How much statistics is required for medical research?

[UrmiSinhaRay]

Will learning the Mann -Whitney test, ROC curve , Bland-Altman plot, Deming regression and Passing Bablok regression, control charts necessary for advanced research in medicine?
Basically how much stats I need if I am wanting go into research medicine.
I am a first Year MBBS student and I am looking for opportunities in medical research.
Also suggest books with these chapters.

---

Members don't see this ad.

 

[PapaGuava]

@QofQuimica may have some insight for your question!

 

[QofQuimica]

I'm afraid I really can't answer that question without knowing more about what kind of research you're interested in doing. I would suggest thinking about this from the other direction: instead of trying to learn everything possible about statistics so that you can be ready for some kind of unknown research gig in the future, first figure out what kind of research you want to do, and then you will know what kind of statistical training you need to have in order to do it. You should also consider that most research is done in teams, and you will likely have a professional statistician (i.e., someone with a PhD in this stuff) working with you. Meaning, if you're a physician working with a statistician on a project, it likely won't be you who's crunching the numbers!

 

[Opticalfuture]

Yeah I'm trying to become a medical scientist in the future. How much math do I need?

 

[Billyrmd]

I think I just took "Basic Practice of Statstics" and "Biostatistics" which were useful and applicable courses.

 

[deleted1006988]

I worked in medical research for several years before med school and come from an engineering background so here are my 2 cents:
It is helpful to learn how to use the basic statistical tests and understand how they work overall (Mann-Whitney was one of them). Similarly know about prediction/classification models and how to evaluate them (ROC, PR, etc). In practice you'll pick these things up while reading papers and looking at your specific hypothesis and data structures. Don't worry too much about the underlying statistical theory - intro biostats was far more helpful clinically than grad-level stats classes.

My main reference book has been "Practical Data Statistics for Data Scientists" from the O'Reilly series, but there's probably better ones out there more focused for medical.

tl;dr Know generally how to interpret papers in your field and how to choose methods for your research. Don't focus on advanced stats theory.

 

[seanbancroftt]

I think I just took "Basic Practice of Statstics" and "Biostatistics" which were useful and applicable courses.

Did you like these courses?