Outcome and implications BUKU | SAIDNA ZULFIQAR BIN TAHIR (VIKAR)

After a hypothesis is confirmed or rejected in one study “one- shot”, replications are needed in order to enhance the robustness and the generalizability of the proposition. A replication strategy must be formulated in accordance with the researcher’s answers to such ques- tions as listed above regarding the outcome of the previous test. Testing of propositions by replication follows the same procedures as initial testing of propositions discussed above.

4.6 Summary

Theory-testing research is testing a proposition of a theory by confirm- ing or rejecting a hypothesis that is derived from that proposition in an instance of the object of study or in a group of instances or a popula- tion. After a hypothesis is confirmed or rejected in one study “one- shot”, replications are needed in order to enhance the robustness and the generalizability of the proposition. Four types of proposition can be distinguished: a sufficient condi- tion If there is A, then there will be B, a necessary condition B exists only if A is present, a deterministic relation If A is higher, then B is higher, and a probabilistic relation If A is higher, then it is likely that B is higher. Many business and management problems are formulated as necessary conditions, but most business research tests probabilistic relations. We argue that the experiment is the preferred research strategy for testing all types of proposition. In an experiment the independent concept A is manipulated and its effect on the dependent concept B is investigated. Confirmation in a well-conducted experiment is strong evidence for the existence of a causal relation. However, in the actual practice of business research, it is often not possible to create experi- mental conditions. If experimental research is not feasible, survey research is a good alternative for testing a probabilistic relation and case study research is a good alternative for testing deterministic con- ditions or relations. The survey is the second-best research strategy for testing a prob- abilistic relation. In the survey, a sample of a population is selected for the test, and a statistical analysis is conducted in order to test for proba- bilistic relations between the independent and dependent concepts. The survey is the third-best strategy for testing deterministic relations. Despite the widespread belief that case study research is not an appro- priate research strategy for theory-testing, we show that the case study is the second-best research strategy for testing deterministic relation. The single case study is the second-best strategy for testing a sufficient condition, or a necessary condition. The longitudinal single case study or the comparative case study is the second-best strategy for testing a deterministic relation. We will discuss in detail how to design and conduct a theory-testing case study in Chapters 5–7.

4.7 References

Goertz, G. 2003, The substantive importance of necessary conditions, Chapter 4 pp. 65–94 in Goertz and Starr 2003, Necessary con- ditions, theory, methodology, and applications. Oxford: Rowman Littlefield. Goertz, G. and Starr, H. eds 2003, Necessary conditions: theory, method- ology, and applications. Oxford: Rowman Littlefield. Ragin, C. 2000, Fuzzy-set social science. Chicago: University of Chicago Press. Van de Ven, A.H. 1989, Nothing is quite so practical as a good theory. Academy of Management Review ; 144: 486–489.