Friday, August 24, 2012

Turn Research Topic to Poster Presentation


I have some experience working with health professions on their research poster presentations. Most of them have not much experience on how the research getting started and how the statistics getting involved. There are some essential steps for present your research in a poster.

1. Research Topic.
This is the key part for the presentation. The research topic can be measuring the disease severity, as well as testing a hypothesis that certain behavior could be associated with increase/decrease the risk of the disease. It could be a question that need to answer. This may required some background knowledge and some literature review. In this process, the health professionals can know what related research has already been done. They can determine the direction for the on going research and modify their topics.

2. Data source.
After set up the research topic, there would be some important factors/variables. The important outcome variable, the most correlated independent variable, and some other related factors that we may need to control for. We will need to identify if we have to collect the data ourselves or we can use some publicly available data source. For example, most of the health behavior or health outcome related research topic can be assessed by the NHANES (The National Health and Nutrition Examination Survey http://www.cdc.gov/nchs/nhanes.htm/) data one line.

3. Statistical Methods.
When both the research topic and data source are determined, you can start to talk to a statistician about the research question. Let him/her know your research topic, your data source, your study design (how do you choose your eligible sample, e.g. age criteria; which are the related variables, e.g. gender, ethnicity, education, etc;), your definition of the variables. In your field, people may have some categorical suggestions. For example, we always defined age into groups for adults as, 18-34, 45-64 and 65+.

4. Statistical Analysis.
You may help the statisticians/biostatisticians to determine the analysis. Usually, the set of analysis can be summarized as Univariate Analysis (the distribution of your dependent/outcome variable, independent variable - the variable you care most, covariates - other variables you would like to adjust for), Bivariate Analysis (the bivariate distribution for each related variable - test if all the variables you identified are actually related to your outcome variable), Multivariate Analysis (the actual model that will help answer your research question).

5. Interpret the Results.
You will have to discuss with the statistician about the results after the analysis. After discuss the results, you have some idea if the interpretation are conflict with what already known in your field. You may help the statisticians to identify whether this is the error in the analysis or the limit of the data source. Some time, this process could be back and forth, until we have the results answering the research question and consistent with current literature. 

6. Present your Findings
There are two ways to present your findings, orally present as a talk (you will need the power point slides) or present your results in a poster. Here I will focus on how to present the research in a poster. You may want to include some figures to illustrate your results, which always quickly orient the audience. You will need the key parts on your poster
  • Title
  • Author/Co-author
  • Abstract
  • Background/Introduction
  • Methods
  • Results
  • Discussion/Conclusion
  • Acknowledgment 
  • Reference

7. Print your Poster
Many free poster presentation pre-designed templates are available for putting together your findings into a fixed size page for large print. For example, 

To summarize, presentation experience is important. More practice, can make a perfect. Reading others' poster presentation and getting feedback from your audience can improve your own skills significantly. 

Good luck on your poster presentation.



No comments:

Post a Comment