Abstract
A sizable body of literature has examined faculty vitality in academic medicine. Tenure-track faculty with a research concentration face high levels of stress (Dankoski, Palmer, Nelson Laird, Ribera, & Bogdewic, 2012), and junior faculty in particular are highly susceptible to discontent. Few studies have examined faculty vitality and research development in the occupational therapy profession (Paul, Stein, Ottenbacher, & Liu, 2002). Vassantachart and Rice (1997) surveyed occupational therapy faculty members about faculty development practices, collegial support, and choice to pursue academia. The results indicated that respondents had entered the field because of a desire to teach. Respondents also reported not taking the time to perform scholarly work or attend research seminars. In a different, qualitative study, occupational therapy investigators were interviewed about the development of their research programs (Case-Smith, 1999). Themes identified included completing a research-based doctoral education, finding a mentor, establishing collaborative relationships, and accepting criticism.
In a study of the benefits of academic mentorship, Paul, Stein, et al. (2002) found that mentored junior occupational therapy investigators had greater research productivity than those who were not mentored. Conference presentations were the most common form of productivity; 70.8% of junior investigators had not yet published in a peer-reviewed journal (Paul, Stein, et al., 2002). Paul, Liu, and Ottenbacher (2002) surveyed occupational therapy faculty about research productivity, institutional factors that influence productivity, and perceived research competence. They found that perceived research competence was significantly lower in junior faculty compared with senior faculty.
Since 2002, the research funding climate has become substantially more competitive. For faculty, establishing and sustaining an independent, externally funded research agenda is daunting. Since 2009, the number of research applications submitted to the National Institutes of Health (NIH) has increased while the award success rate (percentage of applications that receive funding) has decreased (see Table 1; U.S. Department of Health and Human Services, 2015). Furthermore, in part because of the sequestration (mandatory budget cuts ordered in 2013), the NIH cut $1.55 billion in funding with the net result of 700 fewer competitive research grants awarded in Fiscal Year (FY) 2013 than in FY 2012 (U.S. Department of Health and Human Services, 2013), a 41% decrease from 2004 calculated in constant 1995 dollars (correcting for inflation). This cut brought NIH funding down to the level of investment in 1997 (Bourne & Lively, 2012). The award success rate at NIH improved in 2014 and 2015 but is still below 2009 levels. Similar cuts have occurred across other funding streams, such as the Department of Defense, the National Science Foundation, industry, foundations, and support from philanthropic organizations.
Six-Year History of National Institutes of Health Research Project Grant Application Dispositions, 2009–2015
In this article we report the results from an annual evaluation of our academic medical center’s research mentoring program for junior faculty who aspire to be independent, funded investigators. The program includes junior scientists from the Colleges of Health Science (CHS), Nursing (CON), and Medicine (COM). We asked the mentees whether they had considered quitting research in the past year. We present factors that were associated with thoughts about quitting research and propose strategies for intervention and retention.
Method
Participants
An interdisciplinary faculty research mentoring program was developed by Rush University in 2006 to reduce barriers faced by junior faculty seeking research independence. The program facilitates individual mentoring, team mentoring, and peer mentoring–based writing accountability groups (Foucher & Skarupski, 2014) and provides a comprehensive range of resources such as statistical consulting, editing, database management, and grants administration. The program has three translational research tracks: (1) clinical, (2) laboratory, and (3) epidemiological. To enroll, mentees must be guaranteed at least 20% protected research time; however, mentees typically report spending much more time on their research—a median of 25 hr/wk (mode = 40). The program has been successful, as measured by ≥$43 million in external grant funding and ≥350 peer-reviewed publications by mentees since their admission to the program (Skarupski & Keshavarzian, 2013). Of the 63 mentees enrolled in the program, 44 responded to the survey.
Evaluation Tool
An annual program evaluation is conducted that includes assessments of the program mentees. The Rush University institutional review board approved the study. Participants provided written informed consent. We used Rubio et al.’s (2011) Comprehensive Model for Career Success in Physician–Scientists survey instrument, which includes items that measure institutional resources, research training experiences, conflicting demands, and relational factors (i.e., mentoring, networking; Lee et al., 2012; Rubio et al., 2011). Some of the items in the instrument and the accompanying response categories are listed in Table 2.
Number of life events was a sum of 14 possible major life events (e.g., birth of a child, serious personal illness or injury, marital separation).
