The Student Room Group

stats2 confidence intervals

Hi,

I've got the following question which I really can't get my head round....any help would be really appreciated!

Q: Applicants have a physical exercise test and are consequently attributed a score. These scores are normal. A random sample of 110 applicants had a mean of 49.5, sd=16.5. Calculate an interval within which the scores of 95% of applicants will lie.

Ok, so I looked through my stats 1 book and found similar questions, and they way they were worked out was to find the appropriate z value, multiply by the sd, but then do not divide by the sq root of n. Then add and deduct this value from the mean to get the range.

Also, because the sd is taken from the sample, I know I need to use the t-distribution value. As v is large there isn't a precise value, the nearest values are for 100 or 125. I calculated values with both, to see if I could get the right answer, but neither did.

Can anyone give me any tips?

Thanks
Original post by marcsaccount
Hi,

I've got the following question which I really can't get my head round....any help would be really appreciated!

Q: Applicants have a physical exercise test and are consequently attributed a score. These scores are normal. A random sample of 110 applicants had a mean of 49.5, sd=16.5. Calculate an interval within which the scores of 95% of applicants will lie.

Ok, so I looked through my stats 1 book and found similar questions, and they way they were worked out was to find the appropriate z value, multiply by the sd, but then do not divide by the sq root of n. Then add and deduct this value from the mean to get the range.

Also, because the sd is taken from the sample, I know I need to use the t-distribution value. As v is large there isn't a precise value, the nearest values are for 100 or 125. I calculated values with both, to see if I could get the right answer, but neither did.

Can anyone give me any tips?

Thanks


The interval mean value ±1.96 \pm 1.96 times the standard deviation will contain 95% of the observations for a normal variable.
Original post by marcsaccount
Hi,

I've got the following question which I really can't get my head round....any help would be really appreciated!

Q: Applicants have a physical exercise test and are consequently attributed a score. These scores are normal. A random sample of 110 applicants had a mean of 49.5, sd=16.5. Calculate an interval within which the scores of 95% of applicants will lie.

Ok, so I looked through my stats 1 book and found similar questions, and they way they were worked out was to find the appropriate z value, multiply by the sd, but then do not divide by the sq root of n. Then add and deduct this value from the mean to get the range.

Also, because the sd is taken from the sample, I know I need to use the t-distribution value. As v is large there isn't a precise value, the nearest values are for 100 or 125. I calculated values with both, to see if I could get the right answer, but neither did.

Can anyone give me any tips?

Thanks


The interval mean value ±1.96 \pm 1.96 times the standard deviation will contain 95% of the observations for a normal variable.
I don't think you need the t-distribution with a sample as large as 110
Original post by brianeverit
The interval mean value ±1.96 \pm 1.96 times the standard deviation will contain 95% of the observations for a normal variable.
I don't think you need the t-distribution with a sample as large as 110



Hi Brian, thanks for your help.

What is the interval mean value? Isn't that always set to zero? I have the mean value of 49.5, but when I do what you suggest with this, this doesn't get the right answer.
(edited 9 years ago)
Reply 4
Original post by marcsaccount
Hi Brian, thanks for your help.

What is the interval mean value? Isn't that always set to zero? I have the mean value of 49.5, but when I do what you suggest with this, this doesn't get the right answer.



I don't think brian meant that to be read as one phrase - I think he was referring to the interval defined by mean value - 1.96 x sd to mean value + 1.96 x sd :smile:
Original post by davros
I don't think brian meant that to be read as one phrase - I think he was referring to the interval defined by mean value - 1.96 x sd to mean value + 1.96 x sd :smile:




Ahh ok :colondollar:.

Brian and Davros, thanks to the both of you for your help :biggrin:

Quick Reply

Latest