# Thread: how do i approach this problem?

1. ## how do i approach this problem?

how do i approach this problem? thanks in advance.

2. ## Re: how do i approach this problem?

Hi! We are glad that you posted here! This looks like a homework question though. Our homework help policy can be found here. We mainly just want to see what you have tried so far and that you have put some effort into the problem. I would also suggest checking out this thread for some guidelines on smart posting behavior that can help you get answers that are better much more quickly.

3. ## Re: how do i approach this problem?

this is what i did

first find the standard error

standard error = standard deviation divided by square root of sample size
SE = 5 / Square root of 60
SE = 5/ 7.745
SE = 0.645

next, use the z table
for the 90% confidence interval, z= 1.645

1.645 x standard error
1.645 x 0.6 = 0.987

665 – 0.987 = 664.013
665 + 0.987 = 665.987

the 90% confidence interval is 664.013 - 665.987

4. ## Re: how do i approach this problem?

You general approach is I think what they're looking for for part(b). Note that you're given the sample standard deviation though (not the population standard deviation). Have you learned about the T-distribution yet?

5. ## Re: how do i approach this problem?

yea we learned about t distribution.

6. ## Re: how do i approach this problem?

You probably want to use the t-distribution in part (a).

7. ## Re: how do i approach this problem?

what makes you say to use t distribution?

8. ## Re: how do i approach this problem?

You're given the *sample* standard deviation - not the population standard deviation.

9. ## Re: how do i approach this problem?

An interesting aside is that while some text suggest that you use the t distribution whenever you don't know the population standard deviation (which is almost always in practice since you rarely have this), other text do not. They suggest using the Z distribution as long as the population can be assumed to be normal.

Which I always found confusing. It is common in six sigma to use the z distribution with sample data which is likely wrong.

10. ## Re: how do i approach this problem?

Originally Posted by noetsi
It is common in six sigma to use the z distribution with sample data which is likely wrong.
Six Sigma only uses the z - distribution with process capability data, which typically has a sample size of 100. At 100 samples the t-distribution should be very close to the normal distribution. For hypothesis test, the t-tests and ANOVA are used.

11. ## Re: how do i approach this problem?

None of the training I had in six sigma, or the associated materials, mentioned that distinction Miner but I take your word for it.

I don't understand if you mean a sample done 100 times or a single sampling with a hundred points. I have data that has hundreds if not thousands of points that is still highly skewed because of huge outliers. It is not remotely normal.

12. ## Re: how do i approach this problem?

Originally Posted by noetsi
I don't understand if you mean a sample done 100 times or a single sampling with a hundred points. I have data that has hundreds if not thousands of points that is still highly skewed because of huge outliers. It is not remotely normal.
The data doesn't need to be normal. CLT to the rescue.

13. ## Re: how do i approach this problem?

Originally Posted by Dason
CLT to the rescue.
its power is great. it saves me every time.

ALL HAIL CLT!

14. ## Re: how do i approach this problem?

Originally Posted by Dason
The data doesn't need to be normal. CLT to the rescue.
There are various ways in which the CLT can fail though.

15. ## Re: how do i approach this problem?

Originally Posted by noetsi
None of the training I had in six sigma, or the associated materials, mentioned that distinction Miner but I take your word for it.

I don't understand if you mean a sample done 100 times or a single sampling with a hundred points. I have data that has hundreds if not thousands of points that is still highly skewed because of huge outliers. It is not remotely normal.
I'm coming from a manufacturing and business process paradigm. In manufacturing, most processes do follow a normal distribution. Data are typically collected in small subgroups over an extended period of time (e.g., SPC) then estimates are made of the short and long term variation using within subgroup variation and overall variation after determining whether the process is in a state of statistical control. The automotive industry forced sample sizes of 100 on their supply base and that practice has spread throughout the manufacturing world. In business processes, we deal with cycle times which are rarely normally distributed, so nonparametric methods are used instead.

16. ## The Following User Says Thank You to Miner For This Useful Post:

noetsi (03-28-2014)