Re: How to determine the stability?

I suggest that you recheck your formula and your data. Excel help for the
CONFIDENCE function will tell you in which situations you can get the #NUM!
error, and none of them should occur for the formula given, unless all the
sample values are the same (which would, of course, give a STDEV of zero,
and the confidence range has a trivial answer, being of width zero for any
confidence level). The fact that your numbers fall between 0 and 1 should
not be a problem.

What values does your sample give you for STDEV(range) and COUNT(range)?
--
David Biddulph

"Eric" <Eric@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote in message
news:F041AE1B-C8A0-4FFD-9075-E3BFC9F86FD4@xxxxxxxxxxxxxxxx
Thank everyone very much for suggestions
When I use following code, error - #NUM occurs when all data 1 > numbers
>= 0
Does anyone have any suggestions on how to fix it?

=CONFIDENCE(0.05,STDEV(range),COUNT(range))

Thank everyone very much for any suggestions
Eric

"David Biddulph" wrote:

You can use =CONFIDENCE(0.05,STDEV(range),COUNT(range))
--
David Biddulph

"Eric" <Eric@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote in message
news:2DA8022B-61CE-43AE-BE44-1FBCEED2F3F8@xxxxxxxxxxxxxxxx
Could you please tell me where to locate the data of sample for
=CONFIDENCE(0.05,2,30)? It is valid statement to insert a number in
CONFIDENCE(0.05,2,30), but it is not valid statement to insert a list
of
sample data in CONFIDENCE(0.05,2,range)?
Does anyone have any suggestions?
Thank everyone very much for any suggestions
Eric


"Joel" wrote:

For 95% you should be entering .05 like the VBA help indicates.

What confidience means for 0.133 is that 95% of your Normalized Value
will
lie on the x-axis over a span of 0.133. the smaller the number the
more
stable your results.

3 sigma is normally refered to as 95% which means that 95% of your
measured
values will be within the center of a normal curve. 6 Sigma results
which is
98% is better results. 6 Sigma will have a smaller confidence number.

What we are doing is looking only at 95% of the data and throwing the
rest
away and then seeing how the 95% of the data compares between the
three
differrent sets of data. You may want to run the data at 90% and
compare
the
90% results to the 95% results.

"Eric" wrote:

Thank everyone very much for suggestions

After using the function confidence (95%, 2 S.D. 100 data samples)
there are the result, 0.133, 0.155, 0.137.
Does it mean that data under column A will be the most stable?
Do they compare like with like? since they are different set of
numbers?
Does anyone have any suggestions?
Thank everyone very much for any suggestions
Eric

"Joel" wrote:

Try confidence (see spread*** help). Use the 95% and compare
rresults.

"Eric" wrote:

Does anyone have any suggestions on how to select the most
stability among 3
lists?
If I use STDEV(), I cannot compare which one is the most
stability
with
different scaling.
Does anyone have any suggestions?
Thank everyone for any suggestions
Eric

"Gary''s Student" wrote:

Consider the standard deviation
=STDEV()
--
Gary''s Student - gsnu200812


"Eric" wrote:

There are three lists of numbers under column A,B,C
Is there any build-in function to determine the stability
within change?
For example,
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
On above list, there is no change at all, which is
considered
very stable.

Does anyone have any suggestions on following lists?
Thanks in advance for any suggestions
Eric

Under column A:
1064 1099 1116 1154 1105 1118 1092 1089 1087 1055 1057 1058
967
923 874 845
864 950 1024 971
Under column B:
845 840 887 863 865 886 851 843 825 805 818 792 711 687 677
642
642 640 681
702
Under column C:
1064 1065 1100 1129 1098 1120 1086 1084 1070 1022 1034 1003
921
883 862 843
843 882 923 905






.

Loading