4.5. After losing a court battle over a requirement that it reduce its jail population, a county conducted an analysis to predict which offenders would pose the greatest threat of committing a violent offense if released early. A random sample of 500 inmates released from the jail in the last 3 years was analyzed to see what factors predicted arrest for a violent crime in the 12 months after release. For the final model, which included five predictors of violent arrest, the county reported the following statistics: -2LL null = 876.5 -2LLfull = 861.3 PREDICTED NO VIOLENT ARREST PREDICTED VIOLENT ARREST Observed no violent arrest 439 19 Observed violent arrest 27 15 (a) Calculate the percent correctly predicted for this model. What does this statistic indicate about the county's prediction model? (b) Calculate the model chi-square for this model. Interpret this statistic. (c) Calculate Cox and Snell's R for this model. Interpret this statistic. (d) How do you explain the difference in the results for parts (a) through (c)?