Process survey data for a solar energy company.
> Study results from a field study in Lagos, Nigeria.
> ± 3,800 data points collected with 3 surveys in 3 segments; small businesses, lower income, middle income. (expect ±15-20% invalid/incomplete - need to be cleaned up)
> Questions around household characteristics and energy usage.
> Each survey about ± 40 questions/answers.
1) Clean up data sets - delete incomplete rows and invalid answers.
2) Present all in tables / graphs (Q by Q).
3) Lite statistics (on a few questions only) - multiple regression and/or correlate certain variables (e.g. reported income with energy expenditures).
4) 1 page summary with highlights per segment/questionnaire.
5) Per questionnaire: tables/graphs that present results on each question.
Have 3 .CSV files (1 for each questionnaire). The files are complete - data is both coded (numerical) and written out (text).
6) Plot GPS locations on a map (per segment / questionnaire).