How can I figure out a set of values that most of my population will hit 99% of the time?

Question

I am not a math person, I'm trying to put together a spreadsheet in excel that takes the square footage for every property in a county as the only input.

I want to then generate a list of square footage range amounts that in plain english represents;

with almost >90% ( or something) accuracy every property you have will fall in one of these ranges.

So instead of having 900,000 records, I have x number of "ranges"...that almost everyone would fall into....

I'm thinking this will involve stddev and what not but I am clueless.

Answer 1

You are looking for a prediction interval. If you assume a normal distribution you will need to estimate the parameters and then apply the formula: $\overline{X}_n\pm T_a {s}_n\sqrt{1+(1/n)}$ (for a two sided interval)

If you are not a math person, perhaps it would be safer to something premade for excel. A google search for "prediction interval excel" gave some promising results with free software. It would also be more flexible to use, if you wish to modify the $T_a$, which is typically found in a table.

Answer 2

You have a list of all the square footages in a spreadsheet. Why not just sort these values and then pick $x$ ranges that contain 99% of the values?