Career and technical education (CTE) providers play a crucial role in educating and training today’s students for tomorrow’s workforce. Yet many do not have the capacity to make full use of their data for decision-making to keep pace with the rapidly changing labor market. The McKinsey Global Institute expects that about a third of work activities will be automated by 2030, with generative artificial intelligence (AI) enhancing some professions (in science, technology, engineering, and math as well as business, legal, and creative fields) while eliminating many service and support jobs, and CTE providers must be able to adapt to these dynamics.

The Center for Data Insights (CDI) at MDRC partners with CTE organizations—including nonprofit training providers, educational institutions, and government agencies—to augment their data analytics capacity. CDI supports the analysis that happens before, during, after, or separate from impact evaluations (historically MDRC’s concentration) to facilitate focused insights that can be more immediately applied for program improvement. CDI’s partnerships represent a spectrum of involvement in the CTE provider’s data work and generally fall into one of three collaboration models: consultation, cocreation, and coaching.

Consultation Collaboration

In consultation collaborations, CDI is closely involved in the data work, with the CTE partner primarily defining the focus and scope of the research agenda and reflecting on the results. The CTE provider benefits from rapid analysis designed to produce findings that can be translated promptly into action, in the process learning more about the potential for additional insights inherent in its existing data.

As an example, MDRC has enjoyed a long-term engagement with Per Scholas, a nonprofit technology training provider that connects individuals from low-income backgrounds to tech careers. Initially partnering on a randomized controlled trial, MDRC and Per Scholas have gone on to several consultation collaborations. A recent project proceeded as follows:

• Per Scholas laid out an operational goal of improving its applicant-to-participant conversion rate to meet higher enrollment targets. Its hypothesis was that the assessment it administers to determine whether people are a good fit for the program may be too challenging.
• CDI worked with Per Scholas to clarify the project’s goal and scope.
• CDI conducted data analysis independently, frequently sharing interim results.
• CDI found that a protracted intake process preceding the assessment may have suppressed the applicant-to-participant conversion rate more than the assessment itself.
• CDI proposed shifting to a human-centered redesign of the intake process and facilitated a collaborative customer-journey-mapping activity with Per Scholas to identify points in the process at which applicants are likely to drop out.
• Per Scholas made changes to its intake process and pilot tested the new approach.
• CDI supported tracking and measuring of the revised intake process.

The consultation model places more emphasis on quick, relevant, and practical results and less on skills development, making it a good fit for time-strapped organizations looking for rapid-cycle, short-term results.

Cocreation Collaboration

Cocreation collaborations provide opportunities for CTE staff members to learn data coding and documentation best practices on the job by sharing leadership roles with CDI researchers. It can take longer to reach insights than it does with the consultation model, but partners garner technical skill sets as well as coauthored code for their own future analysis efforts.

A cocreation collaboration between CDI and the Washington Student Achievement Council (WSAC), Washington State’s higher education agency, involved a formative evaluation of King County’s Bridge to Finish campaign to help students with food and housing security needs complete community college. The collaboration included:

• Working as one integrated CDI-WSAC team, with representatives from both organizations sharing responsibility for project tasks.
• Making use of MDRC’s Social Policy Research and Operations Unified Technology (SPROUT) platform, which is federally certified for secure multisystem collaboration, to conduct paired coding and code reviews.
• Engaging in agile project scoping by starting with small tasks and iterating often.
• Coauthoring a report for public dissemination.

The cocreation model fosters data analytics capacity at the partner organization but presents some challenges. With shared leadership, it’s vital to establish clear roles and responsibilities, with flexibility for unanticipated events. CTE organizations often need to suddenly change course when institutional or policy shifts happen, and data partners must have fallback plans to take on more project work if CTE staff members get diverted.

The cocreation model can be more productive and cost-effective than the consultation model for both organizations if producing data insights that can be put into action and helping the partner build skills are both priorities.

Coaching Collaboration

Coaching is the most resource-intensive model for the CTE provider, but these collaborations hold the most promise for ongoing data use. The CTE partner leads the work, with CDI providing training, mentorship, and resources to promote long-term self-sufficiency in the management of data projects.

CDI recently took the coaching approach with eight state Temporary Assistance for Needy Families (TANF) agencies, which often partner with education and labor agencies as well as community-based CTE organizations to support working-aged parents. This collaboration explored research questions related to employment, with the goal of facilitating the routine use of agency data to inform policy and practice that would ultimately improve employment and well-being outcomes.

This two-year project, called the TANF Data Collaborative, guided the partner agencies through the life cycle of a data project: laying groundwork; accessing, preparing, and analyzing data; and communicating findings. The agencies performed the data work, with CDI coordinating with teams from Chapin Hall, Actionable Intelligence for Social Policy, and the Coleridge Initiative to offer coaching, training, and resources, including:

• The Coleridge Initiative’s “Applied Data Analytics” course. Agency participants worked with existing TANF data on a secure data platform to learn how to analyze real-world TANF and employment research.
• A dedicated data coach for each agency, who helped agency participants apply data analytics lessons to their work.
• Monthly webinars providing instructional content and facilitating networking among the state agencies.
• Tools and resources developed specifically for state agencies, including toolkits on accessing and using administrative earnings data and sustainable data use.
• Dissemination avenues for project work, including a page on MDRC’s website that documents each state agency’s pilot project.

Some of the agencies continue to perform analyses independently, applying what they learned to other data sources, such as the Workforce Investment Opportunity Act and statewide longitudinal data systems.

The coaching model, a learning-by-doing approach, is ideal for CTE providers who are looking to build sustainable data analytics capacity and able to dedicate consistent and ongoing staff time.