According to Foss and Waters, a well-defined research question has six properties:
There are usually more than two theoretical constructs in any one research question.
Example: Within a thesis titled ‘An empirical investigation of health information system failure in regional Sri Lanka’ the theoretical construct is ‘health information system’.
You need to find and search for synonyms for the theoretical construct to ensure you do not miss finding important research because of superficial differences in vocabulary.
Example: ‘Public health planning’ is similar in meaning to ‘health information system’. Both these phrases are used in the literature and might need to be researched.
Theoretical constructs need to be recognisable and identifiable in the literature, known terms in the field of expertise and/or subject headings in the databases. The theoretical construct must be clear, precise and must conform to the technical usage in your discipline. As you formulate your research question – consider if you will be able to locate and distinguish the theoretical constructs easily from other constructs that appear in the literature.
Example: Instead of the broad topic: ‘health information systems’ choose ‘decentralised health information systems in developing countries’. Focusing on a subset of health information systems enables you, as a researcher, to recognise specific information relevant to a tightly defined construct.
This is because different kinds of data can be used to answer your research question and the significance of your research should go beyond the methods you have used to determine your answers.
Example: A researcher could collect data by: conducting interviews, conducting experiments, writing a meta-analysis or studying a particular region in depth to justify solutions applicable to broader geographic, demographic or socioeconomic groups.
There are some dissertations where the data might be specific enough to include in the research question.
Example: A creative industries dissertation may define a time period, a choreographer or type of artistic practice.
You need to convey what interests you about the theoretical construct and what will be different about it by the time your work is finished. Significance will reflect specialised knowledge in your discipline.
Example: Your new contribution might be to suggest ways to make health information timely and reliable to ensure evidence-based health planning in decentralised, developing world health systems. You begin your research knowing that health information systems lack good information; your new contribution will be to offer evidence-based insight and solutions to impact health information systems in developing nations.
You should not already know the answer to the question you are asking. Valuable research surprises your reader with new ideas or new relationships between existing ideas. Your research question needs to hint at surprising possibilities to increase the probability of original results.
Example of an unsurprising research question: ‘How do health information systems fail to provide reliable, timely information?’
This research question presupposes that all aspects of health information systems fail and that there are no possibilities other than to confirm this statement. A researcher is unlikely to be surprised with this sort of question because their data will simply confirm this judgment.
Choosing an unsurprising, expected approach with your research question compromises your capacity to contribute to your discipline with original publications. If you already know the answer to your research question at the start you are simply documenting known information, rather than researching.
Examples of surprising research questions:
‘How do climate-driven changes in the biophysical environment affect the sustainability of sub-tropical parklands?’
‘What are the impacts of Starbucks on the consuming patterns of its patrons?’
It should not be a question to which the answer is ‘yes’ or ‘no’ because such an answer will not produce a complex result.
Example of a non-robust research question: ‘Are public health information systems in Sri Lanka effective for accurate decision-making?’
Example of a robust research question: ‘How do regional health information systems provide information support for evidence based health planning and interventions?’
A robust question allows for surprising and complex results, such as discovering that there are some effective sub-systems within a broadly inefficient system. A well written research question makes research interesting to write and to read.