Data science can be defined as the process of analyzing, communicating, capturing, and maintaining vast amounts of data.
In fact, we might describe data science as being a five-part cycle of information. Data is mined, modeled, and summarized in the process phase of data science. In the second step, data scientists analyze the data by conducting predictive analyses, regressions, and quantitative analyses. Third, data scientists communicate what they’ve found. This includes data reporting and visualization. The next step in the process is to capture data. Data capture involves data acquisition and entry, signal reception, and data extraction. The process continues in the fifth step, maintenance. During this phase, data is cleansed, processed, and warehoused.
This is an ongoing process, too, given how much data is produced today and how quickly it is produced.
As a result, data science is a profession growing as fast as our passion to generate and consume information. It evolved to meet the demand to collect, coordinate, and analyze all the information being created by individuals, groups, organizations, and companies.
Data science has come to higher education and is creating a huge, quiet revolution that will benefit students, faculty, and the institution’s bottom line.
The Need for Data Science
The birth of the data science profession was inevitable as the amount of information created by humankind has exploded over the past few years.
In social media alone, these are generated every minute of the day:
- 147,000 new photos are uploaded to Facebook
- 347,222 Instagram stories
- 41,666,667 messages shared on WhatsApp
- 479,452 people engage with content on Reddit
What’s more, there are 7,000 tweets on Twitter each minute about TV shows and movies alone. All told, Twitter averaged 474,000 new tweets per minute in 2019.
Additionally, Microfocus notes that each minute, 400 hours of new YouTube videos are created, 3 million Facebook posts are made, and 4 million text messages are sent in the United States alone. That’s a lot of data, and that’s only looking at a small part of one segment of data that’s created each day – social media.
For years, tech pioneers like Google, Amazon, Netflix, and others have been using big data to help us find what we like or want — and to sell us things we didn’t know we needed or wanted.
For decades it’s been standard marketing practice to cluster people in the same age groups or with the same interests and promote products that they might like based on these demographics.
So, in years past, companies would develop marketing strategies for defined age groups, like the coveted 18-49 demographic. And while marketing techniques are still developed based on these general demographics, data science enables marketers to drill down even further.
For example, the huge amount of data being collected every nanosecond enables companies to read you, the individual, like a book, a detailed book, with tons of salient footnotes. They use data like what ads you’ve looked at online, what type of device you’re using to access the internet, location data, what you search for on Google, and so forth.
Additionally, the cookies that websites use generate even more detailed information about you. This includes who you are, where you’ve been, who you communicate with, and things you’re interested in.
Needless to say, data science has cut its teeth in the realm of social media (and many other applications), which makes it a service that is unspeakably valuable. This includes great value for institutions of higher learning.
Why Higher Education Needs Data Science
With a better understanding of what data science is and how it can be used to sort, organize, and analyze data from social media, we can now begin to develop an understanding of how it can be used in other areas, like higher education.
Today, colleges and universities face increasingly sophisticated competition as well as increasing personnel and operating costs. As a result of this, more and more institutions are adopting data science to decrease their margin of error in doing business and helping predict the future.
Any institution offering on-campus and online courses is competing in a global market. To survive and thrive, institutions of higher education need to use big data to:
- Find students: Better data will enable institutions to market their programs to the students suited to each academic degree program. This not only reduces the time and expense of pursuing students that aren’t well-suited to their programs, but colleges can also concentrate on helping students enroll in the right academic program for their specific knowledge, skills, and academic abilities.
- Attract students: Data science will enable institutions to deliver the right talking points to those students that might be interested in attending their school. Much like the ads you see when you browse the internet are tailored to you based on your previous browsing behavior, colleges and universities can tailor their messages directly to students that might be considering attending their school.
- Retain students: With the right kind of data, institutions can allow students to personalize their educational experience. In doing so, colleges and universities enable students to pursue coursework that interests and engages them. A direct consequence of that is that educational institutions thereby improve student retention through graduation.
This is a critically important point. For most institutions, recruiting and retention are the two biggest challenges in the higher education business. Schools that are successful in recruiting students bring in more money in terms of tuition and fees, thus helping the school’s bottom line. Furthermore, the more likely students are to be retained through graduation, the more money the school will receive in the form of tuition and fees, room and board, and so forth.
Data Science is Changing the Business of Higher Education
In addition to improving the manner in which educational institutions attract and retain students, it is also providing new ways to conduct business.
At a growing number of colleges and universities, admissions officers are making faster and more precise decisions on who to admit based on big data. They’re using key performance indicators such as academic history, demographics and geographical data to predict which students are more likely to enroll if they are accepted, thus cutting marketing and advertising costs.
