Effective Research-Oriented Internships to Foster the Next Generation of Statisticians

A JSM Session Sponsored by the ASA SPAIG Committee

By William Mietlowski, John E. Kolassa, Vladimir Geneus, Aiyi Liu, Wei Shen (Session Discussant), Ching-Ray Yu (Guest Discussant), and Kelly H. Zou (Session Organizer and Chair)

According to Merriam-Webster, the origin and etymology of “intern” comes from the middle French word interne, as well as from Latin internus. It has been known since approximately 1500 AD. Recently, more and more students seek internships, and thus, preparing for such opportunities becomes more common.

During a session at the 2016 Joint Statistical Meetings (JSM) in Chicago, Illinois, the ASA—through the Committee on Statistical Partnerships among Academe, Industry, and Government (SPAIG)—showcased a few projects derived from statistical internships across sectors. These projects represent partnerships with a focus on internships, which were established across sectors. Successful partnerships come in many forms, and in particular, statistical internships support the mission of collaboration, enhance both methodological and applied research, and may ultimately be stepping stones to fulfilling careers for a future generation of statisticians.

The four session presenters were William Mietlowski of Novartis Pharmaceutical Corporation, John E. Kolassa of Rutgers University, William Vladimir Geneus of Florida State University, and Aiyi Liu of the National Institutes of Health. Their presentations were elaborated upon by discussant Wei Shen of Eli Lilly and Company and guest discussant Ching-Ray Yu of Pfizer Inc. Kelly H. Zou of Pfizer Inc. and chair of the SPAIG Committee served as the session organizer and chair.

These statisticians from various professional sectors offered valuable thoughts and advice on how young statisticians, especially current graduate students and even undergraduate students, may seek and reap the benefits of internships via vital collaborations across sectors.

Below were a few questions raised and responses provided in the session. There are several examples of institutional collaborations between industry and academia or industry and the U.S. government. Some have resulted in the SPAIG award to recognize outstanding statistical partnerships.

Could the ASA help foster internship opportunities?

The ASA encourages undergraduate-level internships via its curriculum guidance and a white paper. The ASA Workgroup on Master’s Degrees recommends “internships, co-ops, and other significant immersive work experiences.”

For more than 12 years, the ASA has published lists of internship opportunities at no charge. Internships are posted on the ASA website and published in the December print issue of Amstat News and on the website STATtrak. The deadline for submitting internship opportunities for the December issue is October 20 each year. However, later submissions are posted online throughout the year as received. In 2016, 50 organizations posted internship advertisements (35 for graduate students only, while 15 also included undergraduate students).

For example, Novartis Oncology has used the ASA’s internship postings since October 2006. Global biostatisticians submit internship project proposals with clear business benefit each September. “Goodness-of-fit” to the projects is a primary criterion for internship selection. Novartis designs internship projects to be challenging, but feasible, strongly encouraging interns to present/publish.

Novartis employed 48 interns from 1,719 applications. Sixteen interns had 20 presentations (14 at the JSM conferences) and two publications. The company subsequently hired 14 interns as full-time employees. The ASA’s internship program is a “multi-win” endeavor.

Could master’s-level students, as well as international students, easily find summer statistical internship spots, and what are the typical scopes of such internships?

For example, Rutgers MS students generally find paid summer internships, although the process may not be easy. Many of these internships are found through informal student networks; others are found through solicitations made from employers through our program staff and career fairs. Rutgers offers a premium tuition MS program in financial statistics and risk management, which offers resources for internship placement. Rutgers also has a premium-tuition data science program that is too recent to have a record of summer internship placement.

International students are at some disadvantage for internships. One of the motivations for hiring an intern is to assess suitability for future employment. However, immigration barriers to such future employment make this hiring less desirable. Nonetheless, our international students seem to be able to reliably find internships.

These internships routinely involve data set cleaning, assembly, and management—and often, but not always, data analysis. Data management activities are many times a large part of the career of a future statistician, and internships reflect this. Of course, the temporary nature of this employment makes duties requiring long-term involvement and commitment inappropriate.

Internship opportunities also exist in federal government agencies such as the National Institutes of Health (NIH) and U.S. Food and Drug Administration (FDA). Requirements for citizenship or permanent residency vary across agencies; prospective international students with an F-1 visa should contact individual agencies to be clear about the requirements.

