Title: Hypotheses testing with a small N : an example from federalism research
Authors : Hegele, Yvonne
Behnke, Nathalie
et. al : No
Extent : 15
Publisher / Ed. Institution : Sage
Publisher / Ed. Institution: London
Issue Date: 2019
License (according to publishing contract) : Licence according to publishing contract
Series : SAGE Research Methods Cases
Series volume: Part 2
Language : English
Subject (DDC) : 000: Generalities and science
Abstract: In federalism research, scholars are usually confronted with the problem of a small number of cases (small N). Only a limited number of federal or quasi-federal states exist which lend themselves for comparison. The same is true for within-state variation, as federal states have only a limited number of constituent units. What is more, those federal states vary along many institutional variables characteristic of their political systems, which would need to be controlled for in statistical analysis. Thus, federalism researchers frequently use methods such as single or comparative case studies and qualitative methods of data collection and analysis. While some scholars would argue that these methods are not fit for testing hypotheses, we demonstrate that hypotheses testing is also possible in a small N and qualitative data design. Yet, special attention needs to be paid to the problems of case selection and data analysis to enhance internal and external validity of empirical results. In this research case, we explain how we dealt with those problems and which methods we used to test hypotheses with qualitative data.
Departement: School of Management and Law
Organisational Unit: Institute of Public Management (IVM)
Publication type: Working paper – expertise – study
DOI : 10.4135/9781526478009
ISBN: 9781526478009
URI: https://digitalcollection.zhaw.ch/handle/11475/17485
Appears in Collections:Publikationen School of Management and Law

Files in This Item:
There are no files associated with this item.


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.