Statistics Help @ Talk Stats Forum - Applied Statistics
http://www.talkstats.com/
Statistics in finance, economics, engineering. Actuarial science. Econometrics. Operations research.enFri, 19 Dec 2014 15:44:02 GMTvBulletin60http://www.talkstats.com/images/misc/rss.pngStatistics Help @ Talk Stats Forum - Applied Statistics
http://www.talkstats.com/
<![CDATA[stdev hourly average --> stdev individual measurements]]>
http://www.talkstats.com/showthread.php/58937-stdev-hourly-average-gt-stdev-individual-measurements?goto=newpost
Fri, 19 Dec 2014 14:36:40 GMTFor a process variable that's measured every 2 minutes, there's a requirement that the hourly averages of this variable should lie within a normal distribution with a standard deviation of 0.1 around the set point.

I want to find out the standard deviation for the individual measurements that gives a standard deviation of 0.1 for the hourly averages.

Does anybody know how to calculate that?
]]>Applied Statisticsdamionkahttp://www.talkstats.com/showthread.php/58937-stdev-hourly-average-gt-stdev-individual-measurementsMissort Letter Sorting Machine (Statistical Approach)
http://www.talkstats.com/showthread.php/58934-Missort-Letter-Sorting-Machine-(Statistical-Approach)?goto=newpost
Fri, 19 Dec 2014 08:57:31 GMTHi all,
I'm doing some test in a LETTER SORTING MACHINE and I have to prove that with the new modifications to the machine we get a MISSORT REDUCTION(about 5%) from the base.
We have 2% of Missort produit by the machine. So, I have to prove that the modifications made to the machine can reduce the MISSORT of 5% to 2%. As you can see is very small number.

FYI: I have to compare 50.000 live mail in 2 week. In the firts week, I'm gonna do a statistical analyse without the modification. In the second week I have to do the same statistical analyse but with the modification.

What statistical approach can you suggest me?

Thanks
]]>Applied Statisticsicohttp://www.talkstats.com/showthread.php/58934-Missort-Letter-Sorting-Machine-(Statistical-Approach)Simulation
http://www.talkstats.com/showthread.php/58906-Simulation?goto=newpost
Tue, 16 Dec 2014 21:33:52 GMTI am slowly learning how to do this. In case the sources are useful to others - and that will likely mean only beginners at this, I figured I would...I am slowly learning how to do this. In case the sources are useful to others - and that will likely mean only beginners at this, I figured I would post some of what I find. It will be SAS a lot because that is what I use so I try to find sources that use that when I can.

I found this interesting. It makes sense, but I never saw it put this clearly until now.

A typical Monte Carlo simulation involves the following:
• Generate S independent data sets under the conditions of interest
• Compute the numerical value of the estimator/test statistic T from
each data set. Hence we have T1; T2; · · · ; Ts; · · · ; TS

Compare three estimators for the mean of a distribution based on
a random sample Y1; Y2; · · · ; Yn
• The three estimators are:
{ Sample Mean, T(1)
{ Sample Median, T(2)
{ Sample 10% trimmed mean,T(3)

For a particular choice of ; n and true underlying distribution
• Generate independent draws Y1; Y2; · · · ; Yn from the distribution
• Compute T(1); T(2); T(3)
• Repeat S times.