I am planning to perform experimentation in my research work to analyse the effects of certain process parameters on performance measures (using laser machining process). No doubt there is a research gap of some performance measures. But I am confused with the research approach that i can claim for my methodology. Either it is inductive or deductive approach?
During my many years of work in process engineering the methodology I used has been as follows: Step 1. Conduct a screening experiment to see which variables have some impact on the dependent variable. This does not have to involve a lot of experiments. A method such as Plackett-Burman design or Taguchi's L-8 or L-18 design would do. Once you have identified the key independent variables you can run a full factorial or half fractional factorial experiment in Phase 2 and collect the data. You can analyse the data for statistical significance and see which variables have an effect on the dependent variable and in which direction. In Phase 3 you can conduct using a smaller list of independent variables that made the most impact in Phase 2 a Response Surface Analysis (RSA) experiment to fine tune or optimize your process window. RSA is quadratic based.
Both inductive and deductive approaches are good and to be employed depending on the problem. In inductive method we come to a conclusion after our experimentation several times.
Many thanks Srini Vasan and Sankararaman S. for your valuable answers
- Badges
- Science method
- Similar topics
- Methodology