1,721,001 research outputs found
Der Berner Stichprobenplan – Ein Vorschlag für eine effiziente Klumpenstichprobe am Beispiel der Schweiz
Get PDFDie meisten Bevölkerungsumfragen in der Schweiz beruhen auf Stichproben, die aus dem Telefonregister gezogen werden. Da solche Stichproben bezüglich der Abdeckung der Grundgesamtheit als problematisch anzusehen sind, diskutiere ich hier ein alternatives, von Fritschi et al. (1976) entwickeltes Stichprobenverfahren, bei dem in einem ersten Schritt Gemeinden ausgewählt und dann die Adressen der Zielpersonen über die Einwohnerregister bestimmt werden. Das Verfahren ist dabei so angelegt, dass die Stichprobe ähnlich wie bei einer einfachen Klumpenstichprobe auf eine relativ geringe Anzahl Gemeinden verdichtet wird, die Stichprobe aber trotzdem eine möglichst hohe statistische Effizienz beibehält. Eine Analyse der theoretischen Eigenschaften des Berner Stichprobenplans zeigt, dass das ursprüngliche Verfahren zu leicht verzerrten Stichproben führt. Eine korrigierende Modifikation des Verfahrens wird vorgeschlagen. Zudem wird ein alternativer Ansatz mit verbesserten Eigenschaften, die ex ante geteilte Stichprobe, vorgestellt. Mit Hilfe einer Simulationsstudie werden sodann die Vorzüge des Berner Stichprobenplans gegenüber einer einfachen Klumpenstichprobe illustriert
BSNP-A GAUSS Program for Semi-Nonparametric Maximum Likelihood Estimation of Binary Choice Models
No full textUsing Monte-Carlo-algorithms for the estimation of ß-error: alternative inference strategies for multidimensional contingency tables
Get PDFSmall samples and sparse cell frequencies cause major problems for statistical modelling with
categorical data: Sampling zeros or small expected frequencies can lead to situations where asymptotic approximations of test statistics will be inadequate. In such cases one resorts to the use of exact tests or Monte-Carlo-simulations. But also in this ease, inference can yield problematie results, as the power of tests is often extremely low and will therefore lead to the rejection of theoretically plausible hypotheses on the base of poor empirical material. In this paper an alternative modeHing strategy for small samples using Monte-Carlo-algorithms is presented. This strategy is extending the asymptotic power approximations presented by Cohen (1977) or Agresti (1990)
Computergestützte Kategorisierung verbaler Daten
Get PDFThe paper gives an overview about most recent developments in computer-assisted techniques
for the categorization and retrieval of unstructured verbal ("qualitative") data. These techniques allow the analyst to develop an organizing scheme for his or her data by defining linkages between the different elements of a qualitative data base: textsegments, codes and theoretical comments. These structures will offer possibilities for information retrieval which allow the systematic search for empirical evidence or counter evidence in the data. It is argued that the invention of computer-assisted methods represent a far-reaching methodological innovation for qualitative research by enhancing the rigour of the research process and the validity of its results
Going Beyond Counting First Authors in Author Co-citation Analysis
Get PDFThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
- …
