Grouping Observations

#1
Hello! New to the forum here and looking for help.

I have 3 separate "Urban" monitoring locations. 10 observations were made at each location.

Is it possible to combine the observations from the separate locations so I would have 30 total observations in this one "Urban" group? Is it then possible to calculate the mean concentration based on these 30 samples even if they were from 3 different locations? And lastly, would it be possible to compare the mean (or other statistic) from these combined observations to another group (for example, 15 observations made at only 1 "rural" site)?

Thank you all for your help!

-rs123
 
#3
The purpose would be to combine the the 10 observations that were taken at each of the "Urban" sites so that I have 30 total observations to work with instead of 10 at each site. This way the statistics are better (ex. confidence intervals, estimates of the mean, etc.). But would that be ok to do or do I need to keep the "Urban" observations separate since they were taken from 3 separate sites?
 

rogojel

TS Contributor
#4
This way the statistics are better (ex. confidence intervals, estimates of the mean, etc.).
Not necessarily - if the 3 urban sites are quite different, you may end up with a so much larger variance in your combined sample, that it will not be compensated by the larger sample size.