New College is no longer solely an undergraduate institution. As the accredited Master’s in Data Science program welcomes its second cohort, campus is now home to undergraduate and graduate students alike.
How it began
Around three years ago, a person from Venture Capital suggested to Professor of Mathematics Patrick McDonald that New College should do data science — something no one else in Florida was doing thoroughly at the moment. With a background in mathematics, McDonald was hesitant about the suggestion. That hesitation changed the next day when someone in Tallahassee called President Donal O’Shea with similar advice. Realizing this was more than a coincidence, O’Shea asked McDonald to write a description for an undergraduate data science program, which eventually, after much talk, evolved into a graduate program.
In Jan. 2016, Forbes published an article called “Report: Why ‘Data Scientist’ Is The Best Job To Pursue In 2016.” In it, Forbes stated that the average salary for a data scientist is $116,000 and there are almost 2,000 job openings, according to a report by company review site Glassdoor.
“Eventually, we asked for money for a Master’s program, and we got it,” McDonald said. “That was an additional $750,000 per year—to the base budget—plus start-up costs for hardware and other things you need to run a data science program.” The Florida State legislature provided funding for the Master’s in Data Science, which goes directly the School’s budget.
Undergraduate impact
“That’s a decent chunk of change,” McDonald said about the funding. “It came earmarked with certain constraints. We said we were going to spend the money to hire faculty, and we were very explicit and clear that the faculty would spend half of their time teaching in the graduate program and half of their time teaching in the undergraduate program.”
“The idea was always that data science would somehow be directed at least as much towards the undergraduate program as the graduate program,” McDonald continued.
In fact, McDonald stated that there would not be a computer science undergraduate degree without the data science graduate degree’s funding. “I’ve been here 22 years, and since my very first days on campus, I’ve been saying, ‘you know, you really need someone to develop computer science.’ You can have all the correct ideas in the world, but none of them are worth anything until you have money to implement them,” McDonald said.
“Data science gives New College a chance to have a serious computer science offering, and that’s exactly what it’s got, and it’s got it because of that money,” McDonald concludes.
Program development
“Then we actually had to put together a program,” McDonald said. “[Professor of Computer Science David] Gillman had a lot of great ideas.”
Gillman and McDonald traveled and talked to people in data science, including some at Google and Amazon, asking what they were looking for when hiring those with education beyond an undergraduate degree—but without a PhD.
“We built a spreadsheet, essentially, a histogram—trying to figure out what was important to different players in the data science sector. From that, we pounded out a curriculum program and a market niche,” McDonald explained.
According to the college’s website, the Master’s in Data Science is “rigorous, practical, and focused on preparing recent college graduates for work in data analysis for research and industry.” The first year is centered on learning computer science and statistics skills for data analysis while the second year sees the students work on software-based projects with industry and research organization’s real data. During the last semester, students participate in paid internships at one of the program’s partner institutions. The Master’s in Data Science partners with major companies, such as Ancestry.com, Voalte, MobileBits, Star2Star and more.
Second-cohort data science student and alum Charlie Edelson ‘12 will be the first person at New College for six consecutive years.
“A lot of the focus of the data science program is getting people into industry and actually getting people jobs in the real world,” Edelson said.
Students have classes twice a week, from 9:00 a.m. to 5:00 p.m.
“It’s a cohort-based system, which means you enter with your group of buddies, and you all do the same classes in the same order,” Edelson explained. “You don’t move on until the next class until you finish the first class because everything builds upon on each other.”
“The cohort system is so one, you always have the foundation to move on to the next class, and two, data science in the real world—no one does data science alone unless you’re privately wealthy and have too much time on your hands because it’s a team effort,” Edelson continued.
“For anyone who is interested in the program, if you’re not willing to go head first back into it on day one, I would not recommend ending your thesis, graduating and coming straight back,” Edelson advised. “It is grad school. You have to know that if you’re going into the grad program you’re going to be in grad school, and a lot is expected of you. You’re going to have to behave like a graduate student.”
Accreditation
An application for accreditation was submitted to the Southern Association of Colleges and Schools Commission on Colleges (SACSCOC) on Sept. 15, 2015 and was accepted on Dec. 7, 2015. While accreditation was granted, SACSCOC needed to confirm that the institutional capacity for the program was satisfactory and visited campus on Aug. 30, 2016. “They came to make sure that the college could actually support the graduate program that they accredited,” McDonald explained.
