Conducting Empirical Research
Empirical research derives its knowledge directly from scientific observations and measurements. It allows researchers to answer a wide variety of research questions. The idea of empirical research dates back to ancient Greece. Over the years, methods have been developed to make empirical research more rigorous. However, a standardized process for conducting empirical research allows researchers to better compare studies and evaluate them for their rigor and accuracy.
The purpose of this paper is to provide an overview of each step in the standardized process for conducting empirical research. In general, empirical research can be conducted using quantitative or qualitative research methods. The steps described in this paper are mainly applicable to quantitative research in the social sciences but can be adapted to other scientific areas and methodologies with slight modifications. Each step in the process will also become a part of your final research article.
1. Develop Basic Research Questions
Empirical research starts with a broad question about a specific research topic. This question will be refined further along in the process after you perform the literature review. The goal of an empirical research question is usually inference: drawing a conclusion about a research topic without complete certainty (Lavraka, 2008). Davis (1971) provides suggestions on how to develop relevant research questions. There are different types of empirical questions such as descriptive questions, relational questions, and explanatory questions (Andrews, 2003).??
Descriptive questions aim to describe the facts about a certain research topic. Relational research questions examine the relationship between different concepts. A variation of relational research questions is predictive research questions. Predictive research questions seek to determine how variable y would change if variable x happened. Explanatory research questions aim to detect causal relationships between concepts and establish causal inference. The goal is to determine if there exists any cause-and-effect relationship between variable x and variable y. A variety of criteria should be met to suggest causation (Hill, 1965). There are a variety of other types of empirical research questions. For example, design questions ask what the best way is to design a product or a study and explore different metrics to measure the effectiveness of different designs. Comparative analysis research takes competing events or ideas and tries to determine if one is better than the others.
Note: The description of the research question and why it is important is the main body of the Introduction section for your article.
2. Review the Literature?
The next step after developing a broad research question is to find out if anyone has answered that question or similar questions before. To this end, you need to review the literature related to the research area. This step in the research process has several objectives:
Note: This literature review serves as the basis for a Background / Introduction / Literature Review section for your article.
2.1. Build a Knowledge Base Related to the Research Question
Once a general research idea has been developed, a review of the literature is needed to identify existing research related to the research question. This requires searching through the literature and textbooks related to the subject. Databases for scientific literature such as Google Scholar or PubMed serve as a good starting point for a literature search. The goal of this step is to better understand the topic and synthesize relevant research that has already been conducted related to the research question.?
Note: Keep a record of the articles you read and their important points, as you may reference them in your article later on.
2.2. Review Theories Underlying the Research in the Field
Similar to a general review of the literature, a review of related theories is advised. Theory should be a foundation for every research project (Suddaby, 2014), and the goal of the research is often to test or refine a theory (Whetten, 1989). To identify theory, seminal work about the research topic should be consulted. Many authors include a discussion of relevant theory in the introduction section of their articles. In their seminal work, Sutton and Straw (1995) discuss how to identify theory in research (Sutton & Staw, 1995).
In more advanced research fields, it is common to propose a revision of an established theory at the beginning of your research project with the objective of testing your proposed theory.
Note: You would describe the theoretical foundation for your study or your proposed theory in a Theories section in your article.
2.3. Identify Research Gaps?
The last part of the literature review step is to identify gaps in the research. Think about the approaches used in the papers as you read them. What assumptions are made and how do the authors attempt to solve their problem? Are there new data or ideas that could improve the methods used? What other similar questions are there without an answer and how could you use newly available data and methods to answer them better?
Many authors will put forward unanswered questions and avenues for future research in their research articles. They are usually in the future research or conclusion section of the papers. These discussion points can be used as starting points for your research project.
Note: These research gaps will serve as the Problem Statement / Gaps in the Literature for your article.
