Predictive analytics brought the leadership team at Lone Star College-Montgomery together.
It began with the approval and trust from the college president. As the lead for the predictive analytics voyage, I assembled a data team by carefully selecting members from across the campus to serve. There were five elements that made us successful.
This article comes from the August/September issue of AACC’s Community College Journal. Read the entire issue online.
Someone who is passionate about analyzing data and understands the value of data to the institution would be considered the champion. This person will act as a coach to seek out data warriors, organize a team and develop the atmosphere conducive to analytics. As a coach, he or she will advocate for others to have access to data, assess the data needs of the college and disseminate results. The champion also will be a resource and be accessible to coach others with their data needs.
Like all great champions, one needs to align with a team of great data warriors. These data warriors then receive the champion’s interpretation of the data being analyzed and begin creating action-oriented initiatives (i.e., nudges, campaigns utilizing social media and text messages) to improve student success.
Team members from across the college were carefully selected and invited to serve. It is a six-member cross-functional team with an academic dean, faculty, director, student service deans, analyst and the vice president of student success.
In order to take advantage of the data analytics, the team must be fast at producing action-oriented initiatives. Campaigns must be carried forward after three to four meetings.
Once the data are disseminated to team members, a meeting is called. The meetings should be no longer than an hour in duration. This keeps the team functional and action-oriented. Each campaign is focused on enrollment, retention, persistence and/or graduation.
When implementing action-oriented initiatives, the Data Team sought collaboration from various departments. For example, an association rule-mining analysis was conducted on the courses in which students enrolled over a three-year period. Association rule mining analyzes variables that were clustered together. The results indicated that students who enrolled in English courses also enrolled in history courses during the same semester; students who enrolled in biology courses also enrolled in the student success course.
The Data Team organized a meeting with the academic deans from across the college to reveal the results. The outcome of the meeting led deans to work together when scheduling courses to align courses and times. As a result, there has been a 30-percent increase for students enrolling in 15 credit hours.
One of the most revealing actions for the Data Team is how transparency of the data and results transformed the culture at the college. The Data Team analyzed students who were predicted not to persist and were enrolled in the student success courses.
The results were shared with the advising team. During the meeting, an initiative was launched and students were nudged by advisors using email and text messaging. Students quickly responded.
At the end of the semester, the Data Team scheduled a meeting with advisors to reveal the results. They were ecstatic; but, most importantly it was confirmation to the advisors that they made an impact on student success. By the end of the meeting, the advisors generated their own ideas to launch their next campaign using predictive analytics.
The culture at Lone Star College-Montgomery shifted to become data and action-orientated. In an effort to expand use of predictive analytic initiatives to other Lone Star Colleges, the first Data Camp event was organized. It is a conference-type event that teaches others how to use data, predictive analytics, student nudges, social media and use tools (Civitas, PowerBI). The event was for all college employees and was so successful that another one is scheduled for this month.