In Figure 1, the shallow slopes of the IRFs and low peaks of the IIFs at age 11 demonstrate the lack of discrimination (and subsequently, information) in this age group. Table 3 shows that standard errors on the greatest discrimination parameter estimates (i.e., tobacco in males and marijuana and other controlled substances in females) resulted in confidence intervals that included zero (indicating that information provided was not significant) at this age. This overall reduction in discrimination and subsequent information would ultimately result in less reliable trait level estimates for individuals being evaluated with this measure in early adolescence.
