Arzheimer, Kai, Markus Klein, and Jürgen W. Falter. “Läßt sich die Total-Design-Methode auch auf Panel-Befragungen anwenden? Ein Projektbereicht.” Wahlen und Politikvermittlung durch Massenmedien. Eds. Bohrmann, Hans, Otfried Jarren, and Gabriele Melischk. Wiesbaden: Westdeutscher Verlag, 2000. 235-247.
    [BibTeX]
    @InCollection{ arzheimer-klein-falter-2000,
    author  = {Arzheimer, Kai and Klein, Markus and Falter, J{\"u}rgen
    W.},
    title = {L{\"a}{\ss}t sich die Total-Design-Methode auch auf
    Panel-Befragungen anwenden? Ein Projektbereicht},
    booktitle  = {Wahlen und Politikvermittlung durch Massenmedien},
    pages = {235--247},
    publisher  = {Westdeutscher Verlag},
    year = 2000,
    editor  = {Bohrmann, Hans and Jarren, Otfried and Melischk,
    Gabriele},
    address  = {Wiesbaden}
    }

  • Arzheimer, Kai and Markus Klein. “The Effect of Material Incentives on Return Rate, Panel Attrition and Sample Composition of a Mail Panel Survey.” International Journal of Public Opinion Research 11 (1999): 368-377.
    [BibTeX]
    @Article{ arzheimer-klein-1999,
    author  = {Arzheimer, Kai and Klein, Markus},
    title = {The Effect of Material Incentives on Return Rate, Panel
    Attrition and Sample Composition of a Mail Panel Survey},
    journal  = {International Journal of Public Opinion Research},
    year = 1999,
    volume  = 11,
    pages = {368--377}
    }

  • Arzheimer, Kai and Markus Klein. “Die Wirkung materieller Incentives auf den Rücklauf in einer schriftlichen Panelbefragung.” ZA-Info 43 (1998): 6-31.
    [BibTeX]
    @Article{ arzheimer-klein-1998,
    author  = {Arzheimer, Kai and Klein, Markus},
    title = {Die Wirkung materieller Incentives auf den R{\"u}cklauf in
    einer schriftlichen Panelbefragung},
    journal  = {ZA-Info},
    year = 1998,
    number  = 43,
    pages = {6--31}
    }