Question

Hypothesis Testing!!!
When you find the mean and standard deviation, do you use the hypothesis percentage for p?

Answer 1

we assume the sample taken is one that doesnt happen by chance, so we assess the hypothesis based off of comparing it to the sample we took

Answer 2

how well does the hypot fit within our trial

Answer 3

So when do I use the data I collected?

Answer 4

percentage for 'p' is vague since they variable p is used for so many things in stats

Answer 5

Here's what I have so far.
Hypothesis: 60% of people surveyed will have an iPhone.
Significance Level: .05
Data: 28 of 53 people have an iPhone.
Let x=people with an iPhone

Answer 6

p-value? or proportion?

Answer 7

the p value. Like when finding the mean x=np

Answer 8

thats supposed to be x bar my bad

Answer 9

p-value is not proportion value :)

p is from the data you collected

p-value is something differnt, its the probability of the test statistic

Answer 10

p=28/53

Answer 11

So how would I go about doing this?

Answer 12

Sorry, it's for a project and we only briefly talked about this. Plus my teacher rushes through things >.<

Answer 13

how far is .60 from 28/53?

Answer 14

.07

Answer 15

good, .0717

now we divide that by the standard error to determine our test statistic

Answer 16

what's the standard error?

Answer 17

standard error is the standard deviation of the distribution of sample means.

Answer 18

oh so its the square root of p and 1-p right?

Answer 19

in other words, we have obtained one of many different sample values, this specific cample we assume to not have been obtained by chance and that it fits within ...

yeah, but in this case:

sqrt(28(53-28)/53^3)

Answer 20

why did u do 53^3????

Answer 21

\[SE=\sqrt{\frac{pq}{n}}=\sqrt{\frac{a}{n}\frac{n-1}{n}\frac{1}{n}}=\sqrt{\frac{a(n-a)}{n^3}}\]

its just another way to math the data with the values we have instead of converting to decimals and dividing by n

Answer 22

http://www.boxcarchallenge.com/BookPages/page_764.pdf this is how my book does it

Answer 23

n-a not n-1 in the middle there

Answer 24

different authors take different approaches. instead of comparing proportions, they have used a different approach in yours. or maybe your on the wrong page?

Answer 25

... the site is getting very laggy for me at the moment, hard to respond ....

Answer 26

It's the firefighter one right?

Answer 27

the hypothesis would expect 60% of 53, 53(.6) = 31.8 people in the survey right?

Answer 28

Yeah

Answer 29

now we compare how 31.8 fits into the sample in order to compare it with they hypot

Answer 30

Ohh so I use .6 as the probability?

Answer 31

hypot is claiming .60 (60%) of the people do something, the sample has 28 out of 53, the hypot is claiming 31.8 out of 53

Answer 32

okay gotcha

Answer 33

our standard deviation, assuming your book page is the correct one :)

sigma = sqrt(53(.6)(.4))

Answer 34

so, x = the difference of 3.8, divided by sigma; to look up in a table

Answer 35

ok one min

Answer 36

I got .1446.

Answer 37

i dont have your table, so im getting a more exact value of .1433

can you link the table?

Answer 38

yeah ill just take a pic hang on

Answer 39

your book gives us:
\[x=\frac{h-np}{\sqrt{npq}}\]
we can work this into what i was starting with as:
\[x=\frac{\frac{h-np}n}{\frac{\sqrt{npq}}{n}}\]
\[x=\frac{\frac{h}{n}-p}{\sqrt{\frac{npq}{n^2}}}\]
\[x=\frac{H_p-p}{\sqrt{\frac{pq}{n}}}\]

Answer 40

x = 1.06547..

notice since this has a 5 in the round spot that we are really more preciselyin the middle of 1446 and 1423, whose average is closer to the exact i had gotten of 1433

other than that, table of 1446 is fine

Answer 41

its up to you if you want to keep it at 1.06 or round to 1.07, or if you want to average the fields

Answer 42

the question remians as to how we use this now to test the claim :)

Answer 43

Sorry i was afk

Answer 44

hypot is that 60%, no more or less, do something.

what this means is that we have an Ha (alpha hypothesis) that is split between 2 tails