Creativity was measured by responses to one item.
Work preference was measured by eight questions.
Passion and interest was measured by three questions.
Leadership was measured by 12 questions.
Professionalism was measured with seven questions.
Research experience was measured by one question.
Mentoring structure was measured by four questions that asked the participants to rate the perceived effectiveness of their primary mentors.
Experiences with the primary mentor was measured by means of five questions that asked the participant to rate his or her mentor (e.g., accessibility, demonstrating professional integrity, providing useful critiques).
Networking/social capital was measured with three questions.
Job satisfaction was measured by seven items.
Career satisfaction was measured by one question.
Burnout was measured by one question asking participants to select the item that best described their experience with burnout in the past year.
Career goal was measured by one question asking participants to select, from a number of items, their ideal career preference.
Life satisfaction was measured by means of five questions.
Bivariate Analyses Comparing Mentees Who Had Considered Quitting Research in the Past Year With Those Who Had Not Considered Quitting
Note. In some categories, n may not total 39 because respondents selected more than one choice or left the item blank. Respondents who did not indicate a number of dependents were treated as having 0 dependents. All empty cells under p are not significant (p > .05). M = mean; SD = standard deviation.
The Clinical Research Appraisal Inventory–12 (CRAI–12) is a 12-item tool to determine how confident respondents are in performing certain tasks, such as selecting a suitable topic area, determining sample size, designing the best data analysis strategy, interpreting data, writing the Results section of a paper, presenting data, fostering collaborations, communicating to a research team, obtaining informed consent, and determining appropriate funding mechanisms (Robinson et al., 2013). Respondents completed the CRAI–12 as part of the total survey.
Statistical Analysis
We used Student t tests and χ2 tests of association to compare mentees who had considered quitting research with those who had not. We then used logistic regression to model the odds of quitting research. Based on significant associations at the bivariate level, the model included the following covariates: sex, mentoring effectiveness, mentoring experiences, and CRAI–12 score. A p value of .05 was the criterion for statistical significance. All analyses were performed using IBM SPSS Statistics (Version 22; IBM Corp., Armonk, NY).
Results
Of all mentees (respondents and nonrespondents) in the program (n = 62), the majority were from COM (n = 54; 87.1%); 6 were from CHS (9.7%), and 2 (3.2%) were from CON. Forty-seven were in the clinical track, 13 were in the laboratory-based track, and 2 were in the epidemiological track. There were 43 assistant professors, 5 associate professors, 1 clinical instructor, 4 instructors, and 9 fellows or residents. Of mentees in CHS, 3 were from speech, 1 from nutrition, and 2 from occupational therapy. Although occupational therapy mentees were only 3% of all mentees, the total number of occupational therapy department faculty in this university was smaller than most (n = 6, including mentees and chairperson).
Of the 62 mentees enrolled, 44 responded to the survey after three attempts (71% response rate). Participants were mostly female (61%) and married (68%) and had a mean age of 38 (standard deviation [SD] = 6.5). Participants allocated a mean of 57.4% (SD = 32.2) of their effort to research and 40.3% (SD = 26.8) to clinical duties. When asked “In the past year, have you considered quitting research?” 39 mentees responded to the question, and 17 (44%) answered in the affirmative (see Table 2).
Compared with their counterparts, participants who had considered quitting research had significantly lower means on several survey items, including the CRAI–12 total score, job satisfaction, and career satisfaction. Also, the group that considered quitting had higher scores on burnout (52.9% reported “I am definitely burning out,” compared with none of those who had not considered quitting research; see Table 2). There were no differences between groups on marital status, number of dependents, life events, work preferences, passion, professionalism, networking, social support, life satisfaction, creativity, or reports of mentor effectiveness or experiences. We tested associations among sex, mentoring effectiveness, mentoring experiences, and CRAI–12 scores with thinking about quitting research in a logistic regression model (data not shown). We found that only CRAI–12 scores were significantly, inversely associated with thinking about quitting research (odds ratio = 0.44, 95% confidence interval [0.22, 0.85]).
