hi everybody
am working with stata and here below is a question that i need to answer. I do not get well how to generate the variable high and low and how they should be (dummy variable or continuous ones). I really need assistance. Than for advance.

Here is the question:
Define “high” malaria areas as those in the 90th percentile or greater in terms of “malmort1890”.
Define “low” malaria areas as those in the 10th percentile or lower. Generate a new variable (called
sei_diff) which is the difference in the log Duncan SEI index between high and low malaria areas.
Plot this difference on the y-axis with cohorts on the x-axis (note that this will look like the
author’s Figure 4).