Fluff,

You cannot 'backtransform' mean values, either:

Log10
N1 6 0.77815125
N2 8 0.903089987
N3 10 1
Avg: 8 0.893747079

10^0.893747079 = 7.829735282 not 8.

The greater the spread of numbers, the greater the error. You can use an
array formula (entered using Ctrl-Shift-Enter) like

=AVERAGE(10^A2:A4)

to directly calculate the average (mean).

I'm sure Harlan will come around to set me straight, but you calculate the
standard error directly, again with an array formula (entered using
Ctrl-Shift-Enter)

=STDEV(10^Rng)/(SQRT(COUNT(Rng)))

Like

=STDEV(10^A2:B20)/(SQRT(COUNT(A2:B20)))

Bottom line - only use a log transform if there is a meaningful reason to do
so....

Bernie
MS Excel MVP

"fluff" wrote in message
...
Hi everyone,

I'm trying to find the back transformed values of my log10 transformed
standard error values.

The mean is 4.5857. I did 10^4.587 which gave me the back transformed
value of 38521.

The standard error is 0.11893. I have realised you cannot back
transform it in the same way as you can with the mean. Does anyone
know the calculation for backtransforming standard error which I can
put into excel?

Many thanks.