1. Linear increase in # graduated in the years from 1950 to 1972. Then we have a linear decrease from 1976 to 1989. Then we finish with another linear increase from 1989 to 2000. This a cyclic looking scatterplot.
2. Only going to use data from the years 1950 to 1972 to generate the equation (note: coded years as 1950 = 50): grad-hat=-592.16+15.52year. Using this equation we predict for 1969 a graduating class of 478.72.
3. Only going to use datat from 1989 to 2000 to generate the equation: grad-hat = -362.19 +8.16year. Using this equation, we predict for 2005 a graduating class of 484.61. We are hesitent to use this equation becuase 2005 is out of the range of the inputs (x's) were are using. Extrapolation is risky in regression, because we are predicting into the future.
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment