#Chapter 6 statistics
194 messages · Page 1 of 1 (latest)
perfect
so
ASSUMPTIONS:
a fixed number of "trials", where a trial is just like a run of the event
each trial can only either 'succeed' or 'fail'
the probability of success is constant for each trial
the trial are independent (so no trial impacts the other)
so just as an example, the dice roll thing from before about getting a 6 or not getting a 6
yeah I get that
we can consider getting a 6 to be a 'success'
as a way to split it up
ok now
binomials follow a specific probability distribution
im gonna write it out then we can talk thru it
yeah this looks familiar but idk what it means
pretty sure B means binominal
n is the number of trials
p is the probability
I think
yep perfect
and the squiggle just means
"is distributed like"
so the first line basically reads
X is distributed like a binomial random variable, with n trials and probability of success p
ok?
yeah I understand
ok awesome
now if we think logically
if i have n trials
the number of successes i can get are gonna be 0, 1, 2, ..., up to n
as our like possibilities
yeah?
yeah
we cant get like -1 successes, we cant get like n+1 successes , we cant get some weird decimal number of successes
ok so the second line basically reads
so it ahs to be an interger
okk
so the 2nd line reads as
the probability we get exactly r successes
and then that formula
do you know what nCr means?
what is r
r is the number of successes
were asking lik
whats the probability we do all n trials and get r successes
thats what the P(X = r) means
oh
its just like a letter to signify that number of successes
is it mean number of trials C number of success
no
nope
itll be on your calculator somewhere but
yeah it is shift divide
its basically the number of ways to choose r things out of a list of n things
so for example if i have say
ABCDE
and i wanna choose 2 things
well i could choose AB, i could choose AE, i could choose BE
etc
yeah
the choose function tells you how many of those options there are
perfect exactly
ok so to kinda explain why it shows up in our formula
if we think about doing these n trials
with r successes
if i just call S for success and F for fail
and lets just do 5 trials and 2 successes again
just as an example
ok
then you can see that these successes could come in lots of different orders like
SSFFF, or SFFSF, or FFSFS
etc
yeah?
yeah
so the choose function is here because
thats all the different ways we can get 2 successes
we choose 2 successes of the 5 slots we have
and then we fill in the rest with Fs
yup
whats Fs?
fails
oh ok
so we know the probability of success is p just cuz thats what we chose
and because of the like 'not' rule, the probability of failing is gonna be 1-p
because P always means probability of success?
yep
okay yeah
at least thats what i used here
it might be a number normally
but getting since we have r successes
the probability of those bits will be r^p
and then the other n-r trials fail
so they have probability (1-p)^(n-r)
if that makes sense?
i do sorry
yeah I get this
youre right
np
ok awesome so
thats the like frame work
so lets maybe take an example
lets use the dice one again, i roll a dice say 10 times
and i call a success rolling a 6
how would you work out the probability of getting 2 6's?
maybe talk me thru it
sure
I could try to talk you through it but let me ensure I get the calculation part right
one second
ok sure
is this right?
I put it on paint so its clearer maybe
I will now put inmy calculator and tell you the answer
is it 0.2907
could you double check please?
yeah thank god
so now lets maybe extend our ideas a bit
how about getting at most 2 6's
ie 2 or less
whats ideas do we maybe have for that
idk I am especially confused for this because my teacher said the sign flips or something
I dont know sorry
dont worry about the sign so
if its 2 or less successes
what successes are valid?
perfect so
why dont you work out the probability for each
and then since theyre like separate
you can just add them up
does that make sense?
yeah sure
yep perfect
so
heres an exam question
talk me through what your working would look like
I would find that 3/5 is 0.6
so thats the probability of 6 asking for water
and since there are 10 trials
because like 10 people
yep
then
is this it for a
perfect yeah
lets goo
so you could simplify that and then calculate it
yeah
ok how about the other one?