Research Guide: Developing the Research Question

So, you want to understand some sort of natural or social phenomenon in a logical, research-oriented way. How do you begin? Every well-structured research effort answers four questions before collecting a single data point:

• ·Research question (What?): Define a clearly scoped question of interest. Ideally this should be specific, falsifiable, and grounded in something you can actually observe or measure.
• ·Methodology (How?): How will you investigate this? Is the information already out there and needs to be synthesized in secondary research? Do you need to interview people, conduct ethnographic observation, or survey a population? Do you need to analyze existing data or compare how competitors approach the same problem? These methodologies should be clearly defined and matched to your research question.
• ·Definition of concepts (Which?): What concepts in your research question are you trying to understand? Brand strength? Engagement? Developer retention? Customer satisfaction? Understand the concepts inside your research question and what they mean.
• ·Operationalization (How do you measure it?): How will you measure the concepts you've defined? For example, brand strength might be operationalized as social media engagement numbers, content sentiment analysis, or share of voice in a category. Note that there will often be independent variables and dependent variables: we are trying to show that certain independent variables affect certain dependent variables in specific ways. That is, we are trying to show causality.

Once these questions have been answered, you can write a coherent research plan. You know what you are studying from the research question, how you are studying it from the methodology, and the specific things you are measuring from your operationalization. After conducting the study, you need to write up results.

Writing Up Results

Your write-up should follow a clear structure:

• ·Results: What did you find? What new things did your research uncover? Present findings as plainly as possible: separate observation from interpretation.
• ·Discussion: What do your results tell us? What is your interpretation of the answer to the research question? Where can you show causation, and where are you limited to correlation or association? This section should be a substantial part of your paper.
• ·Conclusion: What should the reader take away? What decisions does this research support, and what questions remain open for follow-on work?

Note: many research pieces written for operational teams will not spell out every element of this framework explicitly. But all good research touches on all of this: and being able to articulate these elements makes your work more credible and easier to act on.