[9] Many states developed initial standards for at least some subjects starting after about 1980, while others revised existing standards and/or curricular guidelines; in some states both of these activities occurred.

[10] Although effects were somewhat different for men and women, Trusty's analysis also adjusted for variables such as previous test scores, previous course completions, and confidence about their mathematics and science skills. These factors sometimes interact in both directions, with strong performance in early grades often leading to greater self-confidence and interest in the subjects, which in turn lead to greater coursetaking, which may increase performance, and so on. Studies may not measure other relevant characteristics like students' motivation and career aspirations.

[11] The fairly flat pattern for trigonometry/algebra III does not necessarily mean that fewer students studied these topics; some schools may have reconfigured courses so that rather than providing a full semester of trigonometry, for example, they may include that material in a precalculus or other course.

[12] Except for biology, AP/IB science course data are available only for 2000 and 2005.

[13] In some course categories, the difference between Asian/Pacific Islander and white graduates was not significant, whereas in others, differences between Asians/Pacific Islanders and one or more of the other groups proved to be not significant. These findings are likely due in part to large standard errors associated with smaller population groups.

[14] Poverty rate is defined as the percentage of students in the school who were eligible for the national subsidized lunch program. For reasons explained above, school lunch program eligibility can be an unreliable indicator of individual families' poverty, particularly for high school students. It is used here as a rough proxy for poverty at the school level because it is the only available measure, but the caveat stands.