Calculating a p-value given the mean and SD are already computed

[mppowell]

Rather than a standard dataset of raw observations, I have a dataset of summary statistics for which I've calculated a t-test using the manual formula. Before I go about doing this for the p-value, is there a simpler way to do this in the Data Analysis add-in? The options that I see seem to ask for a range of values so that it can compute the mean, sd, etc. Thanks in advance to anyone who has advice for me!

[vandan_tanna]

I would do the following

1. Calculate pooled Variance: ( (n1-1)*Variance1 + (n2-1)*Variance2 ) / (n1+n2-2)
Pooled Standard Dev = sqrt(pooled variance)

2. Calculate t statistics: (avg1 - avg2) / (Pooled_Stdev * sqrt(1/ n1 + 1/n2) )

3: P Value = TDIST function to calculate P value . In that function X = t-stat from step #2

Source: DOE book by Douglas C. Montgomery & internet