The happiness vs crime index correlation is likely skewed in that there is only 1 measure for happiness per country but each country has multiple (and arbitrarily many) measures of crime index statistics due to it being by city (also as it stands the r correlation between happiness and crime index is -0.22 and perhaps it might show a much stronger correlation b/w them if we averaged the crime index per country, grouped by country, and then found the correlation b/w happiness and average crime index). As it stands countries that have more city crime index data weigh more heavily than countries that have fewer.
The happiness vs crime index correlation is likely skewed in that there is only 1 measure for happiness per country but each country has multiple (and arbitrarily many) measures of crime index statistics due to it being by city (also as it stands the r correlation between happiness and crime index is -0.22 and perhaps it might show a much stronger correlation b/w them if we averaged the crime index per country, grouped by country, and then found the correlation b/w happiness and average crime index). As it stands countries that have more city crime index data weigh more heavily than countries that have fewer.