Comparing quantiles from one sample

#1
Hello! I'm looking to determine factors that play a role in a country's life span. I want to split the countries up using the percentiles from life expectancy (i.e. 0-25%, 25-50%, 50-75%, and 75-100%). My rationale behind this is to understand how countries with lower life expectancies compare to each other and those with higher life expectancies. Is this statistically valid and, if so, is there a name for this method? If not, is there a more established way to compare countries with low life expectancies to higher ones?
 

hlsmith

Less is more. Stay pure. Stay poor.
#2
Please better describe your intent. Do you want to say compare the median life expectancy between two countries? If so, how many comparisons are you desiring to make and how many countries?
 
#3
So I want to split the data into quantiles using life expectancy (group 1 will be countries at the bottom 25th percentile in life expectancy, group 2 will be countries in the 25th to 50th percentile, and so on). I then want to use these groups to compare things like the average percent spent on health care , total population, BMI average, etc. Does this clear somethings up?
 

Karabiner

TS Contributor
#4
AFAICS, for your analyses, you can use life expectancy as it is, without construction of quartiles.
You can perform correlation or maybe simple linear regression analyses between life expectancy
and BMI, population, health care spending etc. And you could also use life expectancy as dependent
variable in a multiple linear regression which jointly includes BMI, population, health care spending etc.
as predictors.

With kind regards

Karabiner