# Hypothesis testing

1. Feb 5, 2009

### MaxManus

1. The problem statement, all variables and given/known data
I have used OLS and found that:
b0 = 6,85
b1 = 3,88 with se= 0,1121
n = 20

Person x claims that b1 = 5
Choose an alternative hypothesis. Does your estimated relationship support this claim?
Use a 5 % significance level

3. The attempt at a solution

H0 = b1 = 5
H1 b1!= 5
t((1-a)/2,18) = 2,101

t = ($$\overline{x}$$ - h0)/se(b1) = (3,88 - 5)/(0.1122) = -9.99

t lies not in the interval +- 2,101 so I reject H0

Is this correct?

2. Feb 7, 2009

### Mastermind_14

se==? and does "b0 = 6,85" means "b0=6.85"?

3. Feb 7, 2009

### MaxManus

Se = standard error

I'm not sure about what you mean with the last question, but b0 was point-estimated to be 6,85 or 6.85 if the comma was what you asked about.

4. Feb 7, 2009

### Mastermind_14

From what I know of Hypothesis test and student T distribution (from Statistics course which I am still undertaking), your answer is correct ie "t" lies not in the acceptance region but in the left rejection region thus "H0" is discarded and "Ha" is accepted instead.

A very small error that i have noticed is that you were supposed to look for the range of acceptance region corresponding to "significance level =5% " and "degrees of freedom= n-1 =19" where as you seem to have looked it up for degrees of freedom = 20 which is not correct and might result in loss of a few marks in exams even though your answer is still correct.

So
t((1-a)/2,18) = 2.09 ..................(At least in the table i posses)

I Hope you are satisfied with my reply.

5. Feb 7, 2009

Thanks:)