Fundamentals of Data Science for Future Data Scientists

 

4. Data Science Education: Current Data Science Programs and Design Considerations

The increasing demand in Data Science workers has motivated rapid growth of Data Science related programs, as presented in this section. Universities and colleges have been actively developing new Data Science programs and courses at different levels to meet the needs of the job market on qualified Data Science workers. It is necessary to review and summarize what has been established so far in order to develop high-quality Data Science programs.

 

4.1 Data Science Programs Overview

As we have discussed earlier, Data Science is an interdisciplinary field. It integrates and applies knowledge and techniques in multiple disciplines and fields. Data Science programs and courses have been established to meet the needs for training effective data scientists. These programs offer a variety of courses to train students with the research and professional skills to succeed in leading edge organizations. This section summarizes existing Data Science programs and trainings.

 

4.1.1 PhD Programs in Data Science

According to a Data Science salary study, 88% of data scientists have advanced degrees and 46% have a Ph.D. (Burtch Works Executive Recruiting, 2017). As of June 2017, the website MastersInDataScience.org lists 18 universities offering doctorate-level programs in Data Science and Data Analytics, as showed in Table 3.

The majority of these PhD programs require students to have undertaken technical coursework or earned a technical degree in Computer Science, Mathematics, or Statistics as a pre-requisite for entrance. These programs are on-campus and require students to be present in class to complete their degrees, only a couple of programs offer online classes but still have the requirement of attending campus activities in order to obtain the degree.

 

4.1.2 Master Programs in Data Science

The same Data Science salary study also mentioned that 59% of early-career data scientists’ highest degree was a Master’s, a significant increase from 48% in 2015 (Burtch Works Executive Recruiting, 2017). As of June 2017, the site MastersInDataScience.org lists 23 schools offering Master’s programs in Data Science, as showed in Table 4.

These Master’s programs have a duration from nine months to about two years, depending on curriculum requirements and the part-time or full-time status of the students. Students usually begin by taking required courses specified in the curriculum in the first semester. Once they obtain basic knowledge and become familiar with technologies in Data Science such as programming, databases, and statistics, they proceed to do practicums, application, and capstone projects. Some of the schools provide internship opportunities and practicums so students can work with partner companies in industry or government agencies during their study.

 

4.1.3 Graduate Certificates Programs in Data Science

Certificate programs provide a flexible way for people to study Data Science. A certificate program generally takes less time and cost less money than a program intended for earning a degree. It is an attractive option for people who are interested in pursuing a career in this field. The website MastersInDataScience.org has a list of 92 universities offering graduate certificate programs in Data Science and related specialties.

 

4.1.4 Massive Open Online Courses (MOOCs)

MOOCs are online courses with open (mostly free) access via the Internet. These courses are similar to university courses, but do not tend to offer academic credit. They allow students to choose their own academic path and complete the courses on their own schedule. To study Data Science or data analysis, students may choose courses from MOOCs – Coursera, for example, and construct a learning plan primarily from the following sample areas:

• Programming languages: The most popular languages currently used are Python and R for Data Science courses

• Statistics and Probability: There are broad theories that are used for making informed choices in analyzing data and for drawing conclusions about scientific truths from data.

• Data collection: Extracting data from the web, from APIs, from databases, and from other sources

• Data cleaning and managing: Manipulating and organizing the data collected to make data set useful for data analysis tasks

• Exploratory data analysis: Exploring data to understand the data’s underlying structure and summarizing the important characteristics of a data set

• Data Visualization: Using plotting systems to construct data graphics, analyzing data in graphical format, and reporting data analyze results

• Machine Learning: Building prediction functions and models by using supervised learning, unsupervised learning and reinforcement learning

 

4.1.5 Bootcamps

Bootcamp programs are non-traditional educational paths. Compared with traditional degrees, these programs are intense and have faster routes to the workplace. It is another education option for considering a career as a data scientist. Data Science Bootcamps provided by Data Science community (Datascience community, n.d.) has a list of bootcamps available for Data Science and data engineering.

In summary, Data Science programs at undergraduate, master, and doctoral level have been developed in the U.S. These programs provide a wide range of choices to students who want to obtain knowledge and skills in Data Science. Still, more Data Science programs are being developed. It maybe the time to explore the characteristics of a competitive Data Science program.

 

4.2 Data Science Program: An Integrated Design

Based on our understanding of Data Science and its related concepts, the knowledge and skills required for Data Science workers, and current Data Science programs, we believe that a high quality Data Science program should provide courses and training in the following areas:

• Fundamental concepts, disciplines, and the profession. The program should offerl one or two courses to introduce the student basic concepts, disciplines, and the profession of Data Science. These courses set up solid foundation for students to learn more advanced concepts and applications. They may teach not only concepts and characteristics of data, information, knowledge, and data life cycle, but also related mathematical concepts, functions, models, and theorems. Students should develop an affection to data and willing to do Data Science;  

• Statistical analysis and research methodology. The program should offerl multiple courses to teach data collection and data analysis skills, which are usually taught in master or doctoral level research methodology classes.  

• Programming languages, algorithms, and tools. Courses such as computerl algorithms, programming langauges, data mining, database design, and machine learning help students to apply computational techniques for data collecting, cleaning, analysis, prediction, and presentation.  

• Business logic and soft skills. Students need to understand the purpose of Datal Science, they need to learn how to communicate and collaborate with different business units such as sales, marketing, and production. Also, learn to effectively present data analysis results is also important.  

• Practicum or a capstone project. Data Science students should do at least onel project that integrates what they have learned to address a challenging real[1]world Data Science problem. The ideal case would be to have the student work in a company or an organization to participate in real Data Science projects that may help the organization to make better business decisions.

Each student can choose different courses or focus on different areas to improve their knowledge and skills based on their backgrounds and personal interests. For example, a student with business background may want to take mainly courses on data management and programming languages, while, an Information Science student should study business logic in addition to statistical analysis and data visualization through coursework or the practicum.

A course may also teach student knowledge and skills in different areas. For example, a database course may teach students data collection, cleaning, analysis and report writing through its term project. The instructor can also design projects that of research values or reflecting real industry needs.

In general, a flexible curriculum and close connections with profit and non-profit organizations provide students much flexibility and opportunities to become successful data workers.