The dependent variable is "Per Pupil Spending, Public K12 ($)." The three independent variables I choose are: "State Per Capita GDP ($)," "State Spending on Higher Education ($ per capita)," and "Population under 18%." I choose each of these variables because I predicted that each would have atleast a moderatly strong relationship to the dependent variable ("Per Pupil Spending, Public K12 ($) ). I also predicted that states with lower averages on "State GDP" and "State Spending on Higher Education" would have lower "Per Pupil Spending," and the opposite for high values of those two independent varibles. I predict a postive relationship for "Under 18" and "Per Pupil Spending;" the higher perceptage of children, the higher the educational spending on K12. The Correlation matrix summarizes the findings with a 1 through 1 system, indicating the strength and direction of the relationships. I will discuss the relationship of each independent variable with the dependent variable, and how well my predicts hold up to the data, individually, with the scatterplots.
Correlation matrix:

"State Per Capita GDP ($)" and "Per Pupil Spending, Public K12 ($)" show a strong, postive relationship. The Scatterplot shows this as well (as the Correlation matrix), but shows in more detail where points fall. The findings do fit my original predict that as State GDP goes up, so will Per Pupil Spending. I was suprised that most states seem to be towards the middlelow on both variables, and most states fall between 12,000 Per Pupil Spending and 45,000 State GDP.
"State Spending on Higher Education ($ per capita)" and "Per Pupil Spending" shows a weak to weak moderate, negative relationship. The nearflat and negative pointing line on the scatterplot also shows this. I was very suprised that this relationship was weak because I had assumed states that invested more into Higher Education would invest as much, if not even more into Public K12 education; and, states that cut Higher Education would have likely also have cut Public K12 education (such as Wisconsin). The states being all over the graph also was as suprising as the weak relationship these variables produced. It does seem from the points (alone) being so scattered that there may be even no relationship, so considering the scatterness of the points, a weak relationship is indicated as well.
"Population under 18 (%)" and "Per Pupil Spending" shows a negative, strong relationship. While I had predicted a strong relationship, the negative pattern does make sense upon further reflection. Prehaps the more children a state has, the less it is able to spend on each child in the public school system because the state has to divide educational funds by more than a state with fewer children. A lot of points are clustered right on the mean line or right below.
Already a member? Sign in.