A study of sex offenders in the Canadian Federal Prison System was published in the British Journal of Criminology (May 2014). The following data were collected for each of 59 male sex offenders: race, type of crime (violent or nonviolent), age at first conviction, and total number of criminal convictions. Suppose you want to use the data to model the total number of convictions (y) as a function of race, type of crime, and age at first conviction. Do you recommend using stepwise regression to find the "best" model for predicting y? Explain. If not, outline a strategy for finding the best model.