Discussion
We have reported the results of a survey administered to junior faculty members who were part of a mentorship program. Of those survey respondents who answered the specific question about quitting research (n = 39), nearly half had thought about quitting research in the previous year. Factors associated with thinking about quitting included lower confidence in one’s research skills, reduced job satisfaction, and higher burnout. Mentoring factors were not associated with thinking about quitting. Because our survey was confidential, we were not able to ascertain whether the mentees who had thought about quitting judged their research skills more harshly as a consequence of not being awarded funding or whether, in fact, their skills were not up to par.
What is particularly disturbing about these results is that these junior investigators represent a highly motivated and self-selected group of faculty who voluntarily enrolled in a program designed specifically to support the development of their research careers. What are the implications for junior investigators whose research is conducted in isolation, without access to research infrastructure or to research mentors? In 2010, only 16% of all occupational therapy programs in the United States were at institutions classified as Research 1 universities (very high research activity) according to the Carnegie Classification of Institutions of Higher Education (2016). Research 1 universities generally provide optimal infrastructure for research, and the majority of occupational therapy junior faculty do not work in these institutions. Also, investigators in nursing and allied health fields have ample opportunities for teaching positions, which provides yet another disincentive for pursuing a research career in occupational therapy.
The increasingly competitive funding climate makes a career in academic research less palatable. We propose the following intervention strategies: (1) emphasize development of transdisciplinary science teams, (2) implement multilevel mentoring of all types (e.g., traditional mentorship, sponsorship, coaching), (3) cultivate leadership skills in present and future departmental leaders, and (4) train junior faculty to be resilient.
Transdisciplinary Science Teams
The formation of team science is expected to inform the scientific community of innovative methods of facilitating multilevel research (Meyers, Begg, Fleming, & Merchant, 2012). Thus, the facilitation of authentic, transdisciplinary collaborative models should be established. Team science requires people with diverse but specialized areas of expertise to collectively determine questions that will facilitate high-impact work from numerous perspectives (Börner et al., 2010). This effort necessitates training scientific leaders who can bring together diverse groups of researchers who use dissimilar research methods (Gray, 2008).
At present, medical, basic science, and allied health students are trained in a discipline-specific manner. However, schools are now charged with providing more interdisciplinary education so that students might experience collective knowledge across specialties (Gray, 2008). Therefore, trainees may become adept at the integration of multilevel questions and produce groundbreaking work. Also, junior investigators would have built-in mentoring teams with whom they could share workload, authorship, equipment, and staffing resources.
Multilevel Mentoring
In this study we found no association between mentoring and thoughts of quitting research; however, mentees were asked to rate only the effectiveness of and experiences with their mentors, not their satisfaction. The value of mentoring in terms of faculty research productivity has been documented in the literature (Eby, Rhodes, & Allen, 2007; Krause, 2007; Paul, Stein, et al., 2002; Skarupski & Keshavarzian, 2013). Multilevel mentoring refers to the wide variety of mentorship at all junctures in a faculty member’s career; that is, mentorship throughout the various stages in an academic career (instructor to professor) using the various types of mentoring (traditional mentorship, sponsorship, coaching). We found no association with mentorship structure and intent to quit research, which may be attributable to the small sample size.
The American Occupational Therapy Foundation (2015) has strongly recommended that junior investigators find mentors, specifically those with a track record of federal funding. Mentors help their mentees establish an independent research career by narrowing their focus; helping articulate a research agenda; identifying strengths and weaknesses in their research training; identifying funding mechanisms; providing feedback on abstracts, manuscripts, and grant applications; and continuously reevaluating and strategizing next steps in the mentee’s career. Mentors may also be selected to serve a psychosocial function of providing social and emotional support.
In the nursing literature, Jacelon, Zucker, Staccarini, and Henneman (2003) described an interesting peer mentoring arrangement consisting of experienced nurses, all of whom assumed tenure-track positions at a Research 1 university. Group members reported an increase in scholarship (individual and collective) and improved mutual expertise.
Cultivation of Leadership Skills
The average tenure of department leadership has declined (Rayburn, Alexander, Lang, & Scott, 2009), likely because of a shift in job priorities, which now emphasize middle-management skills rather than academic ones. New responsibilities include fundraising, more frequent evaluation of educational outcomes, and management of a large pool of part-time adjunct faculty (Williams, 2013).
