Creating a Confidence Interval based on data from a column

[Alexander_Arozi]

Hello

Im trying to create a Confidence Interval but have no idea how to proceed. The CI is based on data from a Z-test, Independent samples with unknown variance, but assumed equal.

Because of the test, I thought of creating all the necessary components as measures (average, variance, standard deviation, n etc.) but this was a problem since I cant aggregate already aggregated data. the original tables data is based on an estimation model we’ve created which contains aggregated measures.

Initially it feels like i’ve approach this wrong, it shouldn’t be this complex to build out a CI (?),

Any recommendations of how I shall proceed?

[GabrielZucchi]

SAME TROUBLE

[milanb]

Hey there!

If you have all the relevant values of the confidence interval built as Looker measures (i.e. if you have the measures for the mean), you can simply calculate the standard deviation within a table calculation using the "stddev" functions. Once you have this in place, you can create a new table calc with z-score * stddev +/- mean to calculate the interval.

There's some additional context around standard deviation in this post if it helps share additional context on stddev calculations.