I am in doubt of whether or not to include outliers in my variable Aid Received pr. Capita. My thoughts are that I would like to keep the observations, as the data tells the actual amount of aid received pr. capita. If not to include them, how do you deal with that?
However, when graphing this the very extreme outliers (high amounts of aid) obscures the rest of the aid received so it looks like it is just $0.
I have also tried logging the variable but here some of the observations turns into negative values, as a lot of the countries received $0 pr. capita.
Thank you in advance!
Code:
. summarize AidPrCap, detail AidPrCap ------------------------------------------------------------- Percentiles Smallest 1% 0 0 5% 0 0 10% 0 0 Obs 600 25% 0 0 Sum of Wgt. 600 50% 1 Mean 1240.273 Largest Std. Dev. 15891.73 75% 25 28564 90% 302.5 29902 Variance 2.53e+08 95% 1482.5 213971 Skewness 17.78929 99% 10392 322165 Kurtosis 331.846
0 Response to Should outliers be included in graph?
Post a Comment