Start with a question

Categories: Other, Research, Research Design

Tags:

MethodSpace will explore phases of the research process throughout 2021. In the first quarter will explore design steps, starting with a January focus on research questions. Find the unfolding series here.


Let’s start at the beginning. What question(s) will you explore?

This month we will be looking at ways qualitative, quantitative, and mixed methods researchers think about research questions. Let’s start with the basics. This description is adapted from Byrne’s introduction to the SAGE Research Methods Project Planner.

Here are four possible approaches to developing your research question(s).

These types of questions are not mutually exclusive, that is, you might use more than one in a give study.

  1. You can develop questions that will allow you to develop a description of a social context or situation. You will need to focus on specific characteristics or relationships so you can develop an understanding of particular aspects of the social context or findings that can be generalized.

  2. You can develop questions as the basis for a study to test a hypothesis. This means you have constructed a specifically formulated statement which can be falsified. This is the dominant approach in experimental research. Usually you will employ statistical methods to test a null hypothesis, which asserts the opposite of the proposition you are testing. If, for example, your hypothesis is that there is a difference in political attitudes among the people of different ethnic backgrounds in a specific country, you might look at a sample of survey data to test this. In this case, the null hypothesis is that no difference in political attitudes can be found.

  3. You can develop questions about ideas you wish to engage with, without not a formal testable hypothesis. Studies of theoretical or literary topics might fit this category.

  4. You can design a study of a problem, with broadly defined questions. For example, your methodological approach may be based on grounded theory. You do not have a hypothesis that you wish to test, but rather you will work in context and the important questions will emerge as you interact systematically with the data you generate from your research.

Types of Research Questions: Why? When? Who? How? Where?

When developing and answering your research questions, be aware that these words have specific meanings in a social science context.

Why questions seek causal explanations. If you ask “why?” the answer begins with “because.”

When questions locate:

  • Events in relation to the time at which they happened
  • Processes in relation to when they happened and their duration
  • The setting of things in temporal order, or sequencing
  • Boundaries of the context of your research.

These last two types of “when” questions have implications for causality and for the generalizations you may be able to make from your research.

Who questions address agency. They seek to identify the persons, institutions, or collective bodies responsible for the things you are researching. You may use who questions simply to identify an informant: “Who told me this?” But you may also use them to indicate specific agency as the cause of an event.

How questions are about mechanisms. The answers describe ways in which things are done, which together result in a given outcome. These questions might cover:

  • How you carry out your own research.
  • The accounts you generate from your research to indicate how something happened.

Note that there is overlap between how and why questions when dealing with events or system states. Both are addressing cause.

Where questions (like when questions) set your research in context. It is important to understand the social world in terms of contextual spaces and circumstances. Both where and when questions should generate answers which help you define the extent to which you can generalize the results of your research.

Byrne, D. (2017). How do I develop a research question?. Project Planner. 10.4135/9781526408525.

Relevant MethodSpace Posts

Leave a Reply