# Testing Hypothoses, P-Value method

• Apr 19th 2011, 06:13 PM
NecroWinter
Testing Hypothoses, P-Value method
hey there guys, this question (and topic) is throwing me through a loop.

Heres the question:
When consumers apply for credit, their credit is rated using FICO scores. A Random sample of credit ratings is obtained, and, the fico scores are summarized w/ these stats: n = 18, x(bar)=660, s =95.9.

Use a .05 significance lvl to test the claim that these credit ratings are from a population with a mean that is equal to 700. If the bank of newport requires a credit rating of 700 or higher for a car loan, do the results indicate that everyone will be eligible for a car loan, why or why not?

So, tbh im not entirely sure what im doing here.

I wrote the problem in the following way
(660-700)/(95.9/(18)^(1/2)) = -1.76961
The degrees of freedom are n-1 (17)
I assume its a two tailed test, and I have the area in two tails 2.567
I take the significance level and divide it by two and get .025

and I have no idea where to go after that. Any help is appreciated
• Apr 19th 2011, 06:29 PM
pickslides
Most of your work looks good.

Quote:

Originally Posted by NecroWinter

I assume its a two tailed test, and I have the area in two tails 2.567
I take the significance level and divide it by two and get .025

It is a two tailed test, but what do you mean by 'the area in 2-tails'? The total area under the curve for the student's t-distribution is 1. Do you mean 2.567 is the critical value?
• Apr 19th 2011, 06:37 PM
NecroWinter
Since its two tailed, it covers the right extreme, and left one (This probably isnt the best way to explain it)
I have a chart with the degrees of freedom and the corresponding area.

it looks like this
• Apr 19th 2011, 07:27 PM
pickslides
Those are the critical values for the test statistic, we'll call it t(crit), so you need to compare that to what you have calculated (x_bar - mu)/(s/sqrt(n)), we'll call that t(calc).

Now if |t(calc)|> |t(crit)| reject H_0.