# Really, Really, Really need help

#### J2331

##### New Member
I'm not a statistics student and know basic, basic principles. The following question is for my research project.

Mass (g) +/- 0.01 grams Drop height (centimeters) +/- 3.00 Shell
53.47 45 No crack
56.78 45 Cracked
59.33 35 No crack
52.96 35 No crack
55.33 40 No crack
58.44 40 Cracked
My question is on how I should go about representing the value of an egg's maximum height before it cracks. I have 5 data tables like the one below. I need to show a point for which an egg is at its highest point without cracking. I was hoping you would know of a solution to my conundrum. I saw that 70% of the eggs broke at the top of the range (this being 5 centimeters above the central value). 20% of the eggs broke at and 5 centimeters below the middle value. I got the percentages from all five of my trials, which meant that ten eggs were dropped from each value range (Middle, Maximum or 5 cm above middle, and minimum 5 centimeters below middle). I was hoping you could shed some light on the situation.

The masses are irrelevant. While there ma be a trend with a higher mass and a higher probability of cracking, I only care about the heights.
If you need the other data tables, I can send them
Again, I'm looking for the best way to represent the possible value for the maximum height an egg can be dropped without breaking

#### obh

##### Well-Known Member
Logistic regression will help you to calculate the probability of crack egg

#### J2331

##### New Member
thank you for the help but I don't think I want probability, I just want the value where the egg will crack with an appropriate value of error bars.

#### obh

##### Well-Known Member
I though you are talking about random process, if not it is very easy.

Will you get the same result each time if you try the same weight and height several times?

If yes, please find the point of crack, say the point when you raise one parameter by 0.1 and this change from non crack to crack.

Just draw a line of crack points

#### Karabiner

##### TS Contributor
Maybe I did not understand the problem correctely, but it might be a case for a receiver operating characteristic (ROC).

With kind regards

Karabiner