To conduct good research, you need a good research protocol. The protocol is a written document that explains what you're doing, how you're doing it, what you hope to achieve and how you'll produce accurate results
A research protocol or proposal is a document describing the objectives, design, methodology, statistical considerations and organization of a research project. The research protocol also covers how you'll ensure the integrity of your data.
Whatever your research goals, you need a plan to carry out the project successfully. The research protocol is that plan, detailing the guidelines and explaining every part of your project.
You may want to design your own research protocol for your small business to ensure quality research done the way you want it to be. Though you may have your own specific procedures and styles, the basic elements of a protocol, are consistent:
- An explanation of why the question you're researching is important.
- Background information about the available research on this topic and its relevance to your own work. For example, if this is the first human trial, you should discuss any previous animal research.
- What are your goals and objectives?
- The study design: Are you seeking a correlation between two factors, making a general survey or testing an experimental drug? How will you set up your control group? Some research just involves questionnaires when you're looking for common environmental factors or life experiences, for instance.
- Your methodology: Write exactly what procedures and systems you'll use to execute your research. What measurements will you take? What lab tests will be done? What time of the day are you performing experiments at? At what intervals are you testing results? It's better to be overly-detailed than not detailed enough, so as to not discredit your results.
- Safety considerations: You must make your volunteers' and researchers' safety a priority. The research protocol should detail how you'll ensure their well-being during research procedures.
- How will you handle the statistics generated by your tests? You need to outline your statistical-analysis methods, include why you chose the sample size and explain how you'll account for missing data or errors.
The more complex the project, the more demanding the protocol-development process. If your proposed research takes place on one site with minimal risk, then the protocol requirements won't be as demanding as a multisite research project that puts the volunteers at some risk.
There are many ugly cases of bias in research. Some projects study men but never consider whether women would react differently. Some researchers have exploited vulnerable populations — the poor, people of color, cognitively impaired individuals, children, convicts — who are less likely to understand the risks or refuse the researchers.
When you select your candidates, keep an eye out for bias. If your research only involves men, is there a justifiable reason? If you draw your test subjects from a vulnerable population, is there a rationale beyond wanting subjects who won't question your directions? Is there a wide mix of ethnic types or a valid reason for not having a mix?
The research protocol should provide clear criteria for your choice of subjects. A reader of your research report should be able to read the criteria and know that the subject population is representative of the whole.
The protocol-development process can look intimidating to a newbie. Fortunately, the institution under which you're working will be able to help. Medical schools, hospitals, research laboratories and independent review boards all offer protocol-writing guidelines and often sample templates that you can download and adapt.
Institutional protocol-writing guidelines cover the format of the document as well as the contents of the research. Documents typically include a title page, the name of whoever's funding the project, the names of the researchers and the locations of their research sites. Following your institution's format faithfully will make your protocol easier to read.