Skip to main content
Cross Validated

Return to Question

deleted 128 characters in body; edited title
Source Link
R Evans
R Evans
  • 11
  • 1
  • 3

Testing for significance of difference between population and sampletwo populations

I am trying to test the significance of the difference between a population median/mean and a sample median/mean. The sample data is non-randomtwo populations.

Here is a link to the population 1 data: http://pastebin.com/YABptSqnhttp://pastebin.com/914mCRqA

and the samplepopulation 2 data: http://pastebin.com/wngVQQPphttp://pastebin.com/PmDgLLAb

Could someone tell me whether it is best to use median or mean and hence what test to use?

I was thinking of using the sign test for medians, but (i) I can't find any tables with large n and (ii) the samples areI'm not independent of each othersure if median is the most suitable measure. I've done a Z-test but not sure if itthat is suitable as well. Would resampling and subsequent testing between samples be a good idea? Thanks.

Testing for significance of difference between population and sample

I am trying to test the significance of the difference between a population median/mean and a sample median/mean. The sample data is non-random.

Here is a link to the population data: http://pastebin.com/YABptSqn

and the sample data: http://pastebin.com/wngVQQPp

Could someone tell me whether it is best to use median or mean and hence what test to use?

I was thinking of using the sign test for medians, but (i) I can't find any tables with large n and (ii) the samples are not independent of each other. I've done a Z-test but not sure if it is suitable. Would resampling and subsequent testing between samples be a good idea? Thanks.

Testing for significance of difference between two populations

I am trying to test the significance of the difference between two populations.

Here is a link to the population 1 data: http://pastebin.com/914mCRqA

and population 2 data: http://pastebin.com/PmDgLLAb

Could someone tell me whether it is best to use median or mean and hence what test to use?

I was thinking of using the sign test for medians, but I can't find any tables with large n and I'm not sure if median is the most suitable measure. I've done a Z-test but not sure if that is suitable as well. Thanks.

Source Link
R Evans
R Evans
  • 11
  • 1
  • 3

Testing for significance of difference between population and sample

I am trying to test the significance of the difference between a population median/mean and a sample median/mean. The sample data is non-random.

Here is a link to the population data: http://pastebin.com/YABptSqn

and the sample data: http://pastebin.com/wngVQQPp

Could someone tell me whether it is best to use median or mean and hence what test to use?

I was thinking of using the sign test for medians, but (i) I can't find any tables with large n and (ii) the samples are not independent of each other. I've done a Z-test but not sure if it is suitable. Would resampling and subsequent testing between samples be a good idea? Thanks.