Systematic Exploration of Gifted Programming: Seeking Promising Practices in Three States
In Phase 1, the Center has been conducting an exploratory study of gifted programs in three states by collecting data from four sources: a) state administrative data (n=362,254 students who were in grade 5 in 2014), b) district (n=332) and school (n=2,250) surveys, c) district gifted education plans (n=293), and 4) school site visits (n=33 school visits to date). These states were chosen because they provide funding for gifted and talented programs and can track the progress of gifted and talented students in their data systems. State administrative data were collected for grades 3-5 for the Years 2012, 2013, and 2014 and include student reading and math achievement as measured on standardized state tests, gifted status, free/reduced price lunch status, English Language learner (EL) status, special education status, race/ethnicity, birthdate, grade, and the school that the student attended during each school year for all students (n=362,254) who were in grade 5 in 2014).
The Center surveyed districts (n=332 responses) in the three states and schools (n=2,250 responses) that had a 5th grade and no fewer than 10 students per grade. We collected information about current gifted education practices, including district policies, program objectives, identification policies and procedures, instructional approaches, program curricula and content, student contact time and duration of services, qualifications of teachers, existing professional development, student assessment practices, and parent and community involvement. We also analyzed district plans for gifted and talented programming in two of the states and coded the data for key program features (n=133 variables).
The data obtained from these four sources (administrative data, district surveys, school surveys, and district plans) have been merged, and secondary data analyses is being conducted to examine the relationships between program practices and program success at the district and school levels. Specifically, the Center is using multilevel and growth curve modeling: we used Empirical Bayes (EB) estimates of the school- and district-level random effects to identify schools and districts that appeared to have particularly effective gifted programs. A separate set of analyses added school and district level programmatic variables as moderators of growth rates. We are using these analyses to identify which observable characteristics of schools and districts are useful for predicting high achievement growth rates among gifted students, especially those from traditionally underserved groups.
Using latent profile analysis based on the EB estimates, we selected four schools in each state that appear to be effective at identifying and serving gifted students from underrepresented groups on the basis of state student achievement data. We selected four additional schools per state with similar demographics that reported using identification and program services that were similar to the first set of schools we selected. We also selected 9 school districts that showed proportional representation of EL students in their gifted programs.
The Center has conducted site visits to 33 schools to date to identify components within programs that are most related to positive outcomes among traditionally underserved students in terms of identification, persistence in gifted and talented programs, and improving academic achievement. The site visits include document review, interviews and focus groups (n=190), and walk through observations of classrooms. Using the results of the secondary data analyses and the onsite visits, the Center will identify key program implementation components associated with the inclusion of traditionally underserved students and high rates of student academic growth (overall and by underserved group).
On the basis of the findings from Phase 1, the Center will develop a plan to rigorously evaluate, using a multisite cluster randomized trial, the impact of a gifted program identified from the site visits as being the most successful at recruiting and serving traditionally underserved students and potentially having the greatest impact on student achievement. The plan will undergo scientific peer review, and if approved, the Center will implement it during Phase 2 with up to $2,000,000 in additional funding.