Correlation Between Unpaired Data?

#1
What type of statistics compares or finds correlation of unpaired data? I'm assuming an XY scatter plot wouldn't work. I'm also assuming a t-test or Krus Wallis would be more appropriate? For example: Group A: 1, 3, 5, 1, 4. (n=5) and Group B: 5, 6, 7, 1, 2, 5, 6. (n=7). Total number of samples are different and from independent experiments, however they are both looking at the same outcome. Is there a way to analyze this type of data?
 
#3
What is the issue with using a t-test or Kruskal Wallis test?
I was curious if there were other types of analysis I could use. I'm assuming I could also use Mann-Whitney U Test. If the data is unpaired, can I see which one is more specific or sensitive in terms of outcome? Let's say both independent group of samples measure the same outcome, is there a type of statistical analysis that can determine which group works best?
 

hlsmith

Less is more. Stay pure. Stay poor.
#4
The outcomes you got are what your are working with, but I am curious what possible response there could have been or how the variable was formatted (continuous, discrete)?
 
#5
The outcomes you got are what your are working with, but I am curious what possible response there could have been or how the variable was formatted (continuous, discrete)?
I'm looking at data that was previously published. I'm doing a meta-analysis. Both studies used the same experiment layout and were looking for the same outcome but different variables (for example, sample type). At first I wanted to see if there was a correlation, but if the data is not paired that wouldn't make much sense. I wonder if there is a way to compare the two data statistically.