Question

hypothesis testing?? Someone help me

Answer 1

might need some more information ....

Answer 2

hold on

Answer 3

at my age, time is a luxury ...

Answer 4

LolxD

Answer 5

this is more readable ... the chinese method is not the standard

Answer 6

Okay...

Answer 7

start with stating your null and alternate hypothesis ... whats your efforts?

Answer 8

in order to run tests, we have to define what it is we are testing, and what constitutes acceptance/rejection.

Answer 9

*

Answer 10

Null hypothesis: M=U

Answer 11

Sample mean= population mean

Answer 12

is the mean of the last 9 games higher than the mean of the season is our question

lets frame the null with an equality: Ho: mean <= 83 games won

the alternative hypothesis is simply the rest of it: Ha: mean > 83

Answer 13

im not sportsy so my interpretation of the question may be faulty lol ... but this is the basis of our testing.

Answer 14

agreed so far?

Answer 15

yes

Answer 16

now, in order to test this stuff, we need to know the means ... what is our sample mean?

Answer 17

83

Answer 18

our sample mean is not 83 sum of the data, divided by how many datapoints there are.

its close to 83 ... but i did not get 83 for the sample mean

Answer 19

84

Answer 20

9 seasons, # of wins:

89, 95, 95, 96, 86, 9, 98, 95, 93

mean is 84, good

now tell me how you would propose we use this information, ill correct you if need be.

Answer 21

we would use the z test

Answer 22

ok, do you know why we would use z and not t? or should we use t instead?

Answer 23

Because we use the z test when we have the population standard deviation

Answer 24

sounds fair enough :)

and we want two z scores, a critical value, and the value of the test statistic.

the critical value establishes a boundary for the rejection/acceptance regions; and the actual value of the test statistic tells us which side of the boundary we are on.

whats our critical value? (itll be associated with the significance level)

Answer 25

do you work this with p-value or critical value? either way is sufficient

Answer 26

mostly the p-value

Answer 27

ok, then with p-value we are comparing the size of alpha with the probability value of the test statistic.

i always found it simpler to compare the z scores and not their associated probabilities. why look up what the z relates to when its simple enough to compare zs eh

Answer 28

critical z score would be 1.645

if the z for the test stat is higher then this then we know we are "greater than" and in our rejection region.

calculate the test stat for me. how would you work it? (formula)

Answer 29

wait don't we have to do the z test first before doing the p value

Answer 30

we have to find the related z score yes

Answer 31

since alpha = .05 ... and is right tailed (Ha > 83)

our left tail area is .9500, which relates to a critical z score of 1.645 ... but this information is pointless if comparing the p-value.

we still need the test stats z value

Answer 32

i don't know how to do the z test. I always mix the numbers up

Answer 33

z = (sample mean) - (hypot mean)
--------------------------
(pop sd) / sqrt(sample size)

Answer 34

84-83 = 1

sqrt(9)/16 = 3/16

Answer 35

1/(a/b) = b/a if that confused you

Answer 36

so N=9

Answer 37

yes, we have a sample of the last 9 seasons that we took the mean of

Answer 38

what about the 110? Where does that fall into

Answer 39

extra information that does noting for us id assume

Answer 40

the population size is 110 ... has nothing to do with the hypothesis testing. at least not in any of the courses i took.

Answer 41

okay

Answer 42

as an answer i got 0.188

Answer 43

me too

if we were comparing z scores

rejection region
------------|------------------------->
0.188 1.645


we are clearly not in the rejection region; and your p-value will be bigger then .05

Answer 44

We retain the null hypothesis

Answer 45

we fail to reject it is what i always refered to it as. but yeah

Answer 46

P(z>3/16) = .4256

since alpha = .0500

you see how the p-value is working out

Answer 47

how did you got the p-value? And this question is a Type 2 error

Answer 48

P(z>3/16) of course

Answer 49

z = 3/16 = 0.1875

normalCDF(0.1875, 9999, 0, 1) is one method

or looking up a table to find row -0.1 and col 0.08 if its left tailed

etc etc ....

Answer 50

how would you find the area related to the right side of z=0.1875 ?