I would like to generate confidence intervals for a dataset... lets use student grades as the metric (range from 1-100%).
There are QV several functions that resemble the word "confidence" intervals... lower confidence interval, upper confidence interval, however, I'm not clear what the differences are for the similar functions. So I picked: TTest1_lower(Grade_N) , There are several others, i.e. ttest_conf( Group, Value ) , ttest_lower( Group, Value ), etc.
The end result is a line chart with upper and control limits overlayed. My intent was to create the two limits: "UCL" and "LCL" via the "Reference Lines" of the Chart Properties. However, this is not working.
so sorry for the late reply. I am just now revisiting the statistics application that I've been working with. What is your "mean"... is it .05...is .05 the average Range? then you add and subtract 1.96*stdev(Range)/sqrt(PopulationSize) to the "mean".... I am just restating your formula to make sure I have it right?
As far as I can remember the range will be the measure, ie amount of grades and the population will be the amount of students. This will give you an absolute value. We then worked out the % interval using a formula like this:
=(0.05 + 1.96*(stdev(Range)/sqrt(count(PopulationSize)))) / Avg(Range) which gave me a average confidence in %.
I am not quite sure how the upper and lower will work in this function as we only had to get to one answer.
If I remember correctly the 0.05 is the variable for the 95%