Few simple statistics questions...

[Cuda1447]

Aight, got an exam tomorrow morning, Ive got basically everythign down, except for a few little things. I want to make sure I know it.


1. When deciding between using a z score and a t score, you use a z score when the standard deviation is known, and a t score when it is not correct? (Assuming either above 30 sample size or normal distribution)

2. How do you know when using a t score whether or not its in two tails or just one?

Thats all I can think of for the moment. Basically the second step, finding the critical value is sometimes throwing me off. Just hoping somewhere here can explain it in a clear way so I won't be the least bit foggy when I go in their tomorrow morning. Gotta ace this bad boy!

 

[Cuda1447]

Come on, someone knows this crap!


Thx in advance guys.

 

[InverseOfNeo]

You are correct for #1. And I should know #2 but I dont.

 

[clamum]

Correct me if I'm wrong, but a two-tailed test is used when your statement includes a not equals sign.

EDIT: Or an equals sign? heh, one of them! Been a couple years since I had statistics.

 

[Cuda1447]

Originally posted by: americangigolo
Correct me if I'm wrong, but a two-tailed test is used when your statement includes a not equals sign.

So when I use an = sign as my null hypothesis it would just be a one tail test?



Ok and I got one more question, just rememberd.


3. When you are determining your orig. claim and alternate claim, lets say for example you want to test the claim that something is exactly = to a 7. Would the original claim be then p (or whatever) = 7 and alt claim be p does not eqaul 7? And then that would mean that the null hypothesis would be p= 7 (The original claim)

 

[clamum]

Originally posted by: Cuda1447

Originally posted by: americangigolo
Correct me if I'm wrong, but a two-tailed test is used when your statement includes a not equals sign.

So when I use an = sign as my null hypothesis it would just be a one tail test?



Ok and I got one more question, just rememberd.


3. When you are determining your orig. claim and alternate claim, lets say for example you want to test the claim that something is exactly = to a 7. Would the original claim be then p (or whatever) = 7 and alt claim be p does not eqaul 7? And then that would mean that the null hypothesis would be p= 7 (The original claim)

No, if the null hypothesis has an equals sign use a two-tailed test.

As for question 3, yeah, as far as I know.

 

[Cuda1447]

Aight thx guys, wish me luck Im sure I'll do fine. Get up tomorrow morn. early and just review a bit, shouldn't be to hard.

 

[parsley007]

good luck, i took my stats final @ 3 today...

 

[PoPPeR]

i had my stat midterm on t/z/etc. a couple weeks ago. My stat final tomorrow is on hypothesis testing or something... not sure

eh...

gl

 

[SethK28]

Z score is any problem with a n>30 or = 30
T score n<30

H0 is your Null Hypothesis
H0: u=n
H0:u>n
H0:u<n
H0:u not equal n

H1 is your ALternate Hypotheses

h1:u=n
h1:u>n right tailed tes
h1:u<n left tailed test
h1:u not equal n two tailed test


Type 1 error: we reject null hypothesis when true
Type 2 error: we reject null hypothesis when false

if you hypothesis falls with your critical value or level of signifcane they you accept, if its outside of the range you reject.

 

[Ryan]

I'm in statistics, and I couldn't begin to tell you. I guess that's why I'm failing the class