Determining which excel T-Test to use

If I understand correctly, you have paired data: you measured blood pressure (systolic, diastolic, MAP) on the same individuals in two different situations, standing and sitting.

If you are interested in comparing standing BP versus sitting BP, you should do a paired t-test. This is because the measurements are naturally paired, one measurement for a person standing and another measurement for the same person sitting. This is useful because it takes into account particular differences among the observed persons and only makes comparisons when keeping those particular differences constant among paired individuals.

But if you are interested in comparing male BP (say, systolic) versus women BP, you can't use a paired t-test, since observations are not paired. You could do the following:

1. Formally test whether variance in BP (systolic) in men is the same as in women. You can do this parametrically, i.e, assuming a normal distribution on the measurements (which you can check with a qq plot), using an F test for equality of variances, or you can do it non-parametrically using Siegel-Tukey test (centering the data first).

If you do this and you conclude the variances can be considered the same, then use the two-sample t-test with equal variances.

If the variances can't be considered the same, use the two-sample t-test with unequal variances.

2. Don't formally test if the variance is the same for men and women, but compute the variance for each group and check if they are comparable.

If you do this and you conclude the variances are comparable, then use the two-sample t-test with equal variances.

If the variances can't be considered the same, use the two-sample t-test with unequal variances.

3. Don't check  if variances are the same or not and just do a two-sample t-test with unequal variances. You will be sacrificing power (ability to detect differences in the two groups, if they exist), but it might not be a capital sin.