Validity measures how accurate and trustworthy the tools, data and outcomes of the research are. Construct validity explains how correct the
Respondent Company and Position Date of Interview Duration of Interview Place Respondent A Case Company Plc. 20.5.2013 65 min Helsinki
Online Retail Manager
Respondent B Case Company Plc. 24.5.2013 60 min Helsinki Online Business
Service Manager
Respondent C Vendor Ltd. 23.5.2013 85 min Helsinki
Principal Consultant
Respondent D Vendor Ltd. 23.5.2013 85 min Helsinki
Member of the Board
operational measures for the theoretical totality are, by comparing how well the data collection questions have been aligned with the research questions of the thesis. (Al Qur’an 2010, p. 113) Construct validity was accomplished in this thesis first of all by rigorously planning and asking questions during the multiple interview occasions, which provided comprehensive understanding about the main research issues in e-tailer business model creation and the technological aspects in formulating and operating the business model. As the in-depth interviews carried out included both business- and technical staff of the case company, and further included expert interviews from both business- and technological consultants of the technology vendor responsible for aligning business processes with the business model and technology with business processes, it could be stated that a more comprehensive and rich understanding would be very hard to maintain from a case study investigating the above mentioned components. Construct validity was further strengthened by the use of several sources of data (Eisenhardt 1989, p. 538), as documents of the business model and the technical solution were analyzed in addition to conducting the in-depth interviews.
External validity explains how widely generalizations can be made from the case study research findings. As the purpose of a case study is to describe the case under analysis which has purposefully been chosen by theoretical sampling to suit the theory as well as possible and not to describe the entire population, only analytical generalizations may be applied. Analytical generalization considers if the case study can be replicated and whether the research findings remain constant in additional case studies in the same research domain. (Al Qur’an 2010, p. 114)
Single case studies have been criticized in methodological literature, for possibly not possessing equal explanatory power for generalizations compared to multiple case studies. In her study called “A Case in Case
Study Methodology”, Meyer (2001, p.332-333) first states that single case studies include restrictions in generalizability and advices to include more than one case in a case study but also points out that one should avoid choosing several cases due to the fact that in such circumstances it may be impossible to gather the desired depth of data and pluralist perspective.
Meyer then continues on describing one of her own case study research methodology, in which she studied two separate cases and states “I chose two cases, which clearly does not support generalizability any more than does one case…”, which could be seen as a direct statement towards the number of cases under study not being a matter of quantity, but rather a matter of quality. Meyer continues her statement by adding that choosing only two cases enabled a deeper study on the cases, which turned out to be a good decision opposed to the situation in which she would have been analyzing a greater amount of cases. (Meyer 2001, p.333) Based on these notions, it seems valid to draw the conclusion of a single case approach being an equally suitable discipline compared to two or three cases, as long as the single case is able to provide the researcher with rich and in-depth data.
Flyvbjerg (2006, p. 224-225) also takes a stand on the issue of single cases not having the ability to provide evidence for generalization, and deems such arguments as incorrect, since the possibility to make generalizations differs on a case by case basis and also, on how the case has been chosen. Flyvbjerg (2006, p.225-226) addresses the issue by describing the famous story about dropping two objects from the Leaning Tower of Pisa and how Galileo was able to prove that Aristotle’s law of gravity was false. Galileo’s experiment did not include “observations across a wide range” or “observations carried out in some numbers”, but first and foremost Galileo’s effort to prove Aristotle wrong consisted of a conceptual experiment and only later of a single practical study.
External validity was reached in this study primarily by conducting thorough purposeful sampling in case study selection, and by gaining access to the company operating an e-tailer business model with a state of the art technology architecture, which could be said to best provide a holistic description of both the business model and the role of information technology in business model components and the operative business processes associated with the e-tailer business model. External validity was also reached by interviewing both business- and technology consultants of the technology vendor, and these individuals have been responsible for planning dozens and dozens of e-tailer business models, operative processes and technological solutions and it is questionable whether any single individuals in the nation do possess an equally holistic knowledge and experience about the research issues of this thesis compared to these two individuals interviewed.
Reliability is the concept which describes how consistent and well the implemented research design is able to measure the theoretical concepts which it is intended to measure. It also measures whether an additional researcher would be able to repeat the research by using the same research design and to duplicate the findings of the current thesis. (Al Qur’an 2010, p. 115) In this research reliability was achieved by using the described and documented case study protocol, in which all the interviewees were asked the same interview questions. The interview questions are illustrated in Appendix 1 of this thesis. The interviews were also recorded. Furthermore, a factor increasing the reliability of this thesis is the in-depth interviews of the senior managers from both the case company’s business- and IT-departments responsible for the implementation of the e-tailer business model. It could be stated that a more accurate description of the case company’s motives, procedures, processes and technologies regarding the e-tailer business model could not have been maintained from any other sources of information. This information was further backed up with documentation of the business
model and technology architecture. In addition to interviewing internal representatives of the selected case study, even experienced business- and IT consultants of the technology vendor were interviewed in order to additionally describe their viewpoints of the case company’s situation with regard to the research issues of this thesis. These consultants possess enormous experience of creating and operationalizing e-tailer business models in terms of concepts, process and information technology architectures which constitute to the reliability of this thesis.