3.1 Research Design 41
41
been conducted in this research so that patterns could be identified. These patterns
have been compared to findings from this study to shape final conclusions.
A further quality assurance criterion is construct validity. It refers to the extent
to which a study investigates what it aims to investigate. Two measures enhance
the construct validity: triangulation and a clear chain of evidence (Gibbert et al.
2008). This study achieved a sufficient level of construct validity through data tri-
angulation by using transcribed interviews, corporate documents and case notes as
source for conclusions. Furthermore, a chain of evidence is used to ensure that the
research process from formulating research questions to drawing conclusions is a
comprehensible process for the reader. Chapter 3.2 provides a detailed description
of the steps employed during the research process. The results are described through
ample data such as interview quotes to make conclusions reasonable. Besides its
importance for construct validity, the chain of evidence plays an essential role in
the context of reliability (Yin 2014; Gibbert et al. 2008). This importance will be
explained later in this chapter.
The third and last type of validity is external validity, often called generalizability.
This describes “the extent to which findings (...) can be analytically generalized
to other situations” (Yin 2014, p. 238). Although case studies are not designed for
statistical generalization they offer the possibility for analytical generalization.
Analytical generalization describes building theory from empirical findings and
qualitative case studies are a good starting point to develop theory (Gibbert et al.
2008; Eisenhardt 1989).
Eisenhardt (1989) further suggests cross-case analysis of four to ten cases to begin
building a theory in order to be able to conduct an analytical generalization. By
doing so, the complexity is kept at a manageable level whilst achieving saturation.
Likewise, it is important that the researcher provides his/her theoretical reasons
for the case selection so the reader is able to “appreciate the researchers’ sample
choices” (Gibbert et al. 2008, p. 1468). For this thesis nine cases were selected. The
theoretical considerations leading to the selection of each case will be explained
in detail in chapter 3.2.1.2.
Reliability describes the fact that subsequent researchers conducting the same
study would come to the same conclusions (Gibbert et al. 2008). It is therefore im-
portant for the researcher to provide a transparent and comprehensible description
of the research process as well as the previously mentioned chain of evidence (Yin
2014). For the purpose of making the study replicable, Gibbert et al. (2008) suggest
the creation of a case study database to store and organize transcripts, notes and
all other documents collected during the course of research.