Financial_Modeling_Midterm_13

# Of the claimants age clmageatt the interaction between

This preview shows page 1. Sign up to view the full content.

This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: T_ 0 MARSTAT_ 1 AGE GENDER = Value 0.6729935853 Std Error 0.5756814265 t Ratio 1.1690382117 Prob>|t| 0.2434308662 SS 3.1018156175 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0.852100353 0.4976430135 1.7122723114 0.0880075466 6.6543284337 Sum of Squares 6.9645989789 Numerator DF 2 F Ratio 1.5342903542 Prob > F 0.2175034266 9 5) MULTIPLE LINEAR REGRESSION We consider automobile injury claims data. The data contain information on demographic information about the claimant, attorney involvement, and economic loss, among other variables. We consider here a sample of n = 1,340 losses. More specifically, we consider the following list of variables: ATTORNEY CLMAGE CLMSEX MARITAL SEATBELT CLMINSUR LOSS Whether the claimant is represented by an attorney (=0 if no and =1 if yes) Claimant’s age Claimant’s sex (=0 if male and =1 if female) Claimant’s marital status (married, single, widowed, or divorced/separated) Whether the claimant was wearing a seatbelt/child restraint (=0 if no, =1 if yes) Whether the driver of the claimant’s vehicle was uninsured (=0 if no, =1 if yes) The claimant’s total economic loss (in thousands) Consider the attached JMP output for the regression model of the logarithm of the losses (LN_LOSS) on all the variables above plus: • CLMAGE_sqr è༎ the square of the claimant’s age, • CLMAGE*ATT è༎ the interaction between claimant’s age and the attorney dummy variable, è༎ • CLMAGE_sqr*ATT the interaction between the square of claimant’s age and the attorney dummy variable. Moreover, the marital status has been split in the dummy variables MARITAL_SINGLE, MARITAL_WIDOWED and MARITAL_DIVORCED. a) Provide an assessment of the attached regression model. b) Provide an interpretation for the CLMAGE*ATT coefficient. c) What was the objective of the analyst who did run this analysis in using so many terms involving the claimant’s CLMAGE_sqr*ATT)? age (i.e. CLMAGE, CLMAGE*ATT, CLMAGE_sqr, Do you think the analyst’s idea is supported by the data? Why? d) Do you have any suggestion on how to improve the model? 10 and Response LOG_LOSS Whole Model LOG_LOSS Actual Actual by Predicted Plot 6 5 4 3 2 1 0 -1 -2 -3 -4 -5 -2 -1 0 1 2 LOG_LOSS Predicte...
View Full Document

{[ snackBarMessage ]}

Ask a homework question - tutors are online