|Of Special Note|
|Unlike the NAAL sample, SAAL does not oversample for minorities or provide county-level estimates. However, a state may if it wishes, augment its sample in order to get a greater representation of its minority population subgroups. Kentucky, which chose to do this in the 2003 assessment, requested a larger sample size of 1,500.|
SAAL sampling was conducted using the same approach as NAAL’s four-stage sample design. Within each participating state, the sample was drawn to yield approximately 1,000 respondents (1,500 in Kentucky) who complete the background questionnaire and at least attempt the cognitive assessment.
In the 2003 SAAL, a state’s sample size was augmented by the portion of the national sample drawn from within that state, which allows for more in-depth analysis of the participating states. In other words, a participating state’s final sample size is the sum of its SAAL sample (approximately 1,000) and the NAAL sample that is located in the state. States with large populations are more likely to have a larger number of adults drawn from the national NAAL sample than are less-populated states.