Re: A question about confidential interval

It was said


I run a simulation program for 30 times and got 30 values. However,
there is very minor different between these values. For example, the
first ten values are as follows:
0.1200, 0.1201, 0.1198, 0.1202, 0.1200, 0.1202, 0.1201, 0.1200, 0.1199,0.1202.

In such case, is it still reasonable to show the confidential interval?

1. If yes, then how to show?
2. If no, then which value should be chosen as the simulation result?


**********
my feedback

if you are interested in the value take the mean
as the *true value*. The standard error
*of this mean* is the values s.d. divided by
sqrt (30)= 5.477...
Therefore our result will be:
________mean +/- s.d. /5.477

It is expected that 64.26% approximately of
the values,19,will be in the interval
___________(mean-s.d. , mean+s.d. )

_________________licas
.

Relevant Pages

  • Re: External Program
    ... > I'ha a simulation program in fortran 77. ... your new program and then run the simulation program. ... The name and format is compiler dependent. ... Prev by Date: ...
    (comp.lang.fortran)
  • Re: Error
    ... Andrew ... >Post a short compilable example that fails.. ... I have been given a simulation program that requires all of the classes in ... Prev by Date: ...
    (comp.lang.java.help)