Bringing Them Back: Data Mining Is Key to College Completion Efforts
An obvious avenue to meeting national, state, and local goals for increasing the number of collegiate completers is finding adults with prior college credits and steering them toward graduation. Their path is much shorter than for a new student. The trick is finding them.
With an increasingly mobile and transient population, developing reliable methods for identifying and contacting these ‘ready adults’ (those with prior college credits but no degree) is a crucial first step in drawing this population back to postsecondary education. In WICHE’s recent project, Non-traditional No More: Policy Solutions for Adult Learners (NTNM), six states (Arkansas, Colorado, Nevada, New Jersey, North Dakota, and South Dakota) looked at the different ways to reach these potential students. Despite differing state higher education structures, data systems, and state action plans, several common themes emerged from these efforts. While these promising practices emerged at the state level, the lessons learned can apply to institutions and other organizations as well.
1. Most states and institutions can mine their data systems to identify some, but not all, ready adults.
Almost all states involved in the Non-traditional No More project were able to use their higher education data systems to identify former students of public institutions who withdrew after earning a substantial number of credits. The process does not, however, capture those who migrated from another state, attended an out-of-state institution, or attended a private institution. While a state or institution may be able to identify all of its former students, there may be significant numbers of adults not captured in the data mining and would be receptive to other forms of outreach and communications efforts.
2. Lists of ready adults must be refined and filtered.
Once obtained from a state or institutional data system, the lists of ready adults must be cleaned and filtered. States and institutions discovered that some students identified through the data-mining process had earned degrees at other institutions or in other states. For example, consulting with the National Student Clearinghouse, South Dakota discovered that 23 percent of the adults with significant prior college credits identified in its initial data-mining efforts had enrolled in or graduated from another institution.
3. Obtaining accurate contact information is a major challenge, but private-sector partnerships are a cost-effective solution.
Although states and institutions have been able to identify adults with significant prior college credit, efforts to reach them directly have been hampered by a lack of current contact information. Many ready adults no longer live at the address they used while in college, making it difficult (and expensive) for states and institutions to carry out targeted outreach campaigns. Arkansas and South Dakota have worked with a private-sector data company that uses old addresses and student names to generate current contact information. This solution has proven remarkably cost-effective compared to other labor-intensive strategies while generating high match rates. Other states have reported similar results through partnerships with other state agencies (i.e., driver’s license bureaus), but sometimes those partnerships are difficult to broker.
4. Once adults with prior college credit are identified, data analysis can aid policymakers in addressing barriers.
States and institutions can also use the data-mining process to develop new policies based on in-depth analyses of the ready adults. In South Dakota, for example, the data showed a high number of former students had left difficult programs, such as nursing, after completing nearly all of the credits necessary for a degree. Based on the data, policymakers hypothesize that a significant number of students are not able to complete the internships, practicum exams, or other final requirements for these types of degrees, so they drop out due to the difficulty and extra time involved in starting a new major. To remedy this, the South Dakota system has now implemented “parachute degrees,” where students who will not be able to complete the requirements of a highly technical degree program will be able to transfer to a more general program and quickly earn a degree.
The South Dakota analysis also showed that these students perform well academically, helping to dispel the myth that they left because they could not successfully manage the academic rigor of postsecondary education. Instead, nonacademic issues proved to be a major factor in explaining why these students left.
Conclusion: Higher education data systems can help states and institutions identify adults with prior college credit and develop sound policies for eliminating barriers to their return. Mining this data to develop lists of former postsecondary students who left when they were close to earning a degree will be invaluable in developing targeted outreach efforts. Although adults who attended postsecondary education years ago may be difficult to locate, private data aggregation firms or partnerships with other state agencies can greatly ease this process and allow states and institutions to directly target these potential students for return. An effective data-mining strategy can be one tool in a broad effort to bring adults back to college and help boost overall degree completion rates.