Where INDIRECT gives me the selection of data I am working with, and B3 is the confidence interval. What I am trying to get is a way to count X data points in from the tail of my distribution, where 1-X is the user selected confidence interval.
The Returns are the main data set here. The excel formula in the box beneath 'Hist VAR' selects the value from the tail according to the user input confidence interval. Perhaps in Qlikview, the data must be sorted from smallest to largest first, then those smaller than the percentile used are taken as a subset, and the maximum of that subset is returned as the value.
The column on the left has the data we wish to analyse. The formula above takes this data, sorts it from smallest to largest, and counts the appropriate number of data points in from the tail (i.e. if we had 100 data points, and the percentage were 90, this would sort them, then count 10 up from the bottom). This data point is the point we wish to find.