The statistical question I am interested in is the relationship of peoples income, the dependent variable, and their relative happiness level, the independent variable. So, x= income level, and y= happiness level. I m interested in this because of the old cliché, money cannot buy you happiness, but it seems like life would be happier if money was not a constant worry. To do this, we could examine a randomly selected group of people from different income levels and give them a survey regarding their happiness level. A linear regression analysis could be appropriate because it would help demonstrate a relationship between the two variables. Trying to predict Y beyond the highest value of X would not yield accurate results.  

How do I examine this problem to assess the appropriateness and accuracy of using a linear regression model? How do I discuss the meaning of the standard error of the estimate and how it affects the predicted values of Y for that analysis?