Data Set 3 Data set name: conf.dat Variable 1 = person id #

2 = sodal dass status-SOC 3 = number of times victimized-VICT 4 =AGE 5 = index of confidence in legal system-CONF 1 low 0 4 25 2 low 0 14 45 3 low 0 15 46 4 low 0 19 25 5 low 0 17 48 6 low 0 16 30 7 med 0 7 39 8 med 0 10 35 9 med 0 12 39 10 med 0 15 43 11 med 0 16 44 12 high 0 8 37 13 high 0 19 40 14 high 0 10 40 15 high 0 17 25 16 low 1 2 24 17 low 1 7 25 18 low 1 18 46 19 med 1 6 40 20 med 19 35 21 med 1 12 43 22 med 1 12 45 23 high 1 7 33 24 high 1 6 32 25 high 1 5 39 26 high 1 3 30 27 high 1 16 31 28 low 2 7 36 29 low 2 8 38 30 low 2 2 36 31 low 2 11 41 32 low 2 12 40 33 med 2 1 28 34 med 2 2 35 35 med 2 4 27 36 high 2 4 27 37 high 2 2 37 38 high 2 8 35 39 high 2 11 41 Sampie Questions A. Define a new variable that dichotomizes the CONF variable into high low groups, where high denotes a CONF value equal to or greater than 10 and low denotes a CONF value below 10. Also, define new variables from SOC and VIcr as folIows: SOCl = 1 iflow, 2 if medium, 3 ifhigh; VICrl = 0 ifVICT is 0 or 1; and VICT1 = 1 ifVICr is 2 or more.
B. Carry out a logistic regression analysis involving the dichotomized CONF variable as the dependent variable with AGE, SOCl, and VIcrl as independent variables. Assume that the goal of the analysis is to assess the effect of the number of times victimized on confidence in the legal system, controlling for the possible confounding and effect modifying effects of age and sodal dass. In your analysis, make sure to:
i. summarize the strategy used to obtain the best model.
ii. carry out an assessment of collinearity and remedy any collinearity problem that might be found.
iii. provide a summary of relevant computer results for the models you have run.
iv. provide estimated odds ratios for the effect of the number of times victimized on confidence in the legal system, controlling for age and sodal dass, based on best model. Also, evaluate the significance of the odds ratios obtained and provide 95% confidence intervals for your odds ratio(s).
v. What condusions do you draw from your analysis?
VI. How might you critidze the use of the variables SOCl and VICT1 in this analysis? What alternatives would you suggest?