Thread replies: 13
Thread images: 6
Thread images: 6
Anonymous
Sources for learning statistics 2017-07-11 10:04:47 Post No. 9029809
[Report] Image search: [Google]
Sources for learning statistics 2017-07-11 10:04:47 Post No. 9029809
[Report] Image search: [Google]
File: lol_stats.jpg (243KB, 1920x1080px) Image search:
[Google]
243KB, 1920x1080px
Hello Anons, I want to better understand the data published in medical journals and the thing that's constantly impeding me is my ability to understand the stats they use.
None of the techniques seem that complicated (It seems to be mostly different types of regression analysis with weird set-ups so they can hack their p-value, something I see a lot is where they have a study where they get nothing significant, then do sub-group analysis to claim they've found something and I'm not really sure how to evaluate this, sometimes there's more complicated stuff like PCA), however, most of the books I've come across that explicitly deal with medical stats are either exorbitantly expensive or look mind-numbingly dull.
tl;dr: Are there any Doctors or people who work in healthcare reasearch that can suggest a good intro to medical stats book for me?
>>
>>9029809
can you give some specific examples? like what tests do you mean? what tests do they use when they don't find anything significant?
>>
>>9029809
Literally any intro stats book is fine. Barely anybody knows or interprets statistics correctly, so you don't need anything more complex 99% of the time.
Save yourself time and read Nassim Taleb first
>>
File: 51Fi8hXVptL._SX324_BO1,204,203,200_.jpg (35KB, 326x499px) Image search:
[Google]
35KB, 326x499px
>>9029809
start by reading this book
>>
File: oj_dingo.jpg (48KB, 620x866px) Image search:
[Google]
48KB, 620x866px
>>9030056
I'm sure they would be fine but all the ones I've looked at are either boring or too expensive. Could you recommend one you like/used to learn. Ideally I was wondering if there's a more engaging source that I haven't found yet. I said book, but I'd be up for an online course or maybe even video/podcast series if it was engaging and detailed.
>>9030078
Thanks for the recommendation but this doesn't really look like what I'm asking for. I actually have a PhD in Plant Science and I think I probably know a little more math than the average biologist. My problem is that I learned all the stats I know in a pretty ad hoc way (if a paper had a technique I didn't understand, I looked it up) so what I'm really hoping to do is to consolidate the various disparate stats I know into a fuller picture.
>>
>>9029809
she thic
>>
>>9030897
True senpai
>>
>>
File: 5185%2B70D8UL._SX348_BO1,204,203,200_.jpg (35KB, 350x499px) Image search:
[Google]
35KB, 350x499px
>>9030271
I was fucking with you. That's not a math book.
I highly recommend this text however, it teaches a lot of statistics from a Bayesian perspective which is incredibly useful. Furthermore, it gives examples in R and Stan (a MCMC language) and is written by an anthropologist - it's not extremely rigorous and very easy to read. Highly highly recommend it.
>>
File: 168872.jpg (58KB, 317x400px) Image search:
[Google]
58KB, 317x400px
>>
File: free_for_chubbies.jpg (76KB, 620x826px) Image search:
[Google]
76KB, 620x826px
>>9031623
Is this actually good? It looks like a meme.
>>
>>9031623
>the cartoon guide
What a meme. Use the manga guide instead. If the japanese were able to get nuked twice and survive then they gotta be good at something.
Thread posts: 13
Thread images: 6
Thread images: 6