The SACSCOC Site Visit Team delivered their final report at a Campus Exit Conference, held at Keating Center, on Sept. 1, 2016, in which they made one recommendation relating to assessment. That recommendation needs to be ameliorated by Feb. 7, 2017.
“In June, SACSCOC reviews their decision of Dec. 7, 2015 and decides to continue [the accreditation] or not,” McDonald said. New College’s accreditation for the program could be pulled if they find the review to be unsatisfactory.
The SACSCOC handles accreditation for degree-granting higher education institutions in the southern United States and operates under six core values: integrity, continuous quality improvement, peer review/self-regulation, accountability, student learning and transparency.
McDonald listed many names who helped with the extensive accreditation project, such as Suzanne Janney, Nikita Bagley, Julie Morris and Glenn Cuomo.
“That was the hard thing—the level change,” McDonald said about accreditation. “We were an undergraduate-only institution, and to add a data science program means that we actually have to change the level of the school. It has to change from undergraduate-only to a graduate-granting institution. It turns out that is a serious amount of work. It’s not just ‘let’s talk about data science’—it’s ‘let’s talk about lots of other things to support a graduate program in general.’”
The students
The graduate program’s inaugural class entered New College in Feb. 2016. Tuition and fees were waived for that entering cohort. Entering in Aug. 2016, the second class will pay tuition and fees set by the New College of Florida Board of Trustees. Seven students are in the inaugural class, and seven students are in the second class. The program is capped at 15 students per year to emphasize small class sizes and close student-professor relationships, two tenants that the college shares in its graduate and undergraduate programs. New College can also offer international student admission to the Master’s in Data Science.
Florida residents pay $474.33 per credit hour in the Master’s program while non-residents pay $1,169.47. This adds up to $5,691.96 per term and $11,383.92 per year for the former and $14,033.64 per term and $28,067.28 per year for the latter.
Edelson, who graduated from New College last year with a double degree in marine biology and physics, is enjoying the variety of backgrounds among the data science students. “Don’t let the lack of a computer science background dissuade you from participating,” Edleson said. “The majority of the students are math majors [previously].”
The data science students can retrieve keys to work in Palmer E’s “whale room,” as they call it, which is an area on the first floor large enough for many students to collaborate in comfortably at once.
“It creates a more distraction-free environment for us to work as a group,” Edelson explained. “It’s great to be able to have your own workspace, but walling off all by yourself is how you make your life very difficult in this program.”
The faculty
“We hired really, really good people the first year,” McDonald said.
Before Fall 2015, four new professors were hired for the program. One of those professors, Simant Dube, unexpectedly resigned for personal reasons late this summer.
“That was not easy,” McDonald said about needing to replace Dube. Dube’s absence left the Master’s in Data Science without one of its most quintessential components: a statistics instructor.
“In general, if you want to hire a statistician at a small liberal arts college— that is a many year process. And we’re getting a taste of that,” McDonald said.
The other professor in statistics, Visiting Instructor Melissa Crow, is currently working in the undergraduate program. “We had to quickly come up with a replacement for Simant for just this year,” McDonald continued.
The program is currently searching for two new faculty members.
This term, Professors McDonald, Matthew Lepinski, Gary Kalmanovich, David Gillman Michael Sutherland and Rebecca Wooten are teaching graduate classes, which include Statistical Inference for Data Science I, Topics in Statistical Inference for Data Science: Time Series & Forecasting, Data Munging and Exploratory Data Analysis, Data Storage and Retrieval, Practical Data Science, Algorithms for Data Science and Topics.
Data science across the curriculum
Workshops to introduce the interdisciplinary nature of data science and its impact in various disciplines across the curriculum were also planned.
“There are still efforts to hire people with interests that are broad enough, so that interests outside of Natural Sciences are developed and get attention,” McDonald said.
“The only discipline to be significantly negatively impacted by data science so far is mathematics because it’s taken me out of the rotation. Math has to give up resources in order for data science to run,” McDonald concluded.
“We really wanted something that other undergraduate institutions don’t have and really won’t have for a while because it takes so much time to marshal that many resources,” McDonald said about the Master’s in Data Science. “We were figuring that we’d be able to build something that distinguishes us amongst our peers—even our aspirational peers—which is something that would be incredibly valuable.”