Study Design (I): The principles of scientific research
This week I took a course on “Study Design”. Here is a summary from the class.
The aim of science is to produce knowledge, to understand and explain some aspect of the world around us. — Singleton & Straits
A scientific process
As some of you may know, research can be categorized into two types: quantitative and qualitative. We can locate these two types of research in the cycle of a scientific process.
On the right side of the cycle, the research is being done by using data validate a hypothesis based on theories; this type of research usually relies on data analysis to find the relationship supported by the theory. On the left side, we first have a bunch of data among which we observe relationships developed into theories; this type of research can be qualitative research using questionnaires to find hidden problems or a machine learning model that reveal unexpected relations.
Generally, a hypothesis testing (e.g. a clinical trial) is a “deductive” approach. (“An inductive method” is a scientific process that generates a framework/theory)
Types of research studies
Taken together, from the scientific process, we can generally categorize research into three types that generate three types of scientific knowledge respectively.
- Exploratory/Descriptive: analyzing a situation or phenomenon; e.g. the prevalence of breast cancer → knowledge as description
- Explanatory (causal inference): analyzing the relationship between factors; ex: the relationship between x variable and y variable → knowledge as explanation and prediction
- Evaluation (causal inference): evaluating the effectiveness of a certain treatment/policy; e.g. clinical trials → knowledge as understanding
The “Exploratory/Descriptive” will locate on the left side of Figure 1. The “Explanatory” and “Evaluation” will locate on the right side of Figure 1.
What is “scientific research question”? — Principles underlying scientific research
To establish scientific research, we need to first learn how to ask a question.
A research question most of the time
1. includes two or more variables
2. is in a question form
3. is repeatedly answerable with evidence
After a question, three principles must be in place for the research to be “scientific”. (by the way, “science” means knowledge or experiement that can be proved with evidence repeatably)
- Empiricism (實證): the research question can be answered with evidence
p.s. Philosophical questions about essence, existence, or morality can not be answered by scientific research
- Objectivity (客觀): free of bias (which is difficult…)
- Control (控制): what factors to control in order to reach an objective result
Scientific research is systematic, controlled, empirical, and critical investigation of natural phenomena guided by theory and hypotheses about the presumed relations among such phenomena. — Kerlinger (1986), p. 10
Formulating a research study
Having all definitions clear for yourself, to start a research, you need to formulate the goal of your research. Start by answering the following three questions.
- Who is to be studied? — unit of analysis (see the next section)
- What are the domains of interest? — e.g. y variable
- What relationships will be examined? — e.g. linear or nonlinear, causality or association
Source: Research design & methods (30800106) by Taipei Medical University