A model was developed to predict the length of a sentence (the response variable) for a male convicted of assault using the following predictor variables: age (in years), number of prior felony convictions, whether the criminal was married or not (1 = married), and whether the criminal was employed or not (1 = employed). The table below shows the regression output.
(a) Write the regression model.
(b) Using 45 degrees of freedom, find the p-value for each coefficient. Using an α = .01, which predictor variable(s) are not significant predictors of length of sentence?
(c) Interpret the coefficient of Married.
(d) How much shorter is the sentence if the criminal is employed?
(e) Predict the length of sentence for an unmarried, unemployed, Sentencing 25-year old male with one prior conviction. Show your calculations.

Variable Coefficient Standard Error t Statistic Intercept Age Convictions Married? Employedi? 3.2563 0.5219 7.7412 -6.0852 14.3402 4.3376 0.1046 1.0358 2.5809 2.5356 0.751 4.989 7.474 -2.358 -5.656