1. How does linear regression differ from analysis of variance (ANOVA)?
2. What does the regression line add above and beyond just correlation?
3. How does the regression line relate to the correlation of two variables?
4. The F statistic is a ratio of what kinds of variance?
5. Why is the F statistic always positive?
6. When is an analysis of variance (ANOVA) a more appropriate test to use than a t-test?
7. Why is multiple regression often more useful than simple linear regression?
8. A meta-analysis found a negative correlation between birth weight and blood pressure later in life.
   1. Explain what is meant by a negative correlation between these to variables.
   2. If you were to examine these two variables with simple linear regression instead of correlation, how would you frame the question? (hint: the research question for the correlation would be: is birth weight related to blood pressure?)
   3. What is the difference between simple linear regression and multiple regression?
   4. If you were to conduct a multiple regression instead of a simple linear regression, what other predictor variables might you include?
9. As regression analyses become more common in situations we encounter every day, there are bound to be problems. Below are two examples of problematic regression equations. For each, state the likely predictor variable or variables and the likely outcome variable or variables. Be as specific as you can. Based on just the information provided, is this an example of a simple linear regression or a multiple regression? Explain your answer.
   1. The Wall Street Journal reported that the office supply store Staples used a regression to determine the prices for customers at different locations Customers who lived farther away from office supply stores that were its competitors were charged higher prices than customers that lived near its competitors. The problem? Customers living further away from competitors were more likely to live in poor or rural areas.
   2. A 2015 report examined the predictive equations used in sentencing those convicted of crimes in an attempt to predict the likelihood that an individual would become a repeat offender. Beyond a person's criminal record, the equations used information like age, job history, and patterns of crime within a person's family. The reporter asked if its fair to score people based on not only their past criminal behavior, but on statistics about others that fit the dame profile?