Need your input on epidemiology learning modules/content

Hi all,

As part of the Data Modality & Methodology Task force, we would like to develop short, digestible yet informative content (e.g. 15 minutes videos or vignettes) on key epidemiologic concepts relevant to Parkinson’s research. These will be designed for both researchers and engaged community members who want to better understand study design and data interpretation.

We’re starting to outline topics, including:

  • Overview of study designs (cross-sectional, cohort, case-control, case-cohort, ecological, etc.)

  • Understanding confounding and mediation (and collider bias)

  • Selection bias (but might be challenging to distill into a concise product)

  • How to interpret an odds ratio, hazard ratio, incidence, prevalence, etc

  • Basics of survival analysis — when to use Kaplan-Meier, CPH, etc (though this might be more of a best practices document)

  • Benefits of matching

Before we finalize the list, we’d love to hear your ideas, and be expansive.

While some might not be feasible for a brief video or a vignette, we can do our best to track down resources that might shed insights. Looking forward to your suggestions.

6 Likes

@fbbriggs ,

That’s a great initiative. Speaking from personal experience and interest, I think short videos explaining concepts like odds ratios and hazard ratios would be very helpful. It would also be valuable to include demonstrations of survival analysis showing not only how to perform it, but also which datasets or data structures are suitable for these methods.

5 Likes

Thanks for this, @fbbriggs!

Tagging community members with stated interest in epidemiology to see if you all have any comments / suggestions / feedback!

@buddhi @SyedMuzammilShah @CamillaKilbane @paularp @Ranam @cdwhelan @rochet071369 @Sherpi @namburin @Ollena5 @mengxi.yang

2 Likes

Hi @fbbriggs, that’s a great initiative. In addition to the outlined topics, I find the following content useful: Sampling, Errors, when to use various statistical tests (for instance paired t-test, ANOVA, Chi-square test etc.,), Markov’s model.

3 Likes

@fbbriggs great initiative

My comment will be to suggest that video content should be at most 10 mins even if it means more videos,that will be better.

I find the shorter and more concise a video is, the more engaged the audience will be.

3 Likes