What's wrong with
=TINV(alpha,n-1)*s/SQRT(n-1)
where
s = sample SD
n = sample size

If you really want a single function, you can write a VBA user defined
function that you can call from the worksheet

Function SConfidence(alpha, standard_dev, size)
SConfidence = WorksheetFunction.TInv(alpha, size - 1) _
* standard_dev / SQRT(size - 1)
End Function

Jerry

Cyler wrote:

Excel only calculates the confidence interval for a population mean through
the =CONFIDENCE() function. I want the flexibility to calculate the
confidence interval for a sample mean as well. Can Microsoft develop a
CONFIDENCE function for sample data and not just population data?