Patrick Hu

Peter Rodgers

eLife assessment
This important study evaluates the outcomes of a single-institution pilot program designed to provide graduate students and postdoctoral fellows with internship opportunities in areas representing diverse career paths in the life sciences. The data convincingly show the benefit of internships to students and postdocs, their research advisors, and potential employers, without adverse impacts on scientific productivity. This work will be of interest to multiple stakeholders in graduate and postgraduate life sciences education and should stimulate further research into how such programs can best be broadly implemented.

anonymous

Reviewer #1 (Public Review):
The goal of this study was to determine whether short (1 month) internships for biomedical science trainees (mostly graduate students but some post-docs) were beneficial for the trainees, their mentors, and internship hosts. Over a 5 year period, the outcomes of trainees who completed internships were compared with peers who did not. Both quantitative results in terms of survey responses and qualitative results obtained from discussion groups were provided. Overall, the data suggest that internships aid graduate students in multiple ways and do not harm progress on dissertation projects. 'Buy-in' from mentors and prospective mentors appeared to increase over time, and hosts also gained from the contributions of the interns even in a short time period. While the program also appeared valuable for post-doctoral trainees, it was less favorably considered by post-doc mentors.
Strengths:
The internship program that was examined here appears to have been very well designed in terms of availability to students, range of internship offerings, length of time away from PhD lab, and assessments. 
Having a built-in peer control group of graduate students who did not do internships was valuable for much of the quantitative analyses. However, as the authors acknowledge, those who did opt for internships are a self-selected group who may have character traits that would help them overcome the potential negative impacts of the internship. 
The quantitative data is convincing and addresses important considerations for all stakeholders. 
The manuscript is well-constructed to individually address the impact of the program on each set of stakeholders, while also showcasing areas of mutual benefit. 
The discussion of challenges and limitations, from the perspectives of participating stakeholders, program leaders, and also institutions, is comprehensive and very thoughtful.
Weaknesses:
The qualitative data that resulted from the 'focus groups' of faculty mentors was somewhat difficult to evaluate given the very limited number of participants (n=7).
Overall, the data support the authors' conclusions with respect to the utility of internship programs for all stakeholders. As the authors note, the data relate to a specific program where internship length was defined, costs were covered by a grant or institutional funding, and there were multiple off-site internship hosts available. Thus, the results here may not replicate for other programs with different criteria.
This work provides a valuable assessment of how relatively short internships can impact graduate students, both in terms of their graduate tenure and in their decision-making for careers post-graduation. As more graduate programs are heeding calls from funding agencies and professional societies to increase knowledge about, and familiarity with, multiple career paths beyond academia for PhD students, there is a need to evaluate the best ways to accomplish that goal. Hands-on internships are valuable across many spheres so it makes sense that they would be for life science graduates too. However, the fear that time-to-degree and/or productivity would be negatively impacted is important to acknowledge. By providing clear data that this is not the case, these investigators have increased the likelihood that internships could be considered by more institutions. The one big drawback, and one that the authors discuss at some length, is the funding model that could enable internship programs to be used more widely.

Reviewer #2 (Public Review):
Summary:
The authors describe five-year outcomes of an internship program for graduate students and postdoctoral fellows at their institution spurred by pilot funding from an NIH BEST grant. They hypothesized that such a program would be beneficial to interns, internship hosts, and research advisors. The mixed methods study used surveys and focus groups to gather qualitative and quantitative data from the stakeholder groups, and the authors acknowledge the limitation that the study subjects were self-selected and also had research advisors who agreed to allow them to participate. Thus the generally favorable outcomes may not be applicable to students such as those who are struggling in the lab and/or lack career focus or supportive research advisors. Nonetheless, the overall findings support the hypothesis and also suggest additional benefits, including in some cases positive impact for the lab, improved communication between the intern and their research advisor, and an advantage for recruitment of students to the institution. The data refute one of the principal concerns of research advisors: that by taking students out of the lab, internships reduce individual and overall lab productivity. Students who did internships were significantly less likely to pursue postdoctoral fellowships before entering the biomedical workforce and were more likely to have science-related careers versus research careers than control students who did not do internships, although the study design cannot determine whether this was due to selection bias or to the internship.
Strengths:
1. The sample size is good (123 internships).
2. The internship program is well described. Outcomes are clearly defined.
3. Methods and statistical analyses appear to be appropriate (although I am not an expert in mixed methods).
4. "Take-home" lessons for institutions considering implementing internship programs are clearly stated.
Weaknesses:
1. It is possible that interns, hosts, and research advisers with positive experiences were more likely to respond to surveys than those with negative experiences. The response rate and potential bias in responses should be discussed in the Results, not just given in a table legend in Methods.
2. With regard to the biased selection of participants, do the authors know many subjects requested but were not permitted to do internships?
3. While the authors mention internships in professional degree programs in fields such as law and business, some mention of internship practices in non-biomedical STEM PhD programs such as engineering or computer science would be helpful. Is biomedical science rediscovering lessons learned when it comes to internships?
4. Figure 1 k, l - internships did not appear to change career goals, but are the 76% who agreed pre-internship the same individuals as the 75% who agreed post-internship? What percentage gave discordant responses?
Appraisal:
Overall the authors achieve their aims of describing outcomes of an internship program for graduate career development and offering lessons learned for other institutions seeking to create their own internship programs.
Impact:
The paper will be very useful for other institutions to dispel some of the concerns of research advisers about internships for PhD students (although not necessarily for postdoctoral fellows). In the long run, wider adoption of internships as part of PhD training will depend not only on faculty buy-in but also on the availability of resources and changes to the graduate school funding model so that such programs are not viewed as another "unfunded mandate" in graduate education. Perhaps the industry will be motivated to support internships by the positive outcomes for hosts reported in this paper. Additionally, NIH could allow a certain amount of F, T, or even RPG funds to be used to support internships for purposes of career development.

Development and Assessment of a Sustainable PhD Internship Program Supporting Diverse Biomedical Career Outcomes

Figures and data

Logistic Regression Model of Internship Participation, Number of Career Interests, and Trainee Type (Graduate student vs. Postdoc) on Identical Interest Match

Matches Between Career Interest and First Job Placement

Rates of Postdoctoral Training for Interns and Non-interns

Job Sector and Career Type of First Job After Graduate Student Internships

Type and Quantity of Publications Between Graduate Student Interns and Non-Interns

Be the first to read new articles from eLife