Standard deviation returning wrong results

[Salmonredd]

When using stdev with three identicle values it should return 0 as a result. When the values 1.2, 1.2, 1.2 are used I am getting a result of .000000021...... Does anyone know why this is happening? It does it again with (3) 1.4's. I am considering trying to write a formula that will return 0 when the three values are identical, but doesn't that defeat the purpose of the stdev formula? Any help with this will be greatly appreciated.

EX: =IF(S44>2,STDEV(S20:S39),"")

 

[Aladin Akyurek]

When using stdev with three identicle values it should return 0 as a result. When the values 1.2, 1.2, 1.2 are used I am getting a result of .000000021...... Does anyone know why this is happening? It does it again with (3) 1.4's. I am considering trying to write a formula that will return 0 when the three values are identical, but doesn't that defeat the purpose of the stdev formula? Any help with this will be greatly appreciated.

EX: =IF(S44>2,STDEV(S20:S39),"")

If S20:S39 is numeric and without any blanks and

[1]

=SUM(1/(COUNTIF(S20:S39,S20:S39))

which is confirmed with control+shift+enter (instead of the usual enter), results in 1...

[2]

=STDEV(S20:S39)

would result in 0.

The point is that what you see in a cell is not necessarily what Excel sees.

If S20:S39 houses formulas, the numerical results they bring about has to be mapped onto the binary representation of numbers. At this point we get into the issue of floating point arithmetics on binary devices. I think you can find a lot on the latter thru Google.