Adolescent Health and Academic Achievement Study

Analytic Advice

The following examples are designed to help analysts understand the structure of the AHAA dataset and its relationship to the Add Health study. These examples expand upon information presented in the Add Health & AHAA page and we suggest researchers familiarize themselves with that information before continuing to the Analytic Advice page. When using AHAA and Add Health data, analysts must pay particular attention to how AHAA transcript information intersects with the Add Health surveys for different Add Health cohorts. Though these examples provide guidance, we encourage researchers to think critically about the correspondence between the timing of the Add Health surveys and AHAA data when selecting appropriate samples and variables to use to address their specific research questions.

This conceptual model gives an overview of all substantive aspects of the AHAA data set – including data intended for future release (i.e. curricular content exposure, and social contexts). Note that survey-based indicators capturing issues pertaining to health, relationships, socio-economic status, work, and post-secondary education are included in the Add Health data base.

This example illustrates how course-taking and survey data match up for students enrolled in 9th grade in the Fall of 1994. For this cohort, their 9th grade course-taking (considered their “Year 1” of high school) occurred during the same year as the In-School and Wave I surveys. Wave II corresponds with their second year of high school course-taking (their “Year 2”). The “linking indicators” in the AHAA data set provides school year and grade level indicators which enable analysts to match Add Health respondents’ Wave I survey data to their transcript-determined academic year placement (i.e. Year 1, Year 2, Year 3, Year 4).

Example 3: Using Add Health & AHAA to Predict Add Health WII Survey Outcomes
When using Add Health (Wave I and demographic information) & AHAA transcript data to predict Wave II survey outcomes, researchers must first determine which information for each cohort is relevant to include in their analyses. This decision hinges upon an assessment of each cohort’s academic placement (i.e. Year 1, Year 2, Year 3, or Year 4 of their high school course-taking) when the Add Health Wave I and Wave II surveys were conducted. Figures A & B below provide suggestions for how analysts can select which cohort data to use in their predictive models. For example, Figure B demonstrates that researchers interested in examining correlations between Wave I and AHAA information, and outcomes measured at Wave II should use Year 1 AHAA data for students considered to be 9th graders at Wave I, Year 2 information for 10th graders, and Year 3 information for 11th graders. For some research questions, it may be appropriate to include the 7th, 8th, or 12th grade Add Health cohorts. However, as Figures A & B graphically illustrate, decisions about including these other cohorts into analyses must be based on a clear understanding of the correspondence between their academic placement (Year 1, Year, 2, Year 3, or Year 4) and the timing of the Wave I and Wave II surveys.

Example 4: Using Add Health & AHAA to Predict AHAA End of High School Outcomes
Figures C & D (below) highlight suggested Add Health & AHAA data to use for predicting end of high school academic outcomes such as highest math level attained or cumulative GPA. These figures, as in the case of Example 3, illustrate the importance of evaluating the relationship between the timing of student coursework and the Add Health surveys when selecting samples for this type of analysis. Significantly, these figures point out that researchers can choose to use the AHAA data that corresponds to Wave I (i.e. 9th graders’ Year 1 AHAA data, 10th graders’ Year 2 data, 11th graders’ Year 3 information), or they can opt to focus on all cohorts’ Year 1 (or Year 2 or Year 3) AHAA information as a baseline or predictive measure of end of high school academic outcomes.

Example 5: Using Add Health & AHAA to Predict Add Health WIII Outcomes
Because students from all cohorts have completed their high school course-taking by Wave III (with the exception of a small number of students who were in high school for longer than four years), predicting Wave III outcomes with AHAA data involves a less complex process than required for other types of analyses. For this type of analysis, analysts may elect to use subjects’ Wave I Survey data or the AHAA data that corresponds to Wave I. Researchers can also use the cumulative, end of high school, AHAA measures. Researchers should pay attention to how long subjects were out of high school when Wave III was conducted (6 years for the Wave I 12th grade cohort vs. 1 year for the 7th grade cohort) and how far along subjects were in their high school years when the Wave I survey occurred (for the 9th grade cohort it was the beginning of high school whereas for the 11th grade cohort it was nearer the end) when designing their analyses.