Everyday I'm Shufflin pt 1 - A statistical analysis of studying with flash cards

I decided to change my series of posts chronicling the things that consume the day and night of a med student from "About That Life" to "Everyday I'm Shuffling." 

This decision was important enough to me that I felt I had to tell you about it. I spend a large part of my days thinking about catchy titles to posts. About That Life sounded cool, but I'm not sure how much sense it makes. Everyday I'm Shuffling sounds cool and makes a ton of sense - med school feels like you do the same thing over and over day after day. Pre-read, zone out during lectures, make coffee, flash cards, take self-assessments, make more coffee. It never ends. 

Appropriately, the first Everyday I'm Shuffling is a small statistical study on shuffling flash cards. Enjoy.

--

#1

I was going through my online flash cards to see how much kidney physiology I remembered. Turns out, not a lot. sidenote - ever since I realized med school was just another hoop before the powers-that-be will let us start learning real medicine, I've been trying to figure out how to game med school. So while I was studying, a question came to me - what's the minimum number of flash cards I could go through to know my score for the whole stack. In other words, what's the smallest significant sample size? 

So I sent this late night email to my mathematically inclined friend:

I have a virtual stack of 461 flash cards on digestive physiology. I can randomize the order I get the cards in. Assuming 461 cards is a big enough sample size to give me a statistically significant percentage correct, how many cards is the smallest significant sample size? ie - how many cards to get rid of the noise? 

My friend's guess was 30-40. Another friend guessed 50-120. Another friend used an online calculator to say with 95% confidence you'd need 210 cards. Wow, if it's really 210 cards I might as well just do the whole pile.

Whatever this magic number is, it's important because I can go through that many cards in any of my stacks - biochemistry, embryology, anatomy - and figure out what I should spend my time studying instead of wasting my time reviewing facts that I already have down.

As a good scientist, I put this hypothesis to the test. Here were the results from round one, Ken vs Gastrointestinal physiology:

30% @ question 10
36% @ q25 
36% @ q300 (WOW talk about consistency)
42% @ q461

My friend was weary of n = 1, so I put in a round two, Ken vs biochemistry:

90% @ q10 (great!)
84% @ q25 (getting sadder)
78% @ q50 (still getting sadder)
75% @ q391 (fuck this)

I did a third trial for good measure, Ken vs anatomy: 

70% @ q10
76% @ q25
85% @ q100
83% @ q410

Based on this data, I'd say somewhere between 50-100 flashcards, although the data from that first experiment is enticing. Anyways, I should probably go back to studying..

sidenote - this idea just occurred to me, anyone up for trying this same experiment on my (or your) flashcard pile?

See you on the other side,

from ken

taking requests for future blog posts. please comment or email.