How to compare two samples with uneven sizes/comparison

[DownNAbbey]

Example, if one group was able to complete 1 out 2, it would be 50% but if another completed 2/5, it would be 40%. It's uneven because it's not taking into account that the second group that had three more tasks. How can I evenly compare the work of both of these groups? Thank you.

 

[lev888]

What exactly do you mean by 'uneven'?
In general, it's perfectly reasonable to compare fractions. 1/2 > 2/5 (50% > 40%).
Of course, there are cases when it would not be a fair comparison. E.g. 2 groups are throwing a heavy object into a target. Since people get more tired and miss more as the number of attempts increases, it would make sense to compare based on the same number of attempts. Is this the situation in your case? Then we need a way to reduce the value of hits in earlier attempts.

 

[Dr.Peterson]

Example, if one group was able to complete 1 out 2, it would be 50% but if another completed 2/5, it would be 40%. It's uneven because it's not taking into account that the second group that had three more tasks. How can I evenly compare the work of both of these groups? Thank you.

One might say that the whole point of percentages (or ratios or fractions) is to "evenly compare", by eliminating the effect of size. As lev888 pointed out, this assumes a certain sort of effect; ratios are not always appropriate for any given situation. In the absence of other information, however, the percentages you mention would be the first thing you'd try.

But if what you want to compare is not the likelihood of completion of a task, but the total number of tasks done, then you would just compare the number of tasks directly: 1 vs 2.

And if each task has a different level of difficulty, you might want to weight the tasks by some sort of scoring.

In order to find the right measure of whatever it is you want to measure, you first have to define what it is that you want to measure, and then consider the specifics that affect that measure. You haven't given us any information by which to answer the question. Details are required.