A LITTLE UNDER two years ago, I sent a letter to various Arkansas companies asking what kinds of tech grads they were looking to hire, and what particular skills these employees would need to have. They were provided a series of choices to check or not check, plus a place where they could amplify their thoughts about the survey choices they’d made.

The purpose of this survey was to help my UA colleagues and me as we worked to define the University‚Äôs data science curriculum to ensure that our future grads would be trained in solving real-world business problems, instead of starting their working lives armed with merely a surfeit of theory. In May 2019 I wrote‚ÄĒin this very newsletter‚ÄĒabout the process of developing our curriculum. Now, as we begin the year in which our new classes will be rolled out en masse, I thought it would be useful to share some of the results of that business survey, to show just what corporate Arkansas is hoping to find in our coming tech grads.

First, a little information on the companies we queried. Although I‚Äôm based in Northwest Arkansas, home of such major companies as Walmart, Tyson Foods, and J.B. Hunt, our survey wasn‚Äôt limited to large companies or to any one geographic area. We covered the state, from Texarkana to Jonesboro, Eldorado to Fort Smith, and we surveyed small to mid-size companies as well. Start-ups were another focus, young companies with anywhere from half a dozen to a few dozen employees. This is a trend in Arkansas‚ÄĒstart-ups that have cropped up to do various types of analytics and data science work for more established companies, not just here but around the country and even the world. Such start-ups reflect the need for specialized knowledge that so far isn‚Äôt widely available, and these up-and-coming companies tend to be looking farther ahead than some of the bigger companies.

This balance of large, medium, and small reflects the balance on our Data Science Advisory Council, which was the point of the Council itself‚ÄĒto be sure we covered all the bases. To that end, we began this process by surveying the Council members about the kinds of technology jobs their companies were trying to fill, and the different skills those positions would require. From that input came a list of 35 unique technology position titles (including levels within titles) and a listing of the common skill needs received from the Council members. Then we simply asked the question: Which of these position titles are you needing to fill, and which of these skills is required for each title?

Of course, it didn‚Äôt turn out as neatly as that. Some recipients didn‚Äôt use the survey form, they just wrote about their own internal company workings and needs and I had to integrate that into our overall findings. Other recipients, especially in the large companies, didn‚Äôt necessarily fill out the survey themselves, or only themselves; instead, they routed it through their company hierarchy to get more feedback‚ÄĒincluding position titles that weren‚Äôt on our list.

From our standpoint as educators charged with preparing future data science grads, understanding the required skills of today’s workplace was key, and I think our survey succeeded in eliciting much critical information. Here’s what we learned:

• 36 position titles required ‚ÄúEvaluating the quality of data‚ÄĚ;
• 36 position titles required ‚ÄúUnderstanding and rigorously analyzing data using relevant software packages‚ÄĚ;
• 34 position titles required ‚ÄúWorking in a team-based environment‚ÄĚ;
• 32 position titles required ‚ÄúCommunicating findings in writing‚ÄĚ;
• 32 position titles required ‚ÄúCommunicating findings via graphical and visualization techniques‚ÄĚ;
• 31 position titles required ‚ÄúApplying critical thinking skills to solve novel challenges‚ÄĚ;
• 31 position titles required ‚ÄúData-cleansing, processing, and wrangling‚ÄĚ;
• 30 position titles required ‚ÄúGeneralizing knowledge from one subject area to another using data science‚ÄĚ;
• 30 position titles required ‚ÄúData science applied to business and economics in an organizational setting‚ÄĚ;
• 29 position titles required ‚ÄúCollecting data via research techniques‚ÄĚ;
• 29 position titles required ‚ÄúRelevant work or internship experience‚ÄĚ;
• 27 position titles required ‚ÄúApplying data science theories to understand the data and make predictions‚ÄĚ;
• 27 position titles required ‚ÄúCommunicating findings via public speaking‚ÄĚ;
• 23 position titles required ‚ÄúData privacy, security, and ethics‚ÄĚ;
• 21 position titles required ‚ÄúManagement of databases‚ÄĚ; and,
• 17 position titles required ‚ÄúProject management skills and leading teams.‚ÄĚ

*

I‚ÄôM OFTEN ASKED if any of these survey findings surprised me. The answer is no, because I‚Äôve worked in the business world. For many of these positions, I‚Äôve either been one or managed one or hired one, so to me these survey results also reflected my own personal experience. If there was any element of surprise, it was that so many organizations recognized that these so-called ‚Äúsoft skills,‚ÄĚ such as working in a team-based environment or communicating technological findings in writing, needed to be such a high priority in comparison to the purely technological skills. In my own business career, being able to do these deep, effective analyses, and then being able to explain them in a way that the non-technical people, or non-domain experts, or the people I worked for, could grasp it and use it‚ÄĒI felt that that was important. And I was pleasantly surprised that so many others saw it as important, too.

The curriculum that will debut at the UA Fayetteville system in September‚ÄĒand that we hope will eventually involve most of the 100+ two- and four-year colleges in Arkansas‚ÄĒowes much of its content to the businesspeople who took the time to give us their professional input. We used their knowledge and experience to build our actual courses, making sure that we teach the kinds of classes that either start with the desired skills or else create a set of problems and exercises that allow our students to build to those skills.

Our Advisory Council‚ÄĒthat group of committed business and educational leaders from Arkansas and beyond‚ÄĒhas been involved every step of the way, and we‚Äôve used their feedback to validate that we haven‚Äôt missed anything. In our last Advisory Council meeting (October 2019), we took them through the learning objectives and outcomes for the first two years‚Äô worth of courses, and got their feedback on that. In the next meeting, we‚Äôll do the same thing for the second two years of courses.

So yes, the participating companies were an integral part of this revolutionary process, and their input continues to resonate as Arkansas attracts more and more knowledge-based companies to locate here. While I sometimes give new companies the full survey to take‚ÄĒI recently did that with ACHI and DXC, which Governor Hutchinson just announced was moving a whole bunch of new jobs to the state‚ÄĒI often find myself just running the results of that two-year-old survey by the new companies, just to make sure somebody from somewhere else isn‚Äôt going to throw us a curve ball. Sure enough, they generally look over the findings and say, ‚ÄúYep, you got it. Those are exactly the skills we‚Äôre looking for.‚ÄĚ

While I‚Äôm sure we‚Äôll need to make changes and evolve as we go along, we feel good about the work we‚Äôve done to put Arkansas at the forefront of data science education in the U.S. In that effort, we consulted educators and business leaders from around the country, and one such leader‚ÄĒDr. Aric LaBarr, Associate Professor of Analytics, Institute for Advanced Analytics at North Carolina State‚ÄĒis a member of our Advisory Council. ‚ÄúI really appreciate how the University of Arkansas has created a bachelor‚Äôs level Data Science and Analytics education,‚ÄĚ says Dr. LaBarr. ‚ÄúThis space is filled with hundreds of graduate level degrees but only a handful of undergraduate level degrees. This unique offering provides a service that is not fully met in the U.S.

‚ÄúAlso, their approach to allowing many different flavors of the degree with concentrations through their ‚Äėhub and spoke‚Äô approach with many other departments is truly unique . . . which makes their degree truly interdisciplinary, just like the field of data science and analytics itself.‚ÄĚ