Replicate the Logistic Regression SPSS Demonstration and Interpretation Example. Use GSS 2018. This example is done using 2004 data. Variables Coding One of the variables in the file is gunlaw, which records whether the respondent favors the gun law. The question that the respondents were asked is: Would you favor or oppose a law which would require a person to obtain a police permit before he or she could buy a gun? We will predict responses on this question. Let's open the file and examine the distribution for gunlaw. Analyze...Descriptive Statistics...Frequencies We see that one third of the respondents answered this question. However, we are not interested in those who refused to answer. Also, since we are interested in predicting who favored the gun law, interpreting the logistic output will be more straightforward if we recode gunlaw to make favor the higher category (this is a general principle to follow with logistic regression). Go to Transform ...Recode into Different Variables .... Recode the new gunlawR variable categories: Now let's examine the distribution for the recoded variable gunlawR . Analyze...Descriptive Statistics...Frequencies Click analyze... regression.... binary logistic Place gunlawR in the Dependent list box. Place age, educ, sex, racecent1, polviews, happy in the Covariates list box. Click Categorical button. In the resulting dialog box, all the variables will initially appear in the list of Covariates (except string variables, which automatically appear as Categorical Covariates). The categorical variables we listed include race, happy. Race is nominal variable, happy is ordinal. We need to add them to the Categorical Covariates list (we do not need to create dummies for sex because it is already dichotomous. It is coded 1 - male, 2- female)