Getting Your Hands Dirty With the Data to Cross the Finish Line and Grab the Flag
“For me, an analytic plan is an overview of what is needed. You don’t know what it will be like until you get your hands dirty in the data.”
An analytical plan is a type of research methodology that often includes the use of statistics, data analytics, and mathematical modeling.
The goal of an analytical plan is to provide insight into the problem being studied. Often, this insight will be used to develop strategies for improving performance or addressing problems. Data analytics are used to analyze past data to forecast what might happen in the future.
A senior epidemiology programmer’s work always starts with an analytical plan.
The Components of a Quality Analytical Plan
It is important to form an analytic plan before starting any research or experimentation. The analytical plan includes:
- The purpose and objectives, clearly outlining the research goals
- Clear and concise research questions
- The data that is to be used
- The time period in which the epidemiologist is interested
- Where to start and how to start
The analytical plan may also include the inclusion and exclusion criteria, variables to be used in the main analysis, and the statistical methods and software to be used.
You Usually Have to Ask for More
Here, again, is where it pays for a senior programmer in epidemiology to understand both the epidemiology and the biostatistics sides. It’s crucial to know how to respond to an epidemiologist and ask for more.
You see, the epidemiologist wants answers based on the clinical side of the coin. The programmer talks in technical language:
- The epidemiologist needs a single-sentence answer
- The programmer will develop thousands of lines of code to give the epidemiologist that answer
In effect, the programmer acts as a bridge between the epidemiologist and the mass of statistics available for analysis.
To do this successfully, the programmer must be able to push back and ask for specifics.
Nghi gives this example:
“The research question asked may be something like, ‘I want the incidence of pediatric patients with psoriasis’.
For a programmer, that’s not enough. We need to get specific, and there are a lot of variables in the mix. So, I might say, ‘That’s great. I can do that. But can you define ‘incidence’?
“You see, I know that incidence usually means new cases. But from a programming standpoint, you must define what ‘new’ means. What are the complexities with which you are working? What timeframe constitutes ‘new’? How is a case of psoriasis defined? And so on.”
It is crucial to know specifically what is needed so that there is no ambiguity.
Communication and Specialty in Epidemiology and Biostatistics in Action
Once more, Nghi’s experience in both epidemiology and biostatistics has come into play. These skills allow her to understand what the epidemiologist wants, and what questions need to be asked to provide the specifics needed to extract, mine, and analyze the biostats and deliver the specific conclusion needed (that single-sentence answer).
Her skills in communication enable her to ask the technical questions in such a way that they do not confuse the epidemiologist.
A senior epidemiology programmer like Nghi can provide the leadership required to direct a project effectively.
“I view myself as a leader in my project. Pushing all forward, to cross the finish line and grab that flag,” she tells us.
“Being able to help others understand their own questions, and cater accordingly. I find joy in helping others achieve their goals, being their support in any way I can, what I call, ‘leading from behind’. Not someone in front pulling the team, but pushing others forward to achieve their goals.”
Do you enjoy delving into the data, getting your hands dirty, and helping the team solve the research question?
Why not send us a message and let us know? We’d love to help you progress your career as a senior programmer in biostatistics.