It is a pity the study referenced in the article says the exact opposite [1]. "In our large sample of adolescents, certain items still did not perform adequate

It is a pity the study referenced in the article says the exact opposite [1]. "In our large sample of adolescents, certain items still did not perform adequately from a psychometric perspective, test–retest correlations were low to moderate, and predictive validities with respect to alcohol use, smoking, and cannabis use were low and limited to the IMP and SS subscales in this sample." 1. https://www.ncbi.nlm.nih.gov/pubmed/26463560 https://www.ncbi.nlm.nih.gov/pubmed/26463560 Edit. I am referring to the 90% indentification study. I am checking out the intervention study now. Edit 2. OK in the intervention study the entire effect seemed to occur in the studied schools relative to the control schools. The overall rate of self-reported binge drinking fell for all students at the intervention schools, both those that experience the intervention and those that didn't. The major problem is the control schools did not seem to match the intervention schools in demographics so it looks like it is very hard to tell if the intervention worked.