Correlation & Regression
This week we learned two ways to describe a relationship between numerical variables: correlation and
regression. In this exercise, you will choose two possible explanatory variables from a data set, and
determine whether they are significantly correlated with the response variable. You have 2 data sets to
pick from:
• Data set 1
o Response variable: baby’s birthweight
o Other variables: mother’s age, # of cigarettes smoked daily by mother, mother’s
height, length of gestation, father’s age, father’s height, and # of cigarettes smoked daily
by father
• Data set 2
o Response variable: crime rate
o Other variables: proportion of young males in population, expenditure on police,
proportion of males vs females in population, youth unemployment, mature
unemployment, median wage
E.g.
If I choose data set 1, I may choose to explore mother’s age and father’s age.
I will compare mother’s age to birthweight, and I will compare father’s age to birthweight to see if
either or both variables have a correlation with the baby’s birthweight
