Determining appropriate range from test data

[ChaoticLlama]

Main Question or Discussion Point

Hello all.

I am performing chemistry experiments and recording the times of various observed results. What I am doing currently is performing the test several times in order to determine what result is "normal."

My goal is to determine a range as a standard by which all future tests will be compared against. I do not know the correct expected time for these results; this will be set by my experiments.

Say for example we use this test data in seconds:

22
25
26
25
25
25
26
27
24
25​

The full range of data is 22 to 27 with the average being 25, But maybe the "acceptable" range should be 25-27.

I would like to know what tools I should use to analyze this data to determine statistically what range of data should be considered normal.

Thank you.

 

[chiro]

Hey ChaoticLlama.

One thing you can do is to use a distribution (i.e. a statistical one based on your sample, or if you make an assumption on the population), and look at the interval corresponding to the a/2 in each tail where 1-a corresponds to the probability that defines your interval.

A lot of people would use a normal distribution and then find the 95% interval for that distribution, but this isn't always right since not everything is always just the "default".