Here is a suggested general research framework to help guide you through the steps involved.
Step 1: An observation
A research study starts with an observation. This could be along the lines of:
- Indication of something hidden that might be revealed by a study
- The occurrence of an event for which there is no adequate explanation
- An apparent relationship between items which is not explained by current theories
Step 2: Define the research question
The observation leads to a question you would like to answer. Formulate the question as clearly as you can, making sure you understand exactly what population you are interested in and what result(s) are important.
When looking for a question, ensure that it is:
- Something about which you are very interested
- Not a question that has already been answered
- Something that is possible to answer with the time and resources at your disposal
- Balanced and without bias
- Grounded in one or more existing fields or disciplines
- Pointed towards clear and focused argument and action
- Something that others will also find interesting
Step 3: Develop study objectives
Convert the research question into predictive and precise statements which must be “answerable”. The study specifics flow from the objectives.
Step 4: Design
The purpose of the design is to determine a method which creates accurate and unbiased data from which valid conclusions may be drawn.
Here are several different types of study designs that investigators may follow:
relies on a literature review or uses a physician’s clinical cases to introduce a clinical abnormality, usually something rare, new, or something that presents in an unusual way
compares the number of people who had a potential risk factor in the case group (those with the disease) with those who had the same potential risk factor in the control group (those without the disease)
Much like a case-control study, this type of study tracks and compares risk factors shared between the case group and the control group. It can be prospective (following subjects as time moves forward) or retrospective (looking back at events that have already happened to subjects).
Aims to describe the relationship between diseases (or other health-related states) and other factors of interest as they exist in a specified population at a particular time. Cross-sectional studies are a useful way to gather information on important health-related aspects of people's knowledge, attitudes, and practices.
Parallel Design Study:
Subjects are randomized to one of two or more arms, each arm being allocated a different intervention.
Randomized Controlled Study:
Subjects are randomly assigned, using a computer or matrix, into the control group or the investigational group. The control group receives the typically used or approved treatment; the investigational group receives the treatment or intervention being studied.
Blind and Double-Blind Studies:
When a study is “blinded,” subjects do not know which group they are randomly assigned to. A double-blind study means that neither the study personnel who interact with subjects, nor the subjects know who is assigned to which group. These study designs are meant to eliminate possible bias by the subject or caregiver toward or against a therapy or a placebo.
When a study is “open-label”, both the researchers and subjects know which intervention/treatment is being administered. An open-label study may be unavoidable under some circumstances, such as comparing the effectiveness of a medication to intensive physical therapy sessions.
Step 5: Collect study data
Procedures for collecting and managing the data will need to be determined. Generally the following information is described:
- How will data be collected and entered into the database.
- Who will have access to the data and how data will be maintained in a secure manner.
- Procedures for maintaining subject confidentiality.
- How long the information will be kept and plans for destroying the data.
Step 6: Analyze data and draw conclusions
After gathering of data, the data is analyzed and specific conclusions are drawn. Appropriate statistical methods are used to analyze and present the results.