3. Refine Your Research Questions
Based on the review of the literature, and especially based on the research gaps you have identified, your next step is to refine your research question. Empirical research is usually focused on two to three specific questions. The research questions are based on the research gaps and the underlying theory identified in the literature review. Research questions might change slightly based on the analytic approaches you will use. In order to answer your research questions, you will need to form your hypothesis, determine your research methodology and develop your models.
Note: Oftentimes, research questions are listed in an Aims and Objectives section at the end of the Introduction section.
3.1. Form Hypotheses
Hypotheses are based on your research questions. Hypotheses state what you expect to find in your data based on the literature you have reviewed. Developing clear hypotheses for your research project can help to determine the analytic approaches you want to use. Davis (1971) can be used as a guide to develop strong hypotheses. An example hypothesis for the research question “What is the relationship between smoking and caffeine consumption?” is: “Smoking leads to higher caffeine consumption.” Not all empirical research articles list specific hypotheses, especially if the research is of an exploratory nature or research questions are bidirectional.?
Note: If you generate hypotheses for your research, you will list those in an Aims and Objectives section at the end of the Introduction section.?
3.2. Determine Your Research Methodology
Empirical research offers various approaches to answer your research questions. The methodology you use will depend on your scientific field and on the type of research questions you have developed. In this step, you describe in detail what type of study you are conducting (e.g., cross-sectional, experimental, longitudinal) and how you select and recruit your population (e.g., online survey platform, patients of an integrated hospital system). In human subjects research, this section typically includes a brief description of ethical concerns describing the consent process and listing the IRB application number.
Note: You will describe your study methodology in the Methods section of your article using subheadings such as “Study Design,” “Setting,” “Study Population,” and “Human Subjects Protocol.”
3.3. Develop a Model
Based on your research questions/hypotheses and research methodology, it can be helpful to develop a model including your main variables and potential relationships among them. Refining your model is an interactive process and will depend on the statistical analysis you apply to test your research questions or hypotheses.
4. Collect Your Data
At this point in the research process, you have determined your research questions and hypotheses as well as your methodology and model. The next step is to collect your data, which you will use to test your hypotheses.
Note: This part of empirical research will make up the Methods section of your article.
You will need to describe which data you will collect and how you will collect it. For empirical research, you can use primary or secondary data. Primary data is data you collect specifically for the purpose of your study, whereas secondary data is collected from large data repositories, like national surveys (e.g., NHANES, MIDUS, NHES). Both data approaches have advantages and disadvantages. Your study methodology will determine which approach is more suitable for your study. To collect the right data for your research questions, a clear delineation of your variables of interest is needed. The variable selection is oftentimes based on the literature review and the theory you chose as a foundation for your study.?
There are two important factors that we are going to discuss now.
Note: This will be the basis for the Data Collection and Variable Description paragraphs within the Methods section of your article.
4.1. Primary Data Collection
Primary data comes in many different forms. Examples are:
4.2. Secondary Data Collection
Secondary data often comes from large data repositories containing hundreds of variables but can also be data you or a colleague has collected for another study. Depending on the research questions, different surveys provide relevant information. For example, the National Health and Nutrition Examination Survey (NHANES) contains information on a variety of health behaviors and health conditions.?The National Household Education Survey (NHES) contains information on education status and outcomes.
5. Test Your Hypotheses
This step of the research process constitutes the statistical analysis of the data you collected. Clearly outline which statistical analyses you will use to answer your research questions and test your hypotheses. The specific analyses will depend on the nature of the research questions and the type of data you have collected. Articles reviewed during your literature review can provide some guidance as to which analytic approaches might be appropriate for your research questions.?
Note: This will be the basis for the Test Your Hypotheses or Data Analysis paragraph within the Methods section of your article.
6. Report Your Findings?
Once you have completed your analyses, you can report your findings. Here, you want to focus on the outcomes of your statistical analyses. Provide basic information on demographics (e.g., the average age of your participants, gender distribution) and your main variables (e.g., the proportion of smokers among your participants, average cups of caffeinated beverages per day). Many research projects also benefit from bivariate analyses to examine the fundamental relationship between two variables (e.g., average age for smokers vs. non-smokers). After providing these descriptive data, a common approach is to report your findings in the order of your research questions/hypotheses. Capture which analyses you have conducted for which research question/hypothesis and what the results of these analyses are.
