Hints for Problem Set #5

Question 1:
In general, parts A B C and D are four separate questions. They do not all refer back to housing in the bay area - only the first one does. Sorry if that was unclear.
Part A: think about what a slope IS, conceptually
Part B: why does it make sense that the slope and the correlation coefficient would have the same sign?
Part C:  Think about what we talked about today in regards to extrapolation
Part D: What type of data must be measured in a scatterplot - and are we given that type of data?

Question Two:
Part C: Think about what must be true about B0 in order for it to not be possible, then examine whether or not this is true within your regression model
Part D: There are a few correct answers for this. My advice is to use statistical language, and use it correctly to justify your answers. BE SPECIFIC, and read your answer aloud to yourself when through to make sure you sound like you're making sense.

Question Three:
Part C: Remember, to calculate a residual: RAP. Residual = Actual - predicted. To get the predicted you have to plug your x's in your regression equation. Your actuals are given to you.
Part D: what must be true for an influential observation? where would you expect it to be located on the graph relative to the rest of the data?
Part E: This one comes down to whether or not you can articulate the difference between outliers and influential observations....if you draw your favorite fruit next to this problem I will give you five bonus points. Ten if it's pretty.

Question Four:
Part A: Just describe the correlation: direction and strength of association
Part B: Think about what we talked about in regards to causation. Can we justify a casual relationship here? What other factors may have led to forest fires besides the ones given?

Question Five:
Part A: what variables is the problem talking about? "Constant" is not a variable - it's your y-intercept. Think about the example we did like this in class.
Part B: can't solve this one without an equation...think what features of the chart show you this
Part C; what important stat concept is this describing? Check your notes - we talked about this specific interpretation.