STATS 499 Lecture Notes - Lecture 8: Exploratory Data Analysis
Document Summary
We cannot say this because we do not have a representative sample of the population of regular smokers. Please think about it for a discussion on thursday. However, actually they may not have used sampling, could have just been an exploratory data analysis. The paper used number of educated women per 100 educated men. Instead, should have used expected number of years of education for women. Could have used number of patients per physician rather than patients per hospital bed. Prediction 2, which states that countries where people are more educated tend to have higher mean age of marriage. This correlation is at the country level, and may not be translated to the individual level. Only know about averages, and can"t make conclusions about individuals from these averages. There could actually be interaction terms that exist due to cultural differences. Models omitted some variables, only included those that would help support argument. Hw 3 draft due friday, february 12.