At NIH, hundreds of students come each year to gain research experiences through the Summer Internship Program. Internships at NIH usually cover a minimum of eight weeks, generally starting in May or June. The summer interns enjoy various activities NIH offers, including lectures featuring distinguished NIH investigators, career/professional development workshops, and Summer Poster Day. The application is available online from mid-November to March 1. Prospective students are also strongly encouraged to contact individual investigators they might be interested in doing their summer internship with.

As an example, the Biostatistics and Bioinformatics Branch in the Division of Intramural Population Health Research at the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) had 14 summer students—mostly graduate students—during 2012–2015. Under the supervision of an investigator, the students worked on an interesting statistical problem originating from studies conducted by the division in the summer. This often resulted in a publication in a biostatistical journal later. The summer internship experience helped the students pursue their further career objectives, such as employment or a higher degree.

What is some advice for current students who have no or very little work experience in an intense and short internship cycle?

One of the most overused statements in real estate is “Location, location, location!” A similar mantra to be considered for graduate students looking for internships: “Apply, apply, apply!” Employers understand that graduate students are seeking work experience. Therefore, those students who are in the first or final year of graduate school should apply to as many positions as possible.

Because preparing application material can be time consuming, it is wise to start this process early (many deadlines are in November). In recent years, statistics internships have been competitive due to a significantly increased number of students in statistics-related master’s and PhD programs. To increase the probability of getting an internship, students should become involved in activities (e.g., workshop presentation, mentorship, network connection, etc.) that increase their visibility.

If internships of interest—such as those listed in Amstat News—aren’t available, look for research opportunities at your school (e.g., statistics research in the college of medicine, college of engineering, college of behavioral sciences, etc.). Within those same departments, many graduate students will be in need of consulting for their thesis and dissertation. Local opportunities are great, as well. Look for short-term advertisements in local government (city and state departments). Finally, volunteering should never be overlooked. Doing some consulting for small businesses is also an option.

Once students start their internship in either a government institute or industry, they should know the needs and expectations of employers. In the pharmaceutical industry, for example, the statistical skill set is essential, but the most important component is to be a good team player, including communication and conversation.

One cannot accomplish projects without another team member, who is quite different from one in academia. So, the better impression for the internship employer, the bigger chance for the success of a possible future career in government or industry.

Are there communication barriers and benefits to students through internship roles?

While qualified students from established academic institutions are well trained in statistical theory and applications, a well-structured internship program offers an opportunity to gain real-world experience and develop collaborations with nonstatisticians. In a real-world setting, in particular, statisticians are often members of a cross-functional team. Strong technical and problem solving skills are essential, but effective communication is also a key factor for success. This includes collaborative work with colleagues in understanding scientific questions, defining research objectives, setting up a research plan, and communicating results that answer the original scientific questions.

What is SPAIG, and how can it help?

The ASA recognizes that the elements of statistical analysis can best be applied in conjunction with subject-matter experts. SPAIG’s mission is to identify, lead, and promote initiatives that foster partnerships between academe and business, industry, and government (B/I/G). The SPAIG aims to help the ASA member through the following activities:

• Establishing and making decisions about a partnership award to recognize outstanding partnership or collaborative efforts across different career sectors
• Promoting, featuring, and encouraging activities, such as organizing or cosponsoring scientific sessions or authoring statistical magazine and newsletter articles, that highlight a diverse set of collaborative case studies and successes
• Recommending and supporting systematic salary surveys of the ASA members from industry and government
• Periodically announcing the partnership award program and communicating the progress of initiatives and activities

The annual SPAIG award recognizes successful collaborations. In 2011, William Mietlowski wrote “The Summer Internship Partnership of ASA and Novartis Oncology: A Multi-Win Endeavor,” which has generated nearly 5,000 views. Every year, December’s issue routinely provides internship listings.

This JSM session was well attended and well received. It had wide appeal, particularly due to the interdisciplinary nature of statistical research and collaborations, as well as the emergence of complex problems within business or government. It was highly useful to junior researchers, applied statisticians, and statistical program directors to listen to these presentations while considering future statistical internship opportunities and collaborations. The authors do hope readers revisit the top 10 tips for successful internships and learn from a few great examples of students who have completed and benefited from them.

Now, let’s go to the list of the latest internships available and start applying for them! Find full descriptions for these internships on this site.

Editor’s Note
The views expressed in this article are the presenters’ own, and do not necessarily represent the views of their respective employers.