You have to first define the characteristics of living place, for example, education level of the community, poverty rate, access to various resources such as transportation, libraries,... and other demographic variables that may have an impact on student achievement. Again "student achievement" has to be defined carefully: is it % of students passing high school in a year or are there many dependent variables that measure student achievement (school vs. college level). Your modeling and analysis will depend to a large extent on what data is available also.
Good idea but feasibility of the study may be a point to think about.
In your case,keeping all other variables constant (like age,merit,teaching method,syllabus,food intake,outdoor activities etc.) u will have to set up experiments at different locations followed by periodic data collection to evaluate the real impact..
The important thing is to be clear about what you want to do and analyse. Like if you are talking about student's achievement, it has to be defined clearly in what context or field these achievements you are considering. Then trace out the geographic locations where higher success ratio is observed and also the locations where fewer achievements are observed. Then, as Srini Vasan suggested took up at least these parameters and may be considered many more and compare that which factor is common between these two geographic locations and design a schedule and collect some primary data, it will add weight to your findings.
I recently helped someone to determine how distance from school affects performance in science subjects. So the questionnaires were given out to students and they stated the distance their house was from the school. Based on the number values I categorised those values into short, medium and long distance. There after, I averaged their marks in different science subjects and applied a linear model to the data and made boxplots. Finding was that students who stayed closer to the school had higher marks.
You can perhaps do the same with variables such as rural and urban area as living place.