Question

Statistics Problem

Answer 1

Subjects with preexisting cardiovascular symptoms who were receiving subitramine, an appetite suppressant, were found to be at increased risk of cardiovascular events while taking the drug. The study included 9836 overweight or obese subjects with preexisting cardiovascular disease and/ or type 2 diabetes. The subjects were randomly assigned to subitramine (4823 subjects) or a placebo (5013 subjects) in a double-blind fashion. The primary outcome measured was the occurrence of any of the following events: nonfatal myocardial infarction or stroke, resuscitation after cardiac arrest, or cardiovascular death. The primary outcome was observed in 570 subjects in the subitramine group and 482 subjects in the placebo group.



Do the data give good reason to think that there is a difference between the proportions of treatment and placebo subjects who experienced the primary outcome?

State hypotheses, find the test statistic and the P-value.



Test statistic (0.001) z =

P-value (0.001) =

Answer 2

for test statistic I got 3.5347

Answer 3

I thought p-value was .107 but its not

Answer 4

@jim_thompson5910 last one for tonight if you got the time please...seems like no one else has a clue =/

Answer 5

I think when it says "Test statistic (0.001) " it is saying to round the test statistic to the nearest thousandth (or 3 decimal places)

Answer 6

it did, but the site is weird, sometimes it rounds, sometimes not so I added the extra number instead of taking a chance, it usually doesn't count it wrong when I do that

Answer 7

but if im one number off then I get it wrong

Answer 8

hmm interesting, ok one sec

Answer 9

ok

Answer 10

I'm getting 3.53467921514759 for the test statistic, so you did that correctly

Answer 11

however you round it, not sure

but you can always bring it up with your teacher and s/he'll know what to do

Answer 12

now the key to this whole test is figuring out what the null and alternate hypotheses are

Answer 13

Yeah, its the bad thing about online homework...

Answer 14

alright...

Answer 15

it says "Do the data give good reason to think that there is a difference between the proportions of treatment and placebo subjects who experienced the primary outcome?"

so it wants to know if p1 = p2 or if p1 =/= p2

Answer 16

so

Null:
H0: p1 = p2

Alternate
H1: p1 =/= p2

Answer 17

this means we have a 2 tailed test

Answer 18

not again... haha

Answer 19

yes revenge of the two tailed test lol

Answer 20

so we use this table

http://homes.cs.washington.edu/~jrl/normal_cdf.pdf

Answer 21

ok, but we don't use the z that we got before, do we?

Answer 22

the table doesn't go high enough...

Answer 23

hmm let me see if I can find a larger table

Answer 24

I tried that before posting, so I figured I had to do something else

Answer 25

but what, I do not know

Answer 26

no you're just finding the area at this point

Answer 27

I guess we should use a calculator

Answer 28

ok I'm using this calculator

http://www.wolframalpha.com/input/?i=normalcdf&a=*C.normalcdf-_*Formula.dflt-&f2=3.5347&x=5&y=8&f=NormalProbabilities.z_3.5347&a=*FVarOpt.1-_***NormalProbabilities.z--.***NormalProbabilities.pr--.**NormalProbabilities.l-.*NormalProbabilities.r---.*--&a=*FVarOpt.2-_**-.***NormalProbabilities.mu--.**NormalProbabilities.sigma---.**NormalProbabilities.z---

Answer 29

and I got

P(Z > 3.5347) = 2.041 * 10^(-4) = 0.0002041

Answer 30

now double this answer to get 2*0.0002041 = 0.0004082

Answer 31

so if you were to round that to 4 places, it would be

p value = 0.0004

Answer 32

I think that's why they rounded earlier to 4 places

if you rounded the p value to 3 places, you would get 0, which is an odd answer (but I guess it's valid somewhat)

Answer 33

hmm ok, I have an example, nearly the exact same as this one, and the p value was .001.... so Im wondering why this one is .0004 if the difference between the numbers of each problem was minimal

Answer 34

show me the example

Answer 35

in other words why its in the 4th decimal place not 3rd

Answer 36

ok let me attach it

Answer 37

well if it was 3 decimal places, then 0.0004082 ----> 0.000 = 0

Answer 38

I think they wanted the p value to be some number (other than 0)

Answer 39

wow that was horrible

Answer 40

lol a bit small...hmm

Answer 41

The subjects were randomly assigned to subitramine (5213 subjects) or a placebo (5020 subjects) in a double-blind fashion. The primary outcome measured was the occurrence of any of the following events: nonfatal myocardial infarction or stroke, resuscitation after cardiac arrest, or cardiovascular death. The primary outcome was observed in 521 subjects in the subitramine group and 452 subjects in the placebo group.

Answer 42

k one sec

Answer 43

I'll be right back gotta drop someone off so I'll see your reply in a bit

Answer 44

ok take your time

Answer 45

well I'm getting 0.08778444061206 for the p value on this example

so I'm not sure how they got 0.001, typo maybe?

Answer 46

I highly doubt it...

Answer 47

hmm seems odd how it's so off

Answer 48

I'm considering puuting .001 or .0004 50/50 if either is correct

Answer 49

but 0.001 is for the example, is there anything else to the example?

Answer 50

other than what you posted

Answer 51

there is, but its just the same top part as the one posted before

Answer 52

hmm well I'm not sure then

Answer 53

I submitted it as .0004 and it was correct :) I wonder how they got .001 before...

Answer 54

I used a few different calculators and I'm getting the same basic thing

Answer 55

who knows

Answer 56

well thanks!

Answer 57

sure thing