95% Confidence Interval with the difference of 2 means

At first glance it appears to be more of a statistics question than an Excel question (or, at least, we need to answer the statistics "how does anyone using any software program calculate confidence intervals?" before we can work on the Excel specific programming part of the problem). Do you already have a formula for calculating these confidence intervals? If so, share with us the formula for confidence intervals, and we should be able to help translate that formula into an Excel formula. If you are uncertain of the formula, this was the first result from my internet search: https://www.statology.org/confidence...between-means/ I note that the given formula asks for a t score (I don't see a t-score anywhere in your spreadsheet, so maybe need to look into Excel's built in t distribution functions) and a pooled variance.

Does that help?

I see now. I have never used the CONFIDENCE.T() function. I see from the help file that it was intended to be used for confidence intervals for a mean, with no guidance for how it would apply to a difference of means. Again, this feels like more about understanding the statistics behind the problem (and the built in function, which MSFT's help files are often very poor at explaining), and then being able to apply the expertise in statistics to the Excel programming problem.

I apologize for not having a solid response. If this were me, I would probably:

0) Ask someone with better expertise in the statistics behind the CONFIDENCE.T() function what is expected here. I don't know if anyone on this forum has that level of statistical expertise, but maybe someone will come along and enlighten us.
1) In the absence of someone to educate me, I would try calculating confidence intervals using the formula in the statology link -- with the assumption that this formula is correct.
2) Then repeat the calculation with the CONFIDENCE.T() function. I would guess that the "standard deviation" argument would be the pooled standard deviation. Not sure about sample size, but I am guessing that one should somehow "pool" the different sample sizes into something meaningful here.
3) Compare (1) and (2) and adjust my guesses at the "standard deviation" and "sample size" arguments until (2) gives the same result as (1).
4) Look at my results and see if I can deduce something about the statistics behind the function that teaches me what I didn't know before.