This is my question 3
Many laboratory experiments in psychology find that media violence increases aggression in the short run. We want to analyse whether media violence affects violent crime in the daily life. We collect data on the level of violence in blockbuster movies for 516 weekends in the US, and study its effect on sameweekend assaults. The variables are defined as follows: assaults: number of assaults and intimidation in the US. attend_v: attendance for stroneg violent movies (in millions]. attend_m: attendance for mildly violent movies {in millions}. attend_n: attendance for nonviolent movies (in millions}. d_chn's: indicator variable for Christmas weekend. d_newyr: indicator variable for New Year's weekend. w_ral'n: fraction of locations with rain. We estimate a few regression models using 0L5, and the results are displayed in the table below, with standard errors in parentheses. Dependent Variable: log(assaults) Coefficient Estimates Independent Variables (Standard Errors in parentheses) Model (1) Model (2) Model (3) intercept 7.6530 7.5353 7.5266 (0.1083 (0.1160) (0.1161) attend v 0.0207 0.0192 0.0195 (0.0125) (0.0124) (0.0124) attend m 0.0435 0.0423 0.0428 (0.0067) (0.0067) (0.0067) attend n 0.0213 0.0203 0.0231 (0.0071) (0.0070 (0.0072) w rain 0.4403 0.3963 (0.1620) (0.1632) d chris 0.5446 (0.3070) d newyr -0.3331 (0.3065) N (number of observations 516 516 516 SSR (sum of squared residual) 188.56 185.87 184.34 R-squared 0.0788 0.0920 0.0994(a) Use the estimated coefficients from model (1) to calculate the predicted number of assaults and intimidation on a weekend that there are 50 million attendance for the strongly violent movies, 70 million attendance for the mildly violent movies, and 25 million attendance the non-violent movies. (2 marks) (b) Use the estimates from model (2) and interpret the coefficient on w_rain. (2 marks) (c) Use the estimated coefficients from model (3) to determine if the number of assaults and intimidation is significantly lower on Christmas weekend at 5% level of significance. (4 marks) (d) Test whether the two indicator (dummy) variables included in model (3) d_chris and d_newyr are jointly significant at 10% level of significance. State all detailed steps. (4 marks) (2 + 2 + 4 + 4 = 12 marks)