estimate of measures

Hello everyone, I am a statistics student, I'm studying for a project in my University. I have to analyze the data, I have available the measures in millimeters corrisponding two different types of objects, of course I have already done tests to groups such as t-test, actually these organized group differ. My question is how do I estimate a threshold value in millimeters noting that the measure allows me to classify it in one group or another? I thought about using a valuation bootstrap to estimate the average with its confidence interval. I would like to know if there is a specific methodology. Thank you


Less is more. Stay pure. Stay poor.
If I follow your description, it sounds like you could use the AUROC the area under the Receiver Characteristic Curve. I would also advice to make two histograms as well to see if there even is a good threshold for distinction.