Co-Principal Investigator: Dalton, Benjamin

Purpose: This exploratory project focused on industry certifications earned by high school students with support from the state of Florida. The study had two aims: (1) to identify which high school practices and labor demand factors are associated with higher rates of earning certifications and (2) to test whether obtaining a certification is associated with increased rates of on-time high school graduation, postsecondary enrollment, and degree attainment.

Project Activities: To address these aims, the project team obtained and linked data from a range of sources. First, they obtained longitudinal data from Florida's Department of Education on students in public high schools across the state, including records of their taking certification exams and earning certifications. To learn about the local labor market, the researchers obtained statewide job posting data and linked that to the districts' certification data at the county level. To learn about program implementation, the project team developed and administered a survey of education leaders and career and technical education staff in Florida's districts and traditional and vocational high schools. Finally, they used data from the National Student Clearinghouse for tracking students through college completion. Linking these data permitted a nuanced analysis of the predictors and consequences of earning certifications.

Key Outcomes:

• Except for the field of health sciences, the researchers found little association between the local demand for field-specific sub-baccalaureate credentials and students' likelihood of earning certifications in those fields (Dalton et al., 2021).
• In the first 2 years of high school, students tended to earn certifications that focused on more general skills (e.g., Adobe Rich Media) while in the last 2 years of high school, students tended to earn certifications linked to specific jobs (e.g., Certified Medical Administrative Assistant) (Glennie et al., 2023).

Setting: This project took place in urban, rural, and suburban schools in Florida. The Florida Department of Education provided longitudinal data for students enrolled in every public high school in the state, and the researchers sent surveys to educational leaders in all 67 of Florida's districts and every public high school that enrolled at least 10 students (524 schools).

Sample: The study population was composed of all Florida students in public high schools who entered as first-time 9th-graders between 2012–13 and 2015–16 (a total of 897,761 students). Of these, 49 percent were female, 22 percent were Black, 41 percent were White, and 31 percent were Hispanic. When they entered 9th grade, 59 percent of these students were eligible for free or reduced-price lunch, 21 percent were limited English proficient, 11 percent had a mental disability, and 3 percent had a physical disability.

Research Design and Methods: This work examined two distinct factors influencing certification-earning: labor market demand and school implementation policies. To address the labor market demand question, the researchers identified Florida job postings requiring a sub-baccalaureate credentials in each career cluster and linked that file to the industry certification data. To address school implementation policies, the researchers designed and administered a survey in which the respondents (one per school) described the practices and their perceptions of the program. Because not every school responded to the survey (77 percent response rate), the researchers applied survey weights to help account for nonresponse in analyses. They linked the survey data to the statewide administrative data to examine the relationship of different implementation practices to certification exam taking and passing. They then conducted analyses of the academic outcomes of certification earning using a quasi-experimental design that compared certification earners to similar students who did not earn a certification. To identify comparison groups, the researchers used multi-level logistic regression to estimate propensity scores, which they then used to create inverse probability weights. These weights permitted the researchers to compare certification earners to non-earners, controlling for the baseline characteristics. Because students not only select to earn a certification but also choose the career cluster of the certification and the timing of earning this certification, the researchers iterated this process overall, by career cluster, and by timing of certification. Because the selection process may differ by student demographic subgroups, they also iterated this process by race and ethnicity, gender, and economic status as well, for a total of 16 iterations. They achieved covariate balance in the weighted sample for each of these models.

Intervention: The research team examined the operation and outcomes of a certification policy in the state of Florida. With the passage of the Career and Professional Education (CAPE) Act in 2007, the state of Florida embarked on a novel and ambitious program promoting high school certification statewide. Since then, the number of certifications earned annually by Florida high school students has grown exponentially (over 6300 percent)

Control Condition: Due to the exploratory nature of the design, there was no control condition.

Key Measures: The researchers used administrative data housed at Florida's Education Data Warehouse for measures of taking and passing certification exams and the career cluster in which a student earned the certification. For analyses examining predictors of certification earning, they used key school implementation measures, including strong leadership, teacher supports, and inclusive recruitment practices, all of which the research team developed from responses to survey questions. They used the percentage of jobs in the county requiring a given type of certification to assess the influence of local labor market conditions on certification earning. The research team created this measure by mapping job listings that require no more than a sub-baccalaureate credential to the certifications approved for this high school program by the state in each county. Key measures of outcomes of certification earning included high school graduation, postsecondary enrollment, and postsecondary degree attainment, all of which came from administrative data.

Data Analytic Strategy: The researchers estimated relationships between local labor market and school implementation factors on student likelihood of certification using two-level logistic regressions (students nested within schools) controlling for multiple baseline student and school characteristics. They answered different questions about the academic outcomes of earning a certification by adjusting the sample. Their analyses of high school graduation focused on students in 10th through 12th grade, and this range permitted them to examine the outcomes of certification earning on students who got off track for graduation in 9th grade. In analyses of postsecondary outcomes, the researchers limited the sample to high school graduates, and this focus permitted focusing more directly on the influence of certification earning on college enrollment and attainment. All analyses employed two-level (hierarchical) logistic regression models with controls for baseline student and school characteristics.

Products and Publications

ERIC citations: Find available citations in ERIC for this award here.

Project website: https://flcertificationstudy.org/

Additional online resources and information:

"What should I know about industry certifications?" NAF Research & Tech Talk: https://www.youtube.com/watch?v=fKFG3BXaCt4

Select publications:

Dalton, B., Glennie, E., Studley, R., Lauff, E., & Warkentien, S. (2021). Do high school industry certifications reflect local labor market demand? An examination of Florida. Career and Technical Education Research, 46(2): 3–22.

Glennie, E.; Lauff, E.; Studley, R.; and Dalton, B. (2023). Pathways to credentials: Does the timing of earning an industry certification in high school influence postsecondary educational outcomes? Journal of Research in Technical Careers, 7(1), 45–67.