Hi DCR Students - Thanks again for participating in our online Zoom session with Deborah Grady yesterday.
I've already had some feedback, delivered by the students for the small group I led yesterday, and indirectly through the other DCR faculty, but I'm happy to take more feedback.
Here is a Google Doc where you can just type in your comments on what you liked and didn't like, and suggestions for future sessions:
https://docs.google.com/document/d/1M4OdajpLGrq0yuleikS-kXtQ0G9Kqgv5HR9UgBtz_CA/edit
I've also had questions about my "worst mistake", which I never got to say because the Zoom session ended...here's the story:
I was doing a cross-sectional analysis of the CARDIA Cohort Study to estimate the association between telomerase activity in stored mononuclear cells and coronary calcium in CARDIA participants without clinical cardiovascular disease. We collected the precious samples from CARDIA and sent them to the lab for the special telomerase activity measurement. The measurement was a new, difficult assay that was not commercialized but run by a research lab that developed the measurement. I unfortunately sent the specimens SORTED BY some of the variables I was most interested in, instead of in random order; it turned out that there appeared to be some measurement drift in the assay over time, and we could never be sure, after that, if the associations we found were due to the drift and the sorting, or due to a physiologic (potentially causal) link. So the study ended up being basically useless. The morals of the story are: 1) If you are doing a study that depends heavily on a special measurement, take LOTS of precautions to make the measurement as good as you can make it; and 2) Randomization and blinding of specimens is good practice, when possible.
Thanks for listening! I'll send out another post later in the week when the next week's lecture content is up and ready to view.