re: Science philosophy question: Which is a better way to measure?

quote:

---

Setup: a bowl of 10,000,000 marbles. 1,000 red marbles, and 9,999,000 green marbles. Which is the more accurate way to get the makeup of the red marble count? Count all the green marbles? OR Count all the red marbles?

---

If you're guaranteeing ahead of time that there are 10,000,000 marbles in the bowl, theoretically counting either one would be equally correct. But from a more human perspective with margin of error, you'd obviously count the reds. I guess I'm confused as to why that's even a question though?

quote:

---

Which, given counting error would be the most reliable method when there is likely error present?

---

Same as above. Relying on human error, it would seem obvious which is more likely to have errors, the one with the insanely higher volume.

quote:

---

If you're guaranteeing ahead of time that there are 10,000,000 marbles in the bowl, theoretically counting either one would be equally correct. But from a more human perspective with margin of error, you'd obviously count the reds. I guess I'm confused as to why that's even a question though?


Same as above. Relying on human error, it would seem obvious which is more likely to have errors, the one with the insanely higher volume.

---

Your answer explores the two ends of the problem the best.


In science, when neither parts are known you have to rely on instruments to do the counting, but there is also error involved in the instrumentation and lack of resolution of the instrumentation.


So when you are dealing with counting things on such a small scale, with those things being in such a larger, more concentrated medium, you have to ask which is more reliable?

Counting the thing that is in the least concentration, or counting the thing in greatest concentration and subtracting it from the whole to get the thing that is in the least concentration?

They both have error.



This isn't really a right or wrong problem, it is an exercise in being deliberate in choosing what/how to do science.


There is no concrete answer like most of the posters supplied. It doesn't matter if you know the concentration or not going in, you are still going to have error when those parts are physically counted.


If you use percents then your percentage needs to be to the same level of significance as your sample size. And if you are off by a fraction of a fraction, that affects marble counts on both the red or green sides.

Scruffy, you're a fellow science person!

Tortoise? What's that?