Statistic Formulas
(6a) find the percentile for a given amount ex-find the percentile for a diastolic pressure of 78
# of values less than x, divided by / total amount, then multiply by 100
What requirements make a probability disribution
...
"0r more"
1-binomial cdf(n,p,r-1)
"atleast "
1-binomialcdf(n,p,r-1)
(12) Construct a conf interval for proportion
1propzint(enter info)
"atmost"
Binomcdf(n,p,r)
Ratio
Distances, weights, measurements, salaries, time
(11c)Find p#
Invnorm(#, mean, sd)
(AND) lets you know to
Multiply
(11b) Normal distribution -greater than or less than
Normalcdf(#, 9999, m,sd) for greater and Normalcdf(-9999,#, m, sd) for less than
(11d) normal distributed ,given an amount , ex-if 40 different applicants are randomly selected, find the probability that their Mean is above 215
Normalcdf(215,9999,m,sd/square root 40)...above means greater than 215, so 215,9999 then mean, and sd given then divide and square root and #.
Odds against
P(-A)/P (A) find the compliment of the probability, and divide by the probability ...cancel denominators
Multiplication rule, with replacement . Ex a batch consist of 15 defective capacitors and 85 good,If six capacitors are chosen with replacement what is the probability that all are Good?
P(a)* P(a)* P(a).*P(a)*P(a)*P(a).....85/100 , 6times... because 85 is the amount of good and 100 is total ..it is with replacement numbers stay the same.
(Or) lets you know to
add
How to find the mean
add all the data then Divide by the number of data
how you construct a cumulative freq. dist
add the frequency to each frequency below the first frequency.
how to find the mean of the data summarized in a frequency distribution
add your class then divide by two then multiple answer by number of frequency ...repeat steps for all classes then add your data and divide by the frequency sum
complement of "none"
at least one
Ordinal
can be placed in order but order does not matter. Ranks, grades, soft-hard, consumer ratings poor,good
How to construct a pie chart
class frequency / sum of all example class frequency amound divided by total amount
Cluster
divide the population into sections then randomly select some of the sections and then choose all the members from those selected
(9c)addition rule a+b-ab, ex getting a republican or a man..or means add
find a (getting a republican) the find B (getting a man) then subtract what they have in common(republican man) a+b-ab
find the percentile of a data point
how many less than data point /divided by total numbers to choose from then multiple by 100
Range
max-min
range rule of thumb
mean +2(sd)...Mean -2(sd)
(5) Box plot
min, q1 med q3 max
complement of "at least one"
none
(11a) Normal distribution -between given mean and stan dev find the probability of ratings between 170,120.
normalcdf(#,#,mean,stand dev)
(8a) Multiplication without replacement. Ex- a batch consist of 15 def and 85 good capacitors, if two capacitors are selected without replacement what is the probability that all are Good
number divided by /total , multiplied by *, 1-number divided by / 1-total ex - 85/100 * 84/99
(9a)find the probability
number of xdivided by total available x/n
standard deviation
number x minus mean then squared
addition rule
p(a) + P(b)-(ab)
conditional probability
p(b?a)= P(Aand B)/ P(a)....GIVEN IS ALWAYS A
(3b)Median
put info in order if odd number then you pick the middle..if even you take two middle..add both and divide by two
(9b)find the probability using the addition rule (or means add) example probability of getting republicans or a independent
republicans/total sum of all, plus independent/total sum
convenience sampling
results that are easy to get
(9d)additional rule using "given"..ex-getting a democrat "given"that person is a male
rule find given amout of a first then find b from a . ex-given person is a male(total male amount from all , then find democrat from a.
Systemic sampling
select a starting point and then select every kth one such as every 50th
standard deviation
squareroot(n*p*q)
Variance
standard deviation 2....sd squared
stratified
subdivide the population into atleast two different subgroups then draw samples from each group
(13) construct a conf interval, mean
t-interval
(6b) How to find P# ( # th percentile) ex-Find the diastolic pressure at the 40th percentile
use formula P/100*n ..then if whole number count that many then add the next number and divide by two. if not whole count the amount and round up 1
(7c) determine if an amount is unusual . ex-would a diastolic pressure of 96 be considered unusual for a woman?
use the range rule of thumb- Mean-2(sd) and Mean+2(sd)
different
usually means without replacement
mode
value that occures more than once
(7a) how to find the zscore
x-mean/sd = number minus mean Divided by stand dev
(7b) given a z score find x . Ex. Find a womans diastolic pressure that would give a zscore of -1.5
x= mean, -zscore(stand dev) (multiply zscore by standard deviation then subtract the mean to get x)
Nominal
you can categorize but can not be placed in order eye color, hair color, Types of movies,political party preference