领英推荐
Note: This will make up the Results section of your article.
7. Discuss the Implications of Your Findings
The next step of the research process is to interpret your findings. Refer to existing literature to compare your results and explain why they are similar or different. What new ideas can we propose with this knowledge? Were you able to address the gaps you listed in your introduction??
Make sure you do not just repeat the results in this section.?
Note: This will make up the Discussion section of your article.
7.1. Discuss Your Results Related to Theory
In this section, you discuss how your results agree or disagree with the theory you have chosen or proposed as a basis for your study. If your results align with the theory you have chosen, you can confirm that the theory applies to your research topic. It is quite likely, however, that your results do not completely align with your theory. Discuss the implication of your findings to refine your theory or propose a new theory and develop new research questions or hypotheses.?
Note: This will make up part of the Implications paragraph within the Discussion section of your article.
7.2. Discuss Avenues for Future Research
As part of your discussion, explore avenues for future research. This section can describe gaps in the literature and can be used as a basis for your own future research. Provide enough information to describe the research gaps but do not disclose your entire research plan. In this section, you are able to think beyond your research questions and consider other factors that might influence your results.
Note: This will make up part of the Future Research paragraph within the Discussion section of your article.
7.3. Discuss Strengths and Limitations of Your Study
The last step in your research process serves as a forum to critically evaluate your research. Highlight specific strengths of your study (e.g., large sample size) and limitations (e.g., used proxy variables). The reviewers of your article will want to see a candid commentary on your study limitations; however, make sure you do not devalue your study.
Note: This will make up part of the Strengths and Limitations paragraph within the Discussion section of your article.
8. State Your Conclusion
The conclusion section summarizes the most critical takeaways from the study. It should be a brief statement and only a few sentences. Your conclusion should be an impactful statement about your study, not a reiteration of your results.
Note: This will make up the Conclusion paragraph of your article.
9. Create an Abstract
A scientific abstract often uses the same structure as the article. An abstract provides a brief overview of the study and contains the most important results and discussion points. The length of an abstract varies depending on the journal.
Abstracts can be unstructured (one coherent paragraph of text), or structured (including subheadings). For the unstructured abstract you can summarize your conclusion. The subheadings for structured abstracts read for example “Background,” “Purpose,” “Methods,” “Results,” “Discussion,” and “Conclusion.” Many scientific journals include guidelines for the format of article abstracts in their submission guidelines/instructions for authors.
10. Format Your Article
A research article generally has four sections: Introduction, Methods, Results, and Discussion. Consult the journal’s guidelines for the structure and formatting of an article. Some journals have distinct requirements for the sections and subsection headings. It can be helpful to create a checklist of formatting and submission guidelines based on the journal’s website. Such a checklist would contain information on:
Below is an example outline for a research paper:
Abstract
Introduction:
Methods
Results
Discussion
Conclusions
Congratulations, you have conducted empirical research!
References
1965, A. B. Hill, "The Environment and Disease: Associations or Causation?," Proceedings of the Royal Society of Medicine.
1971, M. S. Davis, "That's interesting: Towards a phenomenology of sociology and a sociology of phenomenology," Philosophy of the Social Sciences.
1989, D. A. Whetten, "What Constitutes a Theoretical Contribution?," The Academy of Management Review.
1995, R. I. Sutton & B. M. Staw, "What Theory is Not," Administrative Science Quarterly.
2003, R. Andrews, "3. Formulating research questions," Research questions.
2008, P. J. Lavraka, "Encyclopedia of Survey Research Methods," SAGE.
2014, R. Suddaby, "Editor's Comments: Why Theory?," Academy of Management Review.
To read the complete version, visit here:
Author: Daniel Armani & Philip Placek