)

M chon
I'
pop mean
: =

+
basics Ax s --
...

: - # of terS
add all numbers ->
N ->
assigning probabilities :

-
mean :
sample mean :

and then divide by # of numbers classical :

expirement &
median order numbers lowest -
:

highest MIDDLE #= med lan ,

n
outcome - P(E)
=
re/W-# of outcomes
#

I
most frequent
mode ↓ possible of
:

outcomes

/sample
#

[(yY
2
variance : =

range- highest #- lowest relative

frequency
·

n e e
: #of times an event
n occurred
Ch02 1
-

guessing
- e
subjective
:

class midpoint =

arg of two endpoints total # of opportunities

of class
mamannan . for event to occur
relative frequency Any
I
:

Fancy
-
total #items in dataset
cumulative total frequency
total o f frequency distribute i s expirement process that produces outcomes
equency
=

-
· event
=
outcome of an explrement

I
- can't be broken down
law of addition RY
aggregate
elementary -
X &

I
population =
M Ch-03 broken down
:

I can be
sample
=
I measures of variability describe
sample space
=
COMPLETE roster of all
levels of data : the spread/dispersion of data
elementary events for an expirement

-rato absolute ener mean absolute deviation :

multiplication as X AND Y
see ->
independent
(v) combo of events (X &RY)
E. xMome
union
=

happen
oraialerarglordering
in e must
3 . MAD =

u nominal
categorize R
④ elements
sets
of squared intersection (f) common both

-
=
to

dat a ! · population variance larg

31g deviations from
variety
- x)2
means ⑭④ conditional
probability :
i
- -

occurrence of event

actexcursive
one
Velocity coefficent mutually
:
prevents occurrence a ene
veracity standard deviation baye's rule :
·
(No intersection) &P(X(y) =
0
-
volume
-o
=

/Ex
now to create a FD
n complementary
A
P(A) =

1 -
P(A) & YA(B) P(B(A) X P(A)/P(B)

I
events A +
Not a - =
:

able : ,
sample size
↓
->

empirical rule : replacement n Pop

~
1 find range size
:

.
sampling
E
w

# of classes -
determine N!
sampling wo replacement
:
2
Crangel
.

Width * I 10 - 68 %
3 Find class
-

classes) TN-n) !
.

Itzocass
n
for
boundaries
.

n .
find
7 %
each class
* =
30599
-

set
frequencies t
5-count
- We

I
y matrix :
up table
-

I
↓

-
discrete distribution

himdismbuto
probability of

+
normal distribution

It
Chos is always I : A
-

I
vanable mean and sk for i X
random variable

A
continuous
:

that contains the outcomes

of
poisson distribution :
around the
a chance expirement symmetric &X

M
=
X ·
upper t lower bound Es
mean

i
(a) (b) of
parameters (M
and X
discrete random variable
I
:

& =
2 -
values that a re
produce

auda aene
numbers · rules stuff :
negative whole

)
non
+

b - a
continuous random variables
:

unassumed values Baye's rule :

no gaps or

measured , not counted) ab

s
M
=

# of

discrete distributions
tota
pos sible valued 2 zbyz =

2 1 .
list / calculate regular/given
5= P(Xi) probabilities
word
"

P(z( )
x;
mean
:
M =
G(X) =

1
↓
·8
= 2 p(Xxa) conditional
weighted arg ! calculate regular
· =

possible
KeraP
↳ a
2
2
-

value
variance of DD
:
-
using probability to solve for mean
,
SD ,
or probabilities
distribution
exponential
joint probabilites
2N((Xi-m)2] calculate
-
·
PIX In a normal distribution : 3
an X-value
·

often concerned wtime

reverse lookup of I table 1 formula
use get to

events
.

deviation (DD) between I consecutive

:
probabilities
standard
I always & X
=
0
1 find closest probability In E table revised

/2,((Xi P(xi) apex Is decimal place

0 =
-
M) 2) -
2 look up
.
the whole # and
X 1st
i 1
positive parameter of the E
value the column
=

· single in

add

I
and
-

s look up tre second decimal place

binomial distribution : probability density fin :

3 I
'n' identical trials Ak & >0
involves - sample
size f(x) =
Re-NY for x = 0 Where
'sucess' outcome ,
fail' or

· each Inal is independent

is the only parameter
· x =

arg # of occurrences/time interval

mean
M
·
=
n .
B
· 1/X= arg time between 2 consecutive
standard
deviation -
-

Vn
·

o =

p q
customers
P
. .
Q =
1 -
=

Prob of
fallure
·

·
P(Xx 8) =
calculate x =
1 - X 7
Meant Standard Deviation
=

and then add

M Y/x ; 0
=

1/x
poisson distribution
=

events are Independent computing probabilities :

events are discrete

xx
P(x x)
-

>
could be # of discrete
=

events in a of time *
given period ↑ p(X(x) = 1 -

·
X = # of events PER time period xx x2
P(X X(x2)
- -

, =

e -
e
·
can also be
given intervals ,
distance
or area
e
X
pX
n -

x must be constant formula :

nCy q
·
· .

poisson formula :

treal
! probability of
·

=n
p=

et
-
-
-

P(x) (n-x ! Success/thal

x !

Hofthals . e
H -

interval sex-times x =
# of successes desired f or fails
in 5
 I
Ch07
-

from samples +
sampling data gathered
:

made as part of the

conclusions are

inferential stats process

sensus gathering info from EVERY member of

:

the population ->

eliminates chance of misrepresen
tative sample .

random sampling
:

every pop .
unit has the same

probability of
being selected

nourandom sampling :

generally not an appropriate

technique for gathering data for stats

sampling distribution of X :

·
X
=
sample mean

normal If 30
= :
n
central limit theorem : ->

standard /Ze
mean X M
My
of = =

of mean) =
LaKa standard
error the op
=

SD
Fr

n
=
sample size

If I is normal ,
standardize to find probability
answer on z table
and look up

· sample proportion -> P=

x # of items In that have
sample
=
a

the characterish

n
= sample size

p(wq
1 -

P)
proportion is
=

If population
then
is normal If npLS and ng>5
of B M P
= =

· mean

SD (also called standard error of the

proportion) op /
=
=