A research question naturally emerges from a research problem that needs to be resolved. Developing a good research question not only rests on the study of some uncertain phenomenon, but also on the rational need for investigating it. It is therefore essential that a systematic review of extant literature on the research topic be conducted, prior to formulating a research question. Awareness of current trends and latest development in the field of study will certainly assist in formulating a relevant question. There may be several research questions, whether primary or secondary, but they should all be developed during the planning stage of a study. Furthermore, it must be ensured that secondary questions do not compromise the primary research question, which forms the basis of research objectives and hypotheses. Lastly, bear in mind that the number of research questions will inevitably make the research design and data analysis more complex.
Hulley et al. (2001) suggested the use of the FINER criteria in developing a good research question:
A poorly formulated research question may thus affect the choice of study design and hamper the chance of obtaining any significant finding, besides compromising the quality of the study.
First of all, it is essential to understand that it is a hypothesis, not the data, that drives a primary research question. Otherwise, given any dataset, it would be too simple to perform several tests and apply statistical techniques to establish significant associations and/or relationships among variables and/or constructs. In such cases, it would be working backward by using the data to develop the research question, and that would defeat the entire purpose of conducting the study. To make matters worse, spuriously positive findings may result.
Hypothesis testing aims at making inferences about the targeted population on the basis of a random sample taken from that population. A hypothesis must be formulated as a null hypothesis, generally meaning that a prevailing situation has not changed (in the case of finding differences) or that there are no significant relationships among variables and/or constructs. This is the reason why each null hypoythesis must be paired with an alternative hypothesis, should the outcome be significant. The two hypotheses must be mutually exclusive and comprehensively exhaustive, i.e., the acceptance of one would automatically imply the rejection of the other. For a better understanding of the concept of hypothesis testing, you might need to consult our statistician.
At this stage, all you need to know is that the development of a research hypothesis should be supported by a good research question, as it will influence your research design. Once appropriate hypotheses have been developed, you can safely proceed to the formulation of your research objectives.
You must first learn to distinguish between a research aim and a research objective. While an aim is written in broad terms and explains what is to be achieved at the end of the study, an objective is an active statement that is defined in measurable outcomes via a strong positive statement. The primary objective of a study is paired with the hypothesis of the study, and should be clearly stated in the introduction of the research protocol. Objectives usually state exactly the outcome measures that are going to be used within their statements. Strong verbs like determine, measure, assess, evaluate, identify, examine, investigate, etc., are used in the formulation of objectives.
The importance of objectives is that they guide the development of the protocol and design of study, and play a determining role in sample size calculations. Objectives should be focused on outcomes that are important and relevant to the study.
Research aim
To investigate the issue of student indiscipline and its impact on student attainment in Mauritian Secondary Schools
Research question
What are the various types of student indiscipline currently experienced in secondary schools of Mauritius?
Research (null) hypothesis
School management style does not impact on student attainment
Research objective
To identify the most common forms of indiscipline and their level of seriousness
Hulley, SB, Cummings, SR, Browner, WS, Grady, D, Hearst, N and Newman, TB (2001) "Designing clinical research: an epidemiologic approach" in Designing clinical research: An epidemiologic approach, p. 336.
Copyright © 2024, Stat Consul, All rights reserved. Developed by Algopage