kurtosis not kertosis excel ult 4

determine significance

dividealpha by 2 if two tail

zscore crit times var then divide by sqrt n samp 

zscore method

numerator xbar minus mu

denom var divide by sqrt n samp

ttest find sd

xbar minus mu divided by

denom sampsd timessqrtn

if pop mesn used its zscore

convert mormal to zscore is x minus mu then divide by se

t.inv function get t random variable

p value need t distr prob not percent so use t.dist

ttest sample size is small, and the population standard deviation is not known

ztest opposite. differenve is if samp size less than 30

test pop equal var two tail ftest

test alpha with la veen

large f ratio more dif in var

always large var minus small var for fmax

degree freedom samp size minus one

f dist needs two groups only same samp size

always skew right 

to find crit value

select ftable alpha then look dof column group size rows. if larger reject null

var not equal confidence interval not one

f.test function

kertosis skewness between minus one and one means good data

two samp test for expected dif

xbar dif minus mu dif then divide by denominator  their var sq divide by samp size added to each other. pooled se denom

when equal var dof is sum sampsizr minus 2

same sampsize use two samp ztest

ftest test var equal

ttest small normal pop any var

normal means skew between 1 and plus one 

kurt close 0

ttest check mean sig dif

pvalue two tail double one tail 

pvalue checks chance if teo means same you see discrepency same or larger

alpha level percent is big 

matched pairs before after 

or placebo med match subjects same age gender from each group then flip coin determine who gets placebo

this negates blocking variables like ability age weight

And we would divide by total P7.

But now when we copy that down to get it for males,

we want the 7 to change to an 8.

When we copy this across, we don't want the P to change.

So I'll dollar sign that, okay.

same colimn dif row

chi square between sub categories associsted or dependent or dif variance affect categories compared

expected val for see what not indep will looks like

 Expected values are calculated by multiplying the marginal row and

column totals and dividing by the overall total.

Chi square is then the sum of the expected values subtracted from

the observed value squared,

divided by the expected value with the degrees of freedom,

that is the product of the number of rows minus 1.

And the number of columns minus 1.

_____________  basically we

square the difference between the observed and

expected value and divide by the expected value like this.

Finally, we sum these values, obtain our chi square value 

if less than crit value lookup dpf table no difference

eij means expected under independent

female prob times blue eyes prob times total

observe minus exp then sq then div by exp

chisq.dist.rt is pval 

chisq.test need observe and expect value only dont need dof or chi sq total

_______________

multiple match

multiple match

aggregate ignore errors

divide logic tedt by itself since false is zerogive error

multiply row function so true can be ranked

https://youtu.be/fDB1Ktyhp3Y

delimiter split answer with symbol

textjoin need cse remember to keep in memory so much to search

filter isnumber search

hides missing boxes 

click first result to reference

to get entire spill arrays range need hashtag in data analysis box

rrmovr error alert to allow incomplete searches not exact when you click dropdown

to incl corresponding company filter select company column as array and include customer column then type = and click search box

turn data in table to automatically update