There is nothing in the research process that is more important than getting the question(s) right. If the questions are good there is a chance that the study will be good. If the questions aren't good then there is no hope that the study will be good. The nature of the questions range from very deductive focused questions about specific relationships to broad descriptive inductive questions about the system.  The questions should be designed so that the answers to the questions will produce the knowledge identified in the purpose statement.  Cooper and Schindler (2003) propose that there is a question hierarchy that connects what they call the management dilemma (a.k.a. problem statement) with the management decisions (p. 66).

Quantitative Questions

Descriptive and Inferential Questions – In quantitative studies the research may as Creswell (2003) proposes “describe responses to the independent, mediating, or dependent variables” (p. 109). Descriptive questions focus on a single variable. Inferential questions focus on a relationship between variables. Many research studies have a combination of descriptive and inferential questions. In these case the answers to the descriptive question on single variables are inputs to the inferential questions on the relationships between those variables. This maybe the most attractive approach because it is doable and it does result in actionable information. Quantitative questions often manifest themselves in two approaches - correlations and experiments.

The correlation questions maybe the most popular in the organization and management fields primarily because they are possible to study using instruments such as surveys and they do not require as much of a solid research base as descriptive studies. While they do not establish causation (only experiments really do that) the reason that we look into correlations is that we suspect that a relationship exists that we might be able to act upon. They can be a precursor to descriptive studies.    

The experimental questions focus on finding support for causal relationships between independent and dependent variables. We seldom do experiments in organization and management studies simply because it is seldom that we can control all the variables that are factors in organizations. The exceptions are often found in operations research. While it is challenging to use experimental methods in organizations, experiments are the "gold standard" of research and consequently organization and management researchers will often use quasi-experimental approaches.

Qualitative Questions

Exploratory or Discovery – Closely related to the descriptive questions are exploratory or discovery questions that seek to get at the nature of some phenomenon and not only describe it but also “explain how” it works.

Quasi-deductive – The descriptive and exploratory questions are typically open-ended and support a primarily inductive approach. Miles and Huberman (1994) propose qualitative research approach that is not purely inductive but actually takes a quasi-deductive approach. In the quasi-deductive approach questions are formed directly from the conceptual framework which typically includes variables, relationships, context, etc. (p. 22).

Hypotheses?

Questions are just that - questions - and by themselves they do not include an answer. Hypotheses, on the other hand, are the predicted answers to the questions. Questions alone are usually associated with theory building and exploratory studies which are often flexible and often qualitative or mixed.

Questions plus hypotheses are usually associated with theory testing studies which are often fixed and quantitative. Qualitative methods are usually too limited to be credible for theory testing. However, there may be a rare exception. Now, the tricky part is what does it take to create a hypothesis? A hypothesis is not simply a guess. It is a logical conclusion based on the results of previous research. It is the testing of theory that has already been built by previous studies. So, in order to even have a hypothesis there must be existing research that built the theory that is to be tested. There are rare studies that are mixed in that they "finish" the theory building with a qualitative portion and then test a hypothesis based on that preliminary work. This is unusual however particularly for doctoral dissertations because it is simply a lot of work and usually more work than is required to complete a doctoral thesis.

Example - Chad McAllister [Prospectus Version]

Management Question:  Given that misunderstood requirements are largely responsible for software development failures, how do we get the requirements right and decrease our failure rate?

 

Research Questions:  (1) What are the factors/variables that influence users and developers understanding requirements for software products?  (2) How influential are the factors identified in question one?  (3)  How do the factors differ between users and developers? 

Investigative Questions: (1)

1. What do users say the factors are that influence misunderstanding requirements? 

2. What do developers say the factors are that influence misunderstanding requirements? 

3. How do users and developers prioritize the factors that influence understanding requirements?

4. How do the factors identified and prioritized compare between users and developers?

5. What are possible explanations for the differences in the factors between users and developers? 

Example

Example Conceptual Framework with Research Questions:

Click on diagram for larger PDF view.

Align and Integrate

1. The research questions link directly to the literature review. The literature review should identify what we already know about the variables and relationships identified in the research questions.

2. The research questions should be crafted so that the answers produce new knowledge and insights that will fulfill the purpose and will help resolve the problem.

3. As with all the components of the research methodology the research questions (variables, relationships, and context) should be consistent with the variables, relationships, context, etc. identified in the conceptual framework.

References

Cooper, D. R. & Schindler, P. S. (2003). Business research methods (8th ed.) Boston: McGraw-Hill Irwin

Creswell, J. W. (2003). Research design: Qualitative, quantitative, and mixed methods  approaches (2nd ed.). Thousand Oaks: Sage.

Miles, M. B. & Huberman, A. M. (1994). Qualitative data analysis: An expanded sourcebook (2nd ed.). Thousand Oaks: Sage