In Rubio et al.’s (2011) overall survey, mentees were asked whether they perceived their chairperson as supportive of their research career, and 59% strongly agreed. Paul, Stein, et al. (2002) found that chairperson or dean support was an institutional factor positively related to research productivity in junior occupational therapy investigators. We propose that the support of the chairperson is vital, but how is a chairperson’s support of faculty research manifested? One issue that merits a chairperson’s consideration is asset allocation and department structure, which can be a challenge because of the budget constraints that many chairpersons face. However, when faculty members depend on soft money (salary support from research funds), a dynamic characterized by a lack of collegiality and loyalty among these faculty members develops (Bourne, 2013). More important, dependence on soft money deters faculty investigators from pursuing innovative research. One possible solution is for chairpersons to reallocate assets so that a greater proportion of income is dedicated to the investigator’s salary rather than requiring a large percentage of salary covered by grants. When grants are then awarded, more funds are available to perform the actual science. This solution may require hiring fewer faculty so that those remaining will be supported and disposed to mentor other faculty members (Bourne, 2013).
Training of Junior Faculty to Be Resilient
Generally speaking, health care clinicians have a burnout rate of approximately 50% (Decapua, 2017). Less is known about junior investigators in clinical research, who may or may not have associated clinical responsibilities. Primack et al. (2010) reported on burnout of early career investigators, including MDs and PhDs, who were enrolled in a degree-granting or career development program for clinical research (N = 179) at the University of Pittsburgh. Burnout was found in 16% of the respondents. Also, being female, over age 35, or a member of a racial or ethnic minority group was associated with higher burnout. The authors suggested providing greater support for these demographic groups.
Many professionals lack critical self-assessment skills and an understanding of academic political structure. Adjusting to organizational changes, managing conflicts, and building and participating in high-functioning teams are necessary skill sets. Zwack and Schweitzer (2013) described a number of resilience strategies (e.g., continuous self-awareness supported by mindfulness stress-reduction techniques); formal mentorship programs also could help fill these professional gaps. In addition, we found that lower confidence in research skills was associated with thinking about quitting, which speaks to the necessity of adequate training and preparedness for a research career. Junior investigators may be encouraged to craft individual development plans, with help from mentors, to address perceived weaknesses in research skills. Occupational therapy departments often give junior faculty higher course loads than senior faculty, which undoubtedly impinges on scholarly productivity.
Many junior faculty have occupational therapy doctorates, which do not give them the needed skills to become an independent researcher. There are specific programs, however, for health care professionals with primarily clinical backgrounds who wish to conduct research. These include, but are not limited to, clinical research master’s degree programs specifically targeted to clinicians and institutional K career development award programs targeted to a group of clinically trained professionals. In addition, an individual K career development funding mechanism (K08) is targeted to faculty with clinical terminal degrees who wish to become independent investigators. Faculty members with a clinical doctorate may also consider pursuing opportunities to work alongside an individual with a strong track record of securing grant funding and publishing.
Limitations
A limitation of this study is that five participants did not answer the question that addressed quitting research, perhaps because of ambivalence given that committing to an answer may increase discomfort. Another limitation is that these findings come from an academic medical center in the Midwest and may not be generalizable to other institutions. Future studies could survey junior faculty from various types of academic institutions (Research 1 and Research 2 institutions, as well as smaller colleges). Furthermore, a qualitative study could identify specific details about faculty members who have low confidence in their own research skills. Nonetheless, our results do concur with the available occupational therapy literature on faculty research development (Paul, Liu, & Ottenbacher, 2002) as well as that of other disciplines.
Implications for Occupational Therapy Practice
To increase evidence-based scientific practice in our profession, occupational therapy investigators, like junior investigators in other fields, need a variety of supports. The findings from this study have two implications for occupational therapy practice:
Multilevel mentorship and coaching may be useful to support junior occupational therapy investigators and promote resilience.
Occupational therapy chairpersons may consider allocating a greater portion of department funding to the salaries of junior occupational therapy investigators.
Conclusion
AOTA’s (2007) Centennial Vision priorities included increasing research capacity and productivity. The current literature and the results of our study concur that faculty attrition is increasing, which may result in destabilized departments, increased recruitment expenses, and a decrease in research productivity. Faculty attrition is a concern in academic medicine and allied health education. It is of particular significance for the field of occupational therapy because our current body of research is underdeveloped.
Footnotes
Acknowledgments
At the time of this study, the first author was supported by K12 Award HD055931 from the National Institutes of Health, National Center for Medical Rehabilitation Research, National Institute of Child Health and Human Development.