Figure 5.4b

Title:

Prevalence of past 30-day cigarette smoking (with 95% confidence band) for high school students with high and lower perceived risk of smoking as a function of historical time: 1990 – 2017.

Syntax:

Copy to Clipboard

%WeightedTVEM(data=smoking_risk,
     cluster=id,
        time=year_centered,
          dv=smk30day,
        dist=logistic, 
tvary_effect=int cig_risk,
       knots=3 1,
      weight=v5);

data results.cig_risk_overall;
  set tvem_plot_data;
  *estimated proportion of 30-day cigarette use among NOT HIGH perceived risk;
  plo_risk=exp(int)/(1+(exp(int)));
  plo_risk_U=exp(int_U)/(1+(exp(int_U)));
  plo_risk_L=exp(int_L)/(1+(exp(int_L)));
  *estimated proportion of 30-day cigarette use among HIGH perceived risk;
  phi_risk=exp(int+cig_risk)/(1+(exp(int+cig_risk)));
  phi_risk_U=exp(int_U+cig_risk_U)/(1+(exp(int_U+cig_risk_U)));
  phi_risk_L=exp(int_L+cig_risk_L)/(1+(exp(int_L+cig_risk_L)));
  *estimated odds ratios;
  OR=exp(cig_risk);
  OR_U=exp(cig_risk_U);
  OR_L=exp(cig_risk_L);
run;
proc export data=results.cig_risk_overall
            dbms=xlsx
            outfile='C:\Users\admin_sxl194\Box\LAB - ADDICTION AND INNOVATIVE METHODS (AIM)\AIM Lab_Research projects\SPRINGER BOOK\Chp5_Historical\Results\cig_risk_overall'
            replace;
run;

Equation:

\ln \left(\frac{P(C I G)}{1+P(C I G)}\right)=\beta_{0}(t)+\beta_{1}(t) \operatorname{CIGRISK}_{i}

Return to all Figures HERE.

Time-Varying Effect Modeling for the Behavioral, Social, and Health Sciences

Authors: Stephanie Lanza and Ashley Linden-Carmichael

Purchase This Book

Reprinted by permission from Springer Nature: Generalized Time-Varying Effect Models for Binary and Count Outcomes by Stephanie T. Lanza and Ashley N. Linden-Carmichael, 2021.