Holding the position of a provost, a chief academic officer, in a university has never been easy. The exhaustive responsibilities that a provost needs to handle, and handle well, may slightly differ from one institution to another but remain challenging nevertheless. The main responsibility of a provost is to make sound decisions regarding academics and research in a university. Mentioning the duties precisely, provosts are responsible for
- retaining students,
- balancing the revenue growth,
- attending and organizing meetings,
- planning the recruitment of top faculty, and
- developing curriculum and training programs.
A provost has to understand and maintain the fine balance between the university’s profitability and, student satisfaction and success, both of which are quite interdependent. And this where most provosts face their biggest challenges.
It is reported that out of the total enrolled students, only about half of the students come out of the university gates with a certificate. The tech billionaire, Bill Gates, puts forward his concern on this issue. Gates in his blog said,
“Based on the latest college completion trends, only about half of all those students (54.8 percent) will leave college with a diploma. The rest—most of them low-income, first-generation, and minority students—will not finish a degree. They’ll drop out.”
Now that’s a serious concern for provosts, isn’t it? However, the emergence of new technologies has enabled provosts to deal with the situation in a more equipped fashion.How? By leveraging predictive analytics in education, provosts can accurately identify students who are more likely to drop-out. Here’s how:
Step 1: Identify and define why your institution needs a predictive analytics model
Don’t follow the trend. Instead, first have a clear picture of why your institution requires a predictive analytics model. Conduct an internal analysis that involves brainstorming with all the important stakeholders in the university to discuss the issue of dropouts, and other similar issues, that your institution is currently facing. Once everyone involved in the academic and the administrative project agrees on the to-be-met goals, start strategizing on how you will accomplish the goals successfully. Can they be accomplished by incorporating solutions other than building a predictive model? If not, then build a vision of how you will build a predictive analytics model to achieve the desired outcomes.
Step 2: Build a flexible infrastructure that is tech-friendly
Building a predictive analytics model in education means creating changes in the current working style of the institution. Hence, provosts should ensure that all involved stakeholders are ready to face the changes that will follow. Stakeholders, such as faculty members, curriculum designers, and the dean, should understand how a predictive analytics model works and prepare to welcome the model on-board.
Step 3: Gather quality and unbiased data for predictive analytics in education
Once everyone gives a green signal, the next step is to start with the first big move of building upon the inputs for the predictive model. While students walk around the corridors of the university, they create and leave behind a trail of data. Data points like course enrollment fee, extracurricular activity fee, money spent on food, academic records, student engagement level, student goals, and other such comprehensive data prove helpful at this point. Comprehensiveness here implies that every minute data that can help the institution identify students who are at risk of dropping out, should be collected. But, collecting data on the current students only will not be not enough for the model to predict accurately. To build a robust model for predictive analytics for student retention, you are also required to gather historical data on dropouts and graduates. Ensure that you don’t collect biased data since it holds higher chances of giving discriminatory outcomes when shared with the model.
Step 4: Ensure that all stakeholders have a profound knowledge of predictive models
The next step is to provide the analytics tool with the collected information. To do so the right way, educators or curriculum specialists should have complete knowledge of how to operate a predictive analytics in education tool. Hence, provosts must strategize and schedule a training session on how to operate a predictive analytics tool.
Step 5: Undertake iterative progress checks on the model
Every day, every minute and every second, new data is generated. So, when this data is used as an input to the model, a new result (which is more accurate than the previous one) is produced. Hence, as and when new data is generated on a particular student, it can be given as an input to the predictive model, which will then provide a better, more customized prediction of the student’s performance in the coming months or years.
Step 6: Choose vendors wisely
Building a model for predictive analytics in education and incorporating it into the university’s existing infrastructure is not an easy task. Along with tech experts, provosts also need to arrange funds, build supportive infrastructure, and imbibe the right attitude in their team. So, what could be done? Universities can depend on vendors to get their work done. Many IoT and big data vendors in the market will provide good services for sure.
Instead of going through the numerous steps listed above, provosts can simply rely on a user-friendly educational analytics platform, like Talismatic, that comes with an easy to export data feature. The platform will enable provosts to:
- develop informed curriculums based on future skill demands,
- explain to students how their courses are rightly aligned with the job market demands, and
- infuse confidence in students by sharing insights on employment opportunities after graduation.
Education analytics platforms help universities to both, churn out employment-ready students and increase the enrolment rate.