Hi everyone. is there anyone that knows if, when I have a subset of a population, I can I standardize the subset?? what I mean is, If I have a unknown number of eggs, I take 30 of that eggs and after I have to use what I know (data obtained from the subgroup) for that 30 eggs on all other eggs, although I do not know how many are the total eggs. How can I do it on R program???

I hope I clearly exposed my question.

thank you

I don't follow. What do you ultimately want to do? Standardisation has several meanings between maths, statistics and science: from the context i'm guessing you mean standard score?

 

If so this is easy in R. What have you tried so far?

What I meant was: from an unknown number of eggs (maybe 100, maybe 2000) I take 50 eggs and want to see how long before they hatch. when they hatch I know that 50 eggs hatch in 3 days...can I use this number, 3 days, to refer to all the 100 or maybe 2000 eggs from the original group?? I mean, can I say that the eggs of a species hatch in 3 days, even if I just have a subgroup?? If I should use a t-test, for example, I should know the initial population number to know if the 3 days is significant or not, right??

OK, i see the problem. It's difficult. You are correct that you should account for the fact you are sampling from a finite population. However, since you don't know the population number, you can't make the usual corrections.

 

I would just assume an infinite population and proceed with the usual analyses, bearing in mind you will have a biased estimate. The result will be that you have a larger estimate for the standard error than you would otherwise have, meaning you are less likely to detect statistically significant differences in means when comparing two species.