Some extra fund we do not have. We have census data (that is coss-sectional study also) regarding the age of people in each 5 years bin and reported as histogram. I made a simulated dataset of 2000 people (comparable to our sample) in which each 5 years group has a number of people proportional to the data reported in census statistics and the age is simulated picking a random number in uniform distribution (runif in R) for each age group. Can I use this simulated group directly or may I have to bootstrap this data to have a correct W statistics? Could someone provide some reference? I could not find anything similar.

More Enzo Maria D'ambrosio's questions See All
Similar questions and discussions