There is also a growing trend of admissions offices looking at a potential student’s use of social media to collect more information. (Remember the social media stats above?) This real-time information gleaned from a student’s content and photos can give admissions officers insights into a student’s character, habits, discipline and potential for success.
There has been debate about whether this kind of analysis of students’ personal information is ethical. On the one hand, posts to social media are out there for the public to see, so it’s not like admissions officers are doing their best detective impression to “dig up dirt” on applicants – it’s all out there for everyone to see.
But on the other hand, what kids post online doesn’t necessarily indicate who they are in real life, let alone what kind of student they might be. One could argue that exams like the ACT and SAT are far more predictive of a student’s potential academic success in college than what they post to Instagram and Facebook. However, all that data is out there, and educational institutions now have the ability to mine it thanks to data science.
Data Science Can Help Monitor Student Success
Once a student is accepted to a college or university, big data analytics can monitor that student and intervene with help if he or she is having problems. This helps ensure a student’s success in the classroom and helps the institution’s bottom line.
When a student drops out of college, several areas of the campus are affected, from dining to housing to sports. And, as admissions officers know, every student who drops out adds to the number of new students that need to be recruited. So, better retention cuts an institution’s operating expenses in many ways.
Big data has also given rise to predictive analytics which combines and analyzes information from several sources to make predictions about the future and perhaps even shape the future.
Some institutions are using predictive analytics to:
- Advise students to help them succeed academically, thus increasing academic success and improving the likelihood of retention to graduation
- Identify and advise students who are at-risk academically, socially or financially so interventions can be devised and implemented
- Build models to more accurately forecast enrollments
Institutions are also learning that letting first and second-year students fend for themselves is also a bad business model. Tracking and analyzing student behavior and performance – and stepping in to help students in trouble — will benefit students personally and professionally as well as lend to the institution’s business success.
Data Science is Changing Course Offerings
Most higher education professionals are familiar with Blackboard Inc., an educational technology company known for its pioneering Blackboard Learn. The company’s researchers are continually conducting learning analytics research on a huge scale. One of the latest developments is the quantum jump of categorizing types of courses into five “archetypes.”
The ability to classify courses by these five archetypes enables faculty and departments to identify course patterns that provide the greatest impact for different populations of students.
According to Timothy Hartfield in his April 2017 Blackboard Blog: “Instructional designers and instructors can work together to fit course patterns to learning goals, engaging in evidence-based conversations about teaching methods in a way that, until now, was all but impossible.”
Blackboard’s information that led to the development of archetypes was based on a sample that included data from 3,374,462 unique learners in 70,000 courses at 927 institutions. That type of sample would have been impossible a few years ago. But, because of advancements in data science, all that data can be gathered, analyzed, organized, and interpreted with relative ease.
What the Future Holds for Data Science and Higher Education
Unfocused marketing to a large and varied audience is both lumbering and expensive. It’s also becoming a technique of the past. It is being replaced by smaller, more sharply focused campaigns that result in a larger yield for a smaller investment. This gives educational institutions a significant edge in an increasingly competitive field.
This is especially important because of the steady decrease of high school populations and international students. With fewer kids electing to go to college, it could sharpen the competition for students into something close to cutthroat. Institutions with slow, creaky bureaucracies and leaders cautious of change will find themselves shrinking or even ceasing to exist while those that embrace modern methods of using data-driven strategies to recruit students will thrive.
For the most part, higher education has existed the last few centuries using the lecture model. The professor stood at the front of the class and imparted knowledge as students dutifully took notes. Yes, labs and discussion groups came on board in the mid-20th century, but central to everything was the professor lecturing, giving tests, and handing out grades. If a student couldn’t cut it, he or she could leave.
In recent years, though, professors have faced classrooms in which every student is carrying a mobile device – the product of and contributor to data science. Armed with mobile devices, college and university students have the ability to access an untold amount of data on any matter of subject. While professors still hold the title of “expert,” the ability to access library materials, journal articles, and other academic sources from the palm of one’s hand has made the lecture model less efficient.
In fact, it could be argued that data science will be the death knell of the lecture as a universal educational delivery model. The institutions that incorporate data science into their business models are much more likely to succeed in an environment that demands more efficiency in all aspects of higher education business, from recruitment to retention to how academic courses are delivered to students.
Developing ways through data science to attract students, retain them and help them succeed is vital to the institution’s bottom line. And in the last couple decades, institutions have discovered what corporations have always known: the bottom line is everything.
Data Science Degree Programs Staff
Updated May 2021