Post Your Level I Quant Questions Here

[Arif Irfanullah]

I you have any Level I Quant questions you can post them under this thread by end of day today (8 May). I will try to address as many of these questions as possible.

 

[Galli]

Thanks for volunteering!
I consistently have issues with the meaning behind a confidence interval.
As I understand it, in general, a confidence internval is usually:
Mean +/- Reliability factor(Standard error)
I have trouble understanding what this tells mean, does this mean that a 90% confidence interval, (5%*standard error) of the sample will fall within +/- of the mean? This seems to be an important stepping stone that is limiting my ability to understand hypothesis testing and how to check if a valuefalls in this range

 

[S2000magician]

When you have a 90% confidence interval for the mean (which is what you’ve described), it means that there is a 90% probability that the mean of the population lies in that interval (which is constructed based on the mean of the sample). There’s a 5% probability that the population mean lies in the region less than the interval, and a 5% probability that the population mean lies in the region greater than the interval.
You can also have a 90% confidence interval for a random observation (which uses the standard deviation instead of the standard error); it will be wider than the confidence interval for the mean.

 

[Galli]

To say it simply, the confidence interval is an indication of how accurate the sample is in comparision to the population mean? In this case, we’re 90% confident the sample mean falls within a 5+/- minus range of the population mean? Is this for accuracy or testing? I just don’t get how this applies…
thanks for the help