Alcohol use appears to be decreasing from 1991 through 1996 among US high school students. However, this may partially be due to changes in the wording of the new question that assessed alcohol use from 1993 onward. The new question required reports of heavier use to qualify as having used alcohol in the last year, however the exact wording of the question is not given on the NIDA web site. For 1993 (when both questions were asked), the new question results in 3% to 6% fewer students being classified into the category of "Having used alcohol in the last year" than did the old question. If we assume that the differences observed between the two alcohol use questions in 1993 would have been around the same size in 1994 through 1996, then adding that difference to the percent obtained from using the new alcohol use question should approximate the findings if the old alcohol use question had been used. This difference is represented on the graph by the white bars at the top of each grade level bar. The following descriptions of the graph are based on adding the difference between the old and new alcohol use questions to the new alcohol question percents to estimate the values that would have been obtained if the old alcohol question had been asked. If we add this constant to each of the estimates since 1993, the percent of high school students who report having used alcohol in the last year is fairly stable over time. The percent of 8th Grade High school students who report having used alcohol during the past year decreased slightly from 54% in 1991 to 52.7% in 1996 (note that for 1993 when both the old and new alcohol question were included, the new question resulted in a 6.2% lower rate of students who report using alcohol in the last year Old:51.6% vs. New:45.5%). The percent of 10th Grade High school students who report having used alcohol during the past year decreased slightly from 72.3% in 1991 to 70.9% in 1996 (note that for 1993 when both the old and new alcohol question were included, the new question resulted in a 5.9% lower rate of students who report using alcohol in the last year Old:69.3% vs. New:63.4%). The percent of 12th Grade High school students who report having used alcohol during the past year decreased from 77.7% in 1991 to 75.8% in 1996 (note that for 1993 when both the old and new alcohol question were included, the old question resulted in a 3.3% higher rate for students who report using alcohol in the last year Old:76.0% vs. New:72.7%).

---