Inquiry Strategy: inductive versus deductive

Sometimes called the research approach, our inquiry strategy is the spirit in which we will undertake our research project. It needs to fit with our research philosophy, our research design, and research method, or methods (Creswell's "research onion", 2009). Some researchers lump this with the research method as part of the research approach, but I know it as a separate decision step, as inquiry strategy.

The inquiry strategy falls largely into three main clusters: inductive, or deductive, or a combination of both, as follows (Heit & Rotello, 2010; Veal, 2005):

1. Inductive research begins from a having a vague idea of what may be a causal factor, but we sit back and wait to see what comes up out of the data, once we have collected it. We use very open-ended questions, and, in some ways, allow our participants to direct our research findings. Inductive research fits with subjective research philosophies, qualitative data and interview, focus group or open-ended survey data gathering tools. This type of approach works best when we have a small number of participants, be more in-depth, and look for common factors across our dataset. We use textual analysis, and CAQDAS software to analyse our data (perhaps using software such as NVivo, QDAMiner or even MS Excel to filter data and look for similarities). Once we have analysed our data, then we draw conclusions and perhaps find a theory to match our findings.
2. Deductive research starts with a clear theory that we are testing for, before we collect our data. We use what is called 'syllogistic reasoning', which starts with an idea of our outcome, then a second outcome, then the result; ie, 'Management is a social science. Social sciences are about human relationships. Therefore, management is relationship-oriented'. We test for the presence of a causal factor, or for correlation between two ideas. Deductive research fits with objective research philosophies, statistical testing using SPSS (or similar) and experimental or lab testing data gathering tools. This type of approach works best when we have a large number of participants or samples.
3. Mixed methods are a mix between the two, where a project perhaps requires two data sets, and one can be approached inductively, and the other deductively. This gives us another opportunity to triangulate our data within a research project and can be particularly useful for reliability when exploring a new area of research.

Once we know the differences, it isn't that hard to work out what our approach should be :-)


Sam

References:

• Creswell, J. (2009). Research Design (Third Edition). USA: Sage
• Heit, E., & Rotello, C. M. (2010). Relations between inductive reasoning and deductive reasoning. Journal of Experimental Psychology: Learning, Memory, and Cognition, 36(3), 805.
• Veal, A. J. (2005). Business Research Methods (Second Edition). Australia: Pearson Education