Re: [Phys-l] New gravitational constant

[Brian Whatcott <betwys1@sbcglobal.net>]

At 12:51 PM 1/22/2008, Jack responded to Antti in this way:

> Hi Savinainen-
>         1.5 sd is not usually considered a significant difference.
>                                                 Regards,
>                                                         Jack
>
>
> On Tue, 22 Jan 2008, Savinainen Antti wrote:
> ...
> > (6.67428 +/- 0.00067) * 10^(-11) m^3kg^(-1)s^(-2)
[and]
> > (6.67259 +/- 0.001) * 10^(-11) m^3kg^(-1)s^(-2)
> > Antti Savinainen


An interesting assertion.

If I take samples of Gaussian distributions
which produce Mean1 = 667259 +/- 150
and Mean2 = 667428 +/- 67

...and if I suppose the uncertainty, as expressed, is representing
a standard deviation (SD) in each case, and I further suppose
 that there were 400 trials in each case,
am I justified in saying that Mean1 - Mean2 must be greater than
1.5 times the SD in order for there to be a reasonably small
 probability (p<0.05) that the means are not significantly
 different?


Brian Whatcott    Altus OK